U.S. patent application number 15/597147 was filed with the patent office on 2017-09-07 for scalable local cache in distributed streaming platform for real-time applications.
The applicant listed for this patent is DataTorrent, Inc.. Invention is credited to Gaurav Gupta, Pramod Immaneni, Amol Kekre, Chetan Narsude, Chandi Singh, Thomas Weise, David Yan.
Application Number | 20170257425 15/597147 |
Document ID | / |
Family ID | 55537483 |
Filed Date | 2017-09-07 |
United States Patent
Application |
20170257425 |
Kind Code |
A1 |
Narsude; Chetan ; et
al. |
September 7, 2017 |
Scalable Local Cache in Distributed Streaming Platform for
Real-Time Applications
Abstract
Software for a distributed streaming platform receives an
application. The application is structured as a directed acyclic
graph (DAG) with instances of operators as nodes and streams as
edges between nodes. Multiple instances of an operator access a
shared database. The software receives a pre-defined hint
associated with the application. The pre-defined hint sets a
maximum period of time for local caching of a result from a query
of the database by each of the multiple instances. The software
launches the application by assigning the instances of operators to
one or more containers provided by the streaming platform and
initiating the streams. Each container is associated with a local
cache. The software then receives a request from the application to
make a dynamic adjustment that increases the maximum period of time
for local caching of a result from a query of the database by each
of the multiple instances.
Inventors: |
Narsude; Chetan; (Sunnyvale,
CA) ; Singh; Chandi; (Foster City, CA) ;
Weise; Thomas; (Mountain View, CA) ; Immaneni;
Pramod; (San Jose, CA) ; Yan; David;
(Campbell, CA) ; Gupta; Gaurav; (San Jose, CA)
; Kekre; Amol; (Fremont, CA) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
DataTorrent, Inc. |
San Jose |
CA |
US |
|
|
Family ID: |
55537483 |
Appl. No.: |
15/597147 |
Filed: |
May 16, 2017 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
14205320 |
Mar 11, 2014 |
9654546 |
|
|
15597147 |
|
|
|
|
61776545 |
Mar 11, 2013 |
|
|
|
61838870 |
Jun 24, 2013 |
|
|
|
61957267 |
Jun 25, 2013 |
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
G06F 16/258 20190101;
G06F 11/1402 20130101; H04L 67/10 20130101; H04L 65/60 20130101;
G06F 16/248 20190101; H04L 67/26 20130101; G06F 16/252 20190101;
G06F 16/951 20190101; H04L 29/06027 20130101; H04L 65/605 20130101;
G06F 16/24568 20190101; G06F 9/541 20130101; G06F 11/00
20130101 |
International
Class: |
H04L 29/08 20060101
H04L029/08; H04L 29/06 20060101 H04L029/06 |
Claims
1. A method, comprising the operations of: receiving an application
that runs on a streaming platform, wherein the application is
structured as a directed acyclic graph (DAG) with instances of
operators as nodes and streams as edges between nodes and wherein
multiple instances of an operator access a shared database; setting
a maximum period of time for a local caching of a result from a
query of the database by each of the multiple instances; launching
the application by assigning the instances of operators to one or
more containers provided by the streaming platform, wherein each
container is a process associated with a local cache that is an
allocation of volatile memory created by the streaming platform on
an as needed basis and that performs the local caching; receiving a
request from the application through an application programming
interface (API) to make a dynamic adjustment that increases the
maximum period of time for the local caching of a result from a
query of the database; and executing the dynamic adjustment and
re-launching the application using a recovery policy, wherein each
of the operations is executed by one or more processors in
real-time or near real-time rather than offline.
2. The method as in claim 1, wherein the instances of operators are
program instructions and the streams are unbound sequences of
streaming windows that are ordered in terms of time.
3. The method as in claim 2, wherein each streaming window is an
atomic sequence of sequential structured tuples.
4. The method as in claim 1, further comprising, monitoring a
performance statistic for accesses to the database by each of the
multiple instances and transmitting results of the monitoring to
the application through the API exposed by the streaming
platform.
5. The method of claim 1, wherein the query of the database is
processed by each of the multiple instances.
6. The method of claim 1, wherein said launching further causes
initiation of the streams.
7. A method as in claim 1, wherein the performance statistic
measures one of throughput, or keys read, or keys written.
8. Computer readable media having non-transitory programming
instructions for processing an application, comprising: program
instructions for receiving the application that runs on a streaming
platform, wherein the application is structured as a directed
acyclic graph (DAG) with instances of operators as nodes and
streams as edges between nodes and wherein multiple instances of an
operator access a shared database; program instructions for setting
a maximum period of time for a local caching of a result from a
query of the database by each of the multiple instances; program
instructions for launching the application by assigning the
instances of operators to one or more containers provided by the
streaming platform, wherein each container is a process associated
with a local cache that is an allocation of volatile memory created
by the streaming platform on an as needed basis and that performs
the local caching; program instructions for receiving a request
from the application through an application programming interface
(API) to make a dynamic adjustment that increases the maximum
period of time for the local caching of a result from a query of
the database; and program instructions for executing the dynamic
adjustment and re-launching the application using a recovery
policy, wherein each of the operations is executed by one or more
processors in real-time or near real-time rather than offline.
9. The computer readable media of claim 8, wherein the instances of
operators are program instructions and the streams are unbound
sequences of streaming windows that are ordered in terms of
time.
10. The computer readable media of claim 9, wherein each streaming
window is an atomic sequence of sequential structured tuples.
11. The computer readable media of claim 8, further comprising,
program instructions for monitoring a performance statistic for
accesses to the database by each of the multiple instances and
transmitting results of the monitoring to the application through
the API exposed by the streaming platform.
12. The computer readable media of claim 8, wherein the query of
the database is processed by each of the multiple instances.
13. The computer readable media of claim 8, wherein said launching
further causes initiation of the streams.
14. The computer readable media of claim 8, wherein the performance
statistic measures one of throughput, or keys read, or keys
written.
Description
CLAIM OF PRIORITY
[0001] This application is a continuation of U.S. patent
application Ser. No. 14/205,320, filed on Mar. 11, 2014, entitled
"Scalable Local Cache in Distributed Streaming Platform for
Real-Time Applications", (U.S. Pat. No. 9,654,546, issued on May
16, 2017), and which further claims priority to U.S. Provisional
Patent Application Ser. No. 61/776,545, filed on Mar. 11, 2013,
entitled "Real-Time Streaming Platform for Hadoop" which further
claims priority to U.S. Provisional Patent Application Ser. No.
61/838,870, filed on Jun. 24, 2013, entitled "A Distributed
Streaming Platform for Real-Time Applications" which further claims
priority to U.S. Provisional Patent Application Ser. No.
61/957,267, filed on Jun. 25, 2013, entitled "Distributed Streaming
Platform for Real-Time Applications."
RELATED APPLICATIONS
[0002] This application is related to U.S. patent application Ser.
No. 13/927,108, entitled "Distributed Streaming Platform for
Real-Time Applications", filed on Jun. 25, 2013. This application
is also related to U.S. patent application Ser. No. 13/928,357,
entitled "Dynamic Partitioning of Instances in Distributed
Streaming Platform for Real-Time Applications", filed on Jun. 26,
2013, now issued as U.S. Pat. No. 9,654,538, which issued on May
16, 2017, and U.S. patent application Ser. No. 13/928,351, entitled
"Checkpointing in Distributed Streaming Platform for Real-Time
Applications", filed on Jun. 26, 2013, now issued as U.S. Pat. No.
9,298,788, which issued on Mar. 29, 2016, and U.S. patent
application Ser. No. 13/928,363, entitled "Dynamic Adjustments in
Distributed Streaming Platform for Real-Time Applications", all of
which are incorporated by reference.
[0003] Additionally, this application is related to U.S. patent
application Ser. No. 14/203,551, entitled "Thread-Local Streams in
Distributed Streaming Platform for Real-Time Applications", filed
on Mar. 10, 2014, now issued as U.S. Pat. No. 9,582,365, which
issued on Feb. 28, 2017. And this application is related to U.S.
application Ser. No. 14/205,234, entitled "Formula-Based Load
Evaluation in Distributed Streaming Platform for Real-Time
Applications", filed on Mar. 11, 2014, now issued as U.S. Pat. No.
9,563,486, which issued on Feb. 7, 2017, and U.S. application Ser.
No. 14/205,357, entitled "Partitionable Unifiers in Distributed
Streaming Platform for Real-Time Applications", filed on Mar. 11,
2014, now issued as U.S. Pat. No. 9,648,068, which issued on May 9,
2017.
[0004] The disclosures of all of the applications identified in the
above paragraphs are incorporated herein by reference.
BACKGROUND
[0005] Streaming applications operate on input data which is not
retrieved from persistent storage, but which arrives as one or more
continuous sequence of items. Such input data might be streaming
media such as streaming audio or streaming video. Or such input
data might be other than streaming audio or streaming video, e.g.,
real-time streaming text. Examples of the latter type of input data
include real-time electronic stock tickers published by financial
websites such as Yahoo! Finance, CNBC, Bloomberg, or NASDAQ and
real-time content streams published by websites such as Twitter and
Facebook which leverage interest and/or social graphs.
[0006] As the sources of streaming data proliferate, scalability
has become an issue for streaming applications that process such
data and the platforms which run the streaming applications.
Outside of the area of streaming applications, scalability has been
addressed by distributed batch-processing platforms based on the
Map-Reduce or similar frameworks. However, these platforms
typically operate on input data originating in persistent storage,
e.g., the persistent storage of the commodity servers that make up
a Hadoop cluster. That is to say, in terms of a stock-and-flow
model, these platforms operate on a stock rather than a flow (or
stream).
[0007] Performance is also an issue for streaming applications and
their platforms, since it is often desirable that a streaming
application operate in real time or near real-time. In the past,
streaming applications achieved real-time performance by
sacrificing data integrity or data completeness. For distributed
batch-processing platforms based on Map-Reduce and similar
frameworks, real-time performance is often limited to accessing
(e.g., using Pig, Scalding, Dremel, Drill, etc.) a store of indexed
results that were generated offline.
[0008] Complicating matters still further, streaming applications
tend to be non-stop, almost by definition. And consequently, fault
tolerance is an important issue for streaming applications and the
platforms on which they run.
SUMMARY
[0009] In an example embodiment, a method is described. The method
is executed by one or more processors in real time or near real
time rather than offline. According to the method, software for a
distributed streaming platform receives an application that runs on
the streaming platform. The application is structured as a directed
acyclic graph (DAG) with instances of operators as nodes and
streams as edges between nodes. Multiple instances of an operator
access a shared database. The software then receives a pre-defined
hint associated with the application. The pre-defined hint sets a
maximum period of time for local caching of a result from a query
of the database by each of the multiple instances. The software
launches the application by assigning the instances of operators to
one or more containers provided by the streaming platform and
initiating the streams. Each container is associated with a local
cache. The software monitors a performance statistic for accesses
to the database by each of the multiple instances and transmits
results of the monitoring to the application through an application
programming interface (API) exposed by the streaming platform. The
software then receives a request from the application through the
API to make a dynamic adjustment that increases the maximum period
of time for local caching of a result from a query of the database
by each of the multiple instances. And the software makes the
dynamic adjustment and re-launches the application using a recovery
policy.
[0010] In another example embodiment, an apparatus is described,
namely, computer-readable storage media which persistently store a
program. The program might be software for a distributed streaming
platform. The program is executed by one or more processors in real
time or near real time rather than offline. The program receives an
application that runs on the streaming platform. The application is
structured as a directed acyclic graph (DAG) with instances of
operators as nodes and streams as edges between nodes. Multiple
instances of an operator access a shared database. The program then
receives a pre-defined hint associated with the application. The
pre-defined hint sets a maximum period of time for local caching of
a result from a query of the database by each of the multiple
instances. The program launches the application by assigning the
instances of operators to one or more containers provided by the
streaming platform and initiating the streams. Each container is
associated with a local cache. The program monitors a performance
statistic for accesses to the database by each of the multiple
instances and transmits results of the monitoring to the
application through an API exposed by the streaming platform. The
program then receives a request from the application through the
API to make a dynamic adjustment that increases the maximum period
of time for local caching of a result from a query of the database
by each of the multiple instances. And the program makes the
dynamic adjustment and re-launches the application using a recovery
policy.
[0011] Another example embodiment involves a method. The method is
executed by one or more processors in real time or near real time
rather than offline. According to the method, software for a
distributed streaming platform receives an application that runs on
the streaming platform. The application is structured as a directed
acyclic graph (DAG) with instances of operators as nodes and
streams as edges between nodes. Multiple instances of an operator
access a shared database. The software then receives a pre-defined
hint associated with the application. The pre-defined hint sets a
maximum period of time for local caching of a result from a query
of the database by each of the multiple instances. The software
launches the application by assigning the instances of operators to
one or more containers provided by the streaming platform and
initiating the streams. Each container is associated with a local
cache which is resizable upward or downward. The software monitors
a performance statistic for accesses to the database by each of the
multiple instances and transmits results of the monitoring to the
application through an API exposed by the streaming platform. The
software then receives a request from the application through the
API to make a dynamic adjustment that increases the maximum period
of time for local caching of a result from a query of the database
by each of the multiple instances. And the software makes the
dynamic adjustment and re-launches the application.
[0012] Other aspects and advantages of the inventions will become
apparent from the following detailed description, taken in
conjunction with the accompanying drawings, which illustrates by
way of example the principles of the inventions.
BRIEF DESCRIPTION OF THE DRAWINGS
[0013] FIG. 1 is a diagram of a network with a website hosting a
distributed streaming platform, in accordance with an example
embodiment.
[0014] FIG. 2 is a diagram of a software stack for a distributed
streaming platform, in accordance with an example embodiment.
[0015] FIG. 3 is a diagram showing components of a real-time
streaming application, in accordance with an example
embodiment.
[0016] FIG. 4 is a flowchart diagram that illustrates a process for
launching a streaming application and making dynamic adjustments
based on monitored performance, in accordance with an example
embodiment.
[0017] FIG. 5 is an illustration of the ordered tuples in a
streaming window, in accordance with an example embodiment.
[0018] FIG. 6 is a diagram showing a logical plan, a physical plan,
and an execution plan, in accordance with an example
embodiment.
[0019] FIG. 7A and 7B are diagrams showing examples of the static
partitioning of operator instances in a physical plan, in
accordance with an example embodiment.
[0020] FIG. 8 is a flowchart diagram that illustrates a process for
recovering from a failed container or server, in accordance with an
example embodiment.
[0021] FIG. 9 is a diagram showing several stream modes, in
accordance with an example embodiment.
[0022] FIG. 10 is flowchart diagram that illustrates a process for
dynamically partitioning operator instances, in accordance with an
example embodiment.
[0023] FIG. 11A is a diagram showing the use of dynamic
partitioning of instances to lessen skew resulting from "sticky
key" assignment of tuples, in accordance with an example
embodiment.
[0024] FIG. 11B is a diagram showing the use of a unifier instance
to lessen skew resulting from "sticky key" assignment of tuples, in
accordance with an example embodiment.
[0025] FIG. 11C is a diagram showing the use of cascading unifiers
for more linear scaling, in accordance with an example
embodiment.
[0026] FIG. 12 is a diagram illustrating a stream in a message
queue managed by a container's buffer server, in accordance with an
example embodiment.
[0027] FIG. 13 is a diagram illustrating the flow of tuples in the
streams of an operator instance with two input ports and one output
port, in accordance with an example embodiment.
[0028] FIG. 14A is diagram showing the interactions between a STRAM
and a STRAM Child, in an example embodiment.
[0029] FIG. 14B is a sequence diagram showing the initiation of a
streaming application, in accordance with an example
embodiment.
[0030] FIG. 14C is a diagram showing the ongoing execution of a
streaming application, in accordance with an example
embodiment.
[0031] FIG. 15A is a logical plan for a streaming application that
originates in a stock ticker, in accordance with an example
embodiment.
[0032] FIG. 15B is an execution plan for a streaming application
that originates in a stock ticker, in accordance with an example
embodiment.
[0033] FIGS. 16A to 16E illustrate an application dashboard in a
graphical user interface (GUI) for a distributed streaming
platform, in accordance with an example embodiment.
[0034] FIGS. 17A to 17C illustrate GUI views for debugging an
application running on a distributed streaming platform, in
accordance with an example embodiment.
[0035] FIG. 18 is a flowchart diagram that illustrates a process
for combining two operator instances connected by a stream in a
container, in accordance with an example embodiment.
[0036] FIG. 19 is a sequence diagram that shows the components
involved in combining two operator instances connected by a stream
in a container, in accordance with an example embodiment.
[0037] FIG. 20 is a diagram showing the combination of two operator
instances connected by a stream in a container, in accordance with
an example embodiment.
[0038] FIG. 21 is a diagram showing the call stacks for a combined
operator instance resulting from the combination of two operator
instances connected by a stream in a container, in accordance with
an example embodiment.
[0039] FIG. 22 is a flowchart diagram that illustrates a process
for creating a dynamic partition using a pre-defined hint, in
accordance with an example embodiment.
[0040] FIG. 23 is a sequence diagram that shows the components
involved in creating a dynamic partition using a pre-defined hint,
in accordance with an example embodiment.
[0041] FIG. 24 is a flowchart diagram that illustrates a process
using a scalable local cache in a container and a pre-defined hint,
in accordance with an example embodiment.
[0042] FIG. 25 is a sequence diagram that shows the components
involved in a process using a scalable local cache in a container
and a pre-defined hint, in accordance with an example
embodiment.
[0043] FIG. 26 is a diagram showing a scalable local cache in a
container, in accordance with an example embodiment.
[0044] FIG. 27 is a flowchart diagram that illustrates a process
using partitionable unifiers, in accordance with an example
embodiment.
[0045] FIG. 28A is a diagram showing three use cases that do not
use partitionable unifiers, in accordance with example
embodiments.
[0046] FIG. 28B is a diagram showing a use case that uses
partitionable unifiers, in accordance with an example
embodiment.
DETAILED DESCRIPTION
[0047] In the following description, numerous specific details are
set forth in order to provide a thorough understanding of the
example embodiments. However, it will be apparent to one skilled in
the art that the example embodiments may be practiced without some
of these specific details. In other instances, process operations
and implementation details have not been described in detail, if
already well known.
[0048] FIG. 1 is a diagram of a network with a website hosting a
distributed streaming platform, in accordance with an example
embodiment. As depicted in this figure, a personal computer 102
(e.g., a laptop or other mobile computer) and a mobile device 103
(e.g., a smartphone such as an iPhone, Android, Blackberry, etc.)
are connected by a network 101 (e.g., a wide area network (WAN)
including the Internet, which might be wireless in part or in
whole) with a website 104 hosting a distributed streaming platform.
In turn, website 104 is connected by the network 101 to a website
generating streaming data in real-time (other than streaming audio
or streaming video), such as Yahoo! Finance or Twitter. (In some of
the examples described below, the stock ticker for Yahoo! Finance
is used for illustrative purposes. However, other stock tickers
such as CNBC, Bloomberg, and NASDAQ could easily have been
substituted.) In an example embodiment, personal computer 102 and
mobile device 103 might be used by end users who want to run and/or
view a streaming application (e.g., a GUI dashboard) on website
104.
[0049] In an example embodiment, the website 104 might be composed
of a number of servers connected by a network (e.g., a local area
network (LAN) or a WAN) to each other in a cluster or other
distributed system which might run website software (e.g., web
server software, database software, etc.) and distributed-computing
software. In an example embodiment, the website 105 might also be
composed of a number of servers connected by a network to each
other in a cluster or other distributed system which might run
website software (e.g., web server software, database software,
etc.) and distributed-computing and/or cloud software such as
Hadoop, Hadoop Distributed File System (HDFS), Pig, CloudBase, etc.
The servers at website 105 are also connected (e.g., by a storage
area network (SAN)) to persistent storage 106. Persistent storage
106 might include a redundant array of independent disks (RAID)
and/or flash memory. Persistent storage 106 might be used to store
data related to the data streamed by website 105, e.g., financial
data, content data for social/interest networks, etc.
[0050] In an alternative example embodiment, the servers in website
104 and website 105 and the persistent storage 106 might be hosted
wholly or partially off-site in the cloud, e.g., as a
platform-as-a-service (PaaS) or an infrastructure-as-a-service
(IaaS).
[0051] FIG. 2 is a diagram of a software stack for a distributed
streaming platform, in accordance with an example embodiment. As
depicted in this figure, real-time applications 201 (RT App.sub.1,
RT App.sub.2, etc.) might run on a distributed streaming platform
202, which, in turn, might be integrated with a distributed
computing framework 203 such as YARN, in an example embodiment.
YARN is in the Hadoop family of software but includes functionality
for handling distributed computations that are not structured as
batch jobs for Map-Reduce processing, including distributed
computations that are streaming.
[0052] In turn, the distributed computing framework 203 might be
supported by distributed storage 204, which might be Hadoop
Distributed File System (HDFS), in an example embodiment. And the
distributed computing framework 203 and distributed storage 204
might run on a networked cluster of servers (e.g., commodity
servers) or other hardware computation units (e.g., the hardware
computation units emanating from Facebook's Open Compute
Project).
[0053] FIG. 3 is a diagram showing components of a real-time
streaming application, in accordance with an example embodiment. As
depicted in FIG. 3, such a streaming application might be included
in a specification 301 which might be a Java source code program,
in an example embodiment. Alternatively, specification 301 might be
a Hadoop-style properties file. Specification 301 includes a
logical plan that is a directed acyclic graph (DAG) whose nodes are
operators 302 and whose edges are steams 303. As described in
further detail below, an operator might be a sequence of program
instructions, e.g., to compute a particular statistic. And a stream
might be sequence of streaming windows that contain tuples that (a)
are structured according to a schema and (b) originates in a source
outside of the application program, e.g., a stock ticker or a
content stream broadcast by a social/interest network.
Specification 301 also includes configurable application settings
304 (e.g., with corresponding default values being specified by the
distributed streaming platform), such as streaming window size
(e.g., as measured in terms of units of time or number of tuples),
heartbeat interval or period(e.g., as measured in terms of units of
time or number of streaming windows), frequency of checkpointing
(e.g., as measured in terms of units of time or number of streaming
windows), frequency of purge determinations (e.g., as measured in
terms of units of time or number of streaming windows), aggregate
application window size (e.g., as measured in terms of units of
time or number of streaming windows), sliding application window
size and incrementation (e.g., as measured in terms of units of
time or number of streaming windows), etc., in an example
embodiment. Additionally, specification 301 might include logic 305
(GUI logic) and logic 306 (model or business logic not contained in
the operators).
[0054] In an example embodiment, logic 305 and logic 306 might
include model-view-controller (MVC logic) for displaying the
results of some or all of the operators 302 in a dashboard that is
part of a graphical user interface (GUI). In an example embodiment,
if the origin of the streaming data is a stock ticker, the
dashboard might display statistics related to stock prices and
stock sales. Or if the origin of the streaming data is a content
stream broadcast by a social/interest network, the dashboard might
display statistics related to social signals (e.g., likes,
favorites, shares, etc.) related to posts to the content
stream.
[0055] FIG. 4 is a flowchart diagram that illustrates a continuous
or nonstop process for launching a streaming application and making
dynamic adjustments based on monitored performance, in accordance
with an example embodiment. In an example embodiment, this process
might be performed by the Streaming Application Master (STRAM). In
an alternative embodiment, some or all of the operations of this
process might be performed by the STRAM Childs (e.g., slaves) or
other software in the distributed streaming platform.
[0056] As depicted in FIG. 4, the software (e.g., the STRAM)
receives a specification (e.g., whose location is identified in a
user or script command on a command line interface (CLI)) for an
application that is streaming, in operation 401. In an example
embodiment, the specification might be a source program in Java
created in an integrated development environment (IDE) such as
Eclipse or NetBeans. In another example embodiment, the
specification might be a Hadoop-style properties file. Or the
application might be specified at the CLI, e.g., through user input
such as macros, as described below. In operation 402, the software
converts the specification into a logical plan that includes a
directed acyclic graph (DAG) or other precedence graph with
operators as nodes and streams as edges. One might think of the
logical plan as special form of a data object model (DOM). The
operators are program instructions and the streams are unbound
sequences of streaming windows that are ordered in terms of time.
In an example embodiment, the sequence might originate from
web-services interface, e.g., a web API exposed by Yahoo! Finance
or Twitter accessed by an input adapter for the distributed
streaming platform. It will be appreciated that a Hadoop Map-Reduce
application can be represented as a DAG, though such applications
tend to operate on batches of data rather than streams of data.
[0057] In operation 403, the software translates the logical plan
(e.g., the DAG) into a physical plan using any stream modes
specified in the specification (e.g., in-line, in-node, in-rack, or
other) and with one or more of instances of the operators per the
static partitioning (e.g., as expressed in partition counts) in the
specification. In operation 404, the software obtains a number of
containers (or processes) running on a networked cluster of servers
(or other physical computational units). In an example embodiment,
the software might obtain these containers from the YARN Resource
Manager. One might regard a container as a permission from YARN to
run a process on a particular server (or other physical computation
unit) in the networked cluster. And one might regard the YARN
Resource Manager as a distributed operating system (OS) that
manages processes, memory, and persistent storage for the networked
cluster. One slave (e.g., a STRAM Child) might execute in each
container, in an example embodiment. Then in operation 405, the
software assigns the instances of the operators to the slaves for
execution according to an execution plan that depends on the
physical plan and the number of containers obtained. In operation
406, the software launches the execution plan using input adapters
which convert external input into tuples grouped into streaming
windows. And each slave monitors the throughput of streaming
windows through the instances in its container (e.g., by monitoring
the ordinal streaming window identifiers), along other statistics
related to service level and/or performance in general (e.g.,
latency between servers (e.g., as reported by the container),
network usage (e.g., as reported by the container), CPU usage
(e.g., as reported by the container), memory usage (e.g., as
reported by the container), uptime, errors including data loss
(e.g., as reported by error tuples/ports/streams), size of message
queues managed by buffer servers, throughput other than operator
throughput (such as stream throughput or message-queue throughput),
operator skew, time delay with respect to external system
components, etc.) and reports the results (e.g., in conjunction
with a heartbeat protocol), in an example embodiment. In an example
embodiment, the software might also monitor service level and/or
performance in general using statistics (e.g., latency between
servers, network usage, CPU usage, memory usage, uptime, etc.)
received from or reported by the Resource Manager or other
components of the distributed streaming platform.
[0058] In operation 407, the software makes one or more dynamic
adjustments based on the results of the monitoring (e.g., to reduce
traffic of streaming windows through message queues in containers,
through server NICs, and other bottlenecks or more generally to
improve performance as measured by the reported statistics). In an
example embodiment, the dynamic adjustments might include updating
the physical plan by adding new instances of operators or deleting
existing instances of operators. Or the dynamic adjustments might
update the execution plan by returning a container to the YARN
Resource Manager and/or obtaining a new container from the YARN
Resource Manager for a rack and/or server (or other physical
computational unit) and/or moving instances of operators from one
container or server to another. As described in further detail
below, the making of dynamic adjustments (e.g., failover and
dynamic partitioning) includes re-initiating the streams in
accordance with a recovery policy (e.g., at least once, at most
once, exactly once) through commands (e.g., from the STRAM) to the
slave (e.g., STRAM Child) which, in turn, controls the operators in
the container and the buffer server in the container. Also, in an
example embodiment, the dynamic adjustments might originate from
commands entered by a user or the application at a CLI that
supports runtime modifications to the logical plan (e.g., a macro,
code, an interface or API, a GUI, text input, etc.), the physical
plan, and/or the execution plan.
[0059] In an example embodiment, the distributed streaming platform
might support a macro that is a set of CLI instructions that insert
a sub-DAG (which might be reusable and therefore a candidate for a
library), consisting of multiple operators and streams, into an
application at runtime. The distributed streaming platform might
also support hierarchical operators that are reusable sub-DAGs
which are inserted into logical plans prior to runtime.
[0060] Then in operation 408, the software outputs structured
tuples (e.g., using output adapters) from some or all of the
instances to a display (e.g., a graphical user display or GUI
dashboard for decision support), to persistent storage (e.g., using
HDFS) for subsequent use by another application, to another system,
etc. As noted on FIG. 4, each of the operations in this process
might be executed in real time or near real time rather than
offline, in an example embodiment. Moreover, some of the operations
described in the process, e.g., the monitoring, making dynamic
adjustments, and output operations, might be continuous or non-stop
operations, in an example embodiment.
[0061] In an example embodiment, each container might be a
multi-threaded process that includes one thread for each operator
instance, one thread for the container's buffer server, and one
thread for the slave (e.g., STRAM Child). In an example embodiment,
each container has a single buffer server which manages, for the
operator instances in the container, a message queue (FIFO not
priority) of streaming windows, on a per port basis (e.g., the
buffer server keeps track of which port has read which tuple). Each
of these threads might perform its computations in memory, spilling
over to persistent storage such as HDFS only in the event that
memory is insufficient. It will be appreciated that by using a
single thread for each operator, each operator can execute
asynchronously without creating memory/storage incoherency.
[0062] In an example embodiment, each tuple might be a Plain Old
Java Object (POJO), structured according to a schema or data type.
In an example embodiment, each stream might have one upstream
operator and one downstream operator. In that event, the schema for
a tuple also defines a schema for a stream, e.g., by defining a
schema for the output port of the stream's upstream operator that
is the same as the schema for the input port of the stream's
downstream operator. In an example embodiment, each operator might
have one output port but one or more input ports which are mapped
to the operator's one output port by the logic in the operator. For
example, the input to an instance of an operator that computes an
average price might be a streaming window that consists of a begin
window (e.g., a control tuple with a field for an identifier, say
70), followed by data tuple with a field for a price, say 221.00,
followed by an end window (e.g., a control tuple with a field for
an identifier, also 70). The instance might re-compute an existing
average using the new price to obtain a new average of, say 230.00,
and then output (or emit to an output port) a begin window (e.g.,
with an identifier 70), a data tuple with a field for an average
price set to 230.00, and an end window (e.g., with an identifier
70). It will be appreciated that the tuple input to the instance
might already have a field for an average price which is set to
230.00 by the instance. Or alternatively, the instance might
dynamically allocate the tuple with the field for an average price
and set it to 230.00; the tuple might then be de-allocated by a
downstream operator, e.g., a downstream operator that is an output
adapter that displays the average price of 230.00 in a GUI
dashboard. In an example embodiment, an instance of an operator
might be used to change the schema of a tuple, without making
changes to any values in the schema's fields. In operation 402, the
DAG includes operators that are program instructions. In an example
embodiment, these program instructions might relate to the business
logic for the application, e.g., computing a financial statistic
(e.g., such as the high or low price for a stock within a period of
time) for display in a GUI dashboard for an application fed by a
stock ticker (e.g., through a web API). Or the program instructions
might be more generic, along the lines of the C-runtime library or
the C++ template library. In that regard, a library of reusable
common or standard operator templates (e.g., for use by developers
of applications for the distributed streaming platform) might
include operator templates with functionality for: (1) matching
tuples and emitting results (where the output might be tuples that
matched, tuples that did not match, a Boolean flag, etc.); (2)
recording tuples; (3) counting items such as keys, frequency of
keys, unique counts, etc.; (4) filtering such things as streams
with input schema using keys or rates (e.g., sampling rates); (5)
filtering log file lines from Apache and Tomcat servers; (6)
joining and sorting items; (7) indexing (or mapping, including
hash-mapping) for such operations as search indexing,
word-counting, etc.; (8) consolidating schemas (e.g., to
consolidate multiple streams into one schema); (9) inputting data
into the application (e.g., an input adapter) and outputting data
from the application (e.g., an output adapter), including adapters
using Hadoop Distributed File System (HDFS), MemCache, MySQL,
MongoDB, console, HTTP, Apache ActiveMQ, RabbitMQ, ZeroMQ, Kafka,
Kestrel, Redisetc., Websocket, LocalFile, etc.; (9) performing
mathematical operations such as compare, max, min, average, sum,
quotient, range, except, margin, change, etc.; (10) managing
streams without changing tuples or schema, e.g., stream duplicator,
stream merger, array-list splitter, hash-map splitter,
dev/null/counter, etc.; (11) generate load for testing, e.g., event
generator, random generator, filter event generator, etc.: (12)
computing over application windows that are sliding; (13)
generating data for charts (e.g., in conjunction with CLI macros
that are inserted dynamically at runtime through the CLI); (14)
allow the usage of languages such as Python, JavaScript, Bash,
etc.; (15) issue alerts using SMTP (Simple Mail Transfer Protocol);
and (16) utility functions that are building blocks for other
operator templates, including those listed above. In operation 402,
the DAG includes edges that are streams made up of streaming
windows. In an example embodiment, each streaming window is an
atomic microbatch of sequential tuples that is associated with a
recovery policy for an application. In an example embodiment, the
length of the sequence of tuples in a streaming window is variable
in length, e.g., configurable by the user directly or indirectly; a
streaming window begins with a special "begin window" tuple (e.g.,
a control tuple) and ends with a special "end window" tuple (e.g.,
a control tuple). In an example embodiment, a streaming window
might be specified in terms of time as approximately 0.5 seconds.
An alternative example embodiment might use only a single control
tuple (e.g., "begin window") or some other form of timestamp
ordering for concurrency control (e.g., isolation within the
meaning of the ACID or Atomicity, Consistency, Isolation,
Durability properties for reliable data processing).
[0063] Other control tuples might include checkpoint tuples that
are inserted into the streaming data periodically, per the
checkpointing interval (or frequency) specified by the user or
application (e.g., directly or indirectly through the recovery
policy). In an example embodiment, checkpoint tuples might be
inserted by an input adapter, triggering checkpoints as they work
their way through all of the application's operators, and be
removed by an output adapter. It will be appreciated that
checkpoint tuples can be used to achieve checkpointing at the end
of streaming windows (e.g., align checkpoints with boundaries of
streaming windows).
[0064] In an example embodiment, an instance of an operator might
report errors (e.g., counted per streaming window) using an error
tuple that is emitted through an error port (e.g., an output port
for an error stream) to a log file (e.g., in HDFS). Also, in an
example embodiment, an input adapter might use a sample operator to
perform bucket testing on new application logic on a relatively
small subset of a stream, before deployment to the application.
[0065] In an example embodiment, an application window might be
specified as in terms of streaming windows or using a period of
time. In an example embodiment, an application window specified in
terms of time might range from 5 to 15 minutes. Also, in an example
embodiment, the default application window might be a single
streaming window. An application window is associated with an
operator; thus an application might have multiple application
windows. Also, in an example embodiment, an application might be
either an aggregate application window or a sliding application
window.
[0066] An aggregate application window is constructed by combining
a number of consecutive streaming windows without overlapping. That
is to say, the next application window begins only after the
current application window ends, in an example embodiment.
Aggregate application windows are used for stateless application
operators, e.g., application operators that operate solely on data
in the streaming windows without resort to data read into memory
from persistent storage. It does not follow that the operator
instances in the physical plan are stateless; they are stateful, in
an example embodiment. In an example embodiment, the distributed
streaming platform might enhance performance of an aggregate
application window by using one begin window tuple (e.g., aligned
with the window boundary of the aggregate application window) and
one end window tuple (e.g., also aligned with the window boundary
of the aggregate application window) for all of the streaming
windows in the aggregate application window; that is to say, the
intervening control tuples (e.g., begin window tuples and end
window tuples) might not be processed by the operator associated
with the aggregate application window, though they might be used
for monitoring purposes (e.g., by the STRAM child and/or buffer
server). Also, in an example embodiment, the distributed streaming
platform (e.g., STRAM) might use the last streaming window in an
aggregate application window when making a purge determination as
described in further detail below. An example of an operator that
might be used with an aggregate application window in a financial
application is an operator that charts stock ticker data on a per
minute basis.
[0067] A sliding application window is constructed by combining a
number of consecutive streaming windows with overlapping. That is
to say, the current sliding application window is formed by
dropping a streaming window from the previous sliding application
window and adding a new streaming window, in an example embodiment
(e.g., sliding by one streaming window). Sliding application
windows are used for stateful application operators and the
operator instances in the physical plan are also stateful, in an
example embodiment. An example of an operator that might be used
with a sliding application window in a financial application is an
operator that counts the top 10 trades in terms of volume on a
stock ticker over the past minute (e.g., starting from now).
[0068] Again, the use cases are many, and financial applications
are mentioned because they are a type of process that benefits from
real-time or near real-time processing. So therefore, the types of
applications that can benefit from the processing described herein
can be large, and without limitation and for purposes of example
only, such applications can be for processing technical computing
data, computing statistics, data processing statistics, advertising
statistics, gaming statistics, hospital resource management,
traffic statistics, application load managing, distributed
processing, load balancing of servers and processes, inventory
statistics, data distribution statistics, and other types
technology driven processes.
[0069] Also, in an example embodiment, the recovery policy might be
configurable by the user of the application or the application
itself. Such configuration might occur prior to launch or during
runtime, e.g., through the CLI. The recovery policy might be one of
at least once, at most once, or exactly once, as described in
further detail below. The recovery policy might impact performance
of the distributed streaming platform because it can affect of the
frequency of instance checkpointing, e.g., when the recovery policy
is exactly once, instance checkpointing will occur at the end of
every streaming window. In an example embodiment, instance
checkpointing involves (a) pausing an instance of an operator at
the end of a streaming window, (b) serializing the instance to
persistent storage, e.g., using functionality such as Kryo to
serialize the instance to a file in a file system such as Hadoop
Distributed File System (HDFS), and (c) notifying the STRAM of the
last window completed. Also, in an example embodiment, instance
checkpointing might occur at a specified time period, e.g., every
30 seconds, which might be configurable by the user of the
application, e.g., when the recovery policy is other than exactly
once.
[0070] It will be appreciated that the statefulness of the instance
might determine the amount of data to be serialized, in an example
embodiment. For example, if an operator is stateless (e.g., it
operates solely on the tuples in a streaming window without resort
to data of its own read into memory from persistent storage),
serialization of the operator might be skipped.
[0071] In an example embodiment, the recovery policy might be
specified on a per-operator basis or a per instance basis. That is,
there might be different recovery policies for different operators
or for different instances of the same operator. So, for example, a
stateful instance might have a recovery policy of at least once or
exactly once, whereas a stateless instance might have a recovery
policy of at most once.
[0072] Traditionally, the state of a streaming application is
defined as the state of all operators and the state of all streams.
In an example embodiment, the state of the streaming application
might be defined as the state of all operator instances (e.g., one
or more serializations of the instance obtained through
checkpointing) and the set of all streaming windows in the message
queues maintained by the buffer servers. It will be appreciated
that in such an embodiment, the state of an operator instance is
associated with an identifier for a streaming window. In an example
embodiment, the checkpointing might be asynchronous insofar as the
latest serialization (or checkpoint) for one instance of an
operator might be at the end of a streaming window whose identifier
differs from that of the latest serialization (or checkpoint) for
another instance. Also in an example embodiment, if multiple
serializations are stored for an operator instance, STRAM might
purge earlier serializations on a FIFO basis consistent with the
recovery policy.
[0073] In an example embodiment, the STRAM might dump the current
state of all operator instances (including additions, deletions,
movements to other containers, etc.) to a change file (e.g., in
HDFS). It will be appreciated that the distributed streaming
platform might then use this change file to create an updated
logical plan, which might be used to re-launch the application,
e.g., in the event of a grid outage in the networked cluster. Such
a change file might be updated (a) at a recurring time period that
is configurable by the user or the application, or (b) as a result
of a command at the CLI, e.g., by the user or an application.
[0074] In operation 406, the slaves (e.g., STRAM Childs) might
report the results of their monitoring (e.g., to the STRAM) in
conjunction with a heartbeat protocol. Also, in an example
embodiment, the heartbeat interval or period might be configurable
by the user of the application, e.g., either in terms of units of
time or number of streaming windows. In an example embodiment, the
heartbeat protocol might use YARN RPC (remote procedure call). It
will be appreciated that a heartbeat interval that is too short
might add considerable network traffic and resultant computation to
the distributed streaming platform.
[0075] In any event, the reporting of the results of the monitoring
might be aligned with a streaming window boundary, e.g., through an
end window. That is to say, the reporting of the results might take
place during the period of time between an end window and the next
begin window, in an example embodiment. This period of time might
also be used for restarting operators (e.g., during server
outages), checkpointing, checksumming, and other statistics
generation, etc. In an example embodiment, class-method calls might
be associated with begin window tuples and end window tuples. And
in such an example embodiment, the class-method call for end window
might perform some or all of the reporting of the results of the
monitoring, restarting operators, checkpointing, checksumming, and
other statistics generation.
[0076] It will be appreciated that each streaming window is
identified by an ordinal identifier that increases as the
application runs (e.g., 1, 2, 3, etc.). In an example embodiment,
the results of the monitoring in operation 406 might include (a)
the identifier for the last processed streaming window per operator
in the container, (b) the identifier for the last checkpoint
streaming window per operator in the container, and (c) the
identifier for the committed streaming window. The committed
streaming window is the streaming window that has been computed by
all output adapters (e.g., operators which write to console or
persistent storage). In an example embodiment, the STRAM uses (b)
and (c) to determine which streaming windows can be purged from the
buffer server's message queue in each container and which
checkpoint serializations (or files) can be purged from persistent
storage (e.g., HDFS). In an example embodiment, the user of the
application might configure the period at which this purge
determination is performed by the STRAM, e.g., every 30
seconds.
[0077] In another example embodiment, the results of the monitoring
might also include some statistics related to streams, e.g., tuples
consumed at each input port of an operator per second, tuples
emitted to or by each output port of an operator per second, etc.
Also, each buffer server might also report monitoring results
related to the streaming windows in its message queue, the
identifier of the last streaming window in the message queue,
confirmation of the purge of a streaming window in the message
queue, etc.
[0078] In operation 407, the software re-initiates the streams in
accordance with a recovery policy when making dynamic adjustments.
In an example embodiment, the recovery policy might be one of at
least once, at most once, exactly once, by analogy to the delivery
assurances in the WS-Reliable Messaging Protocol. These recovery
policies are described in greater detail in FIG. 8, in a context
where a dynamic adjustment results from failure of a container or
its server (e.g., failover).
[0079] FIG. 5 is an illustration of the ordered tuples in a
streaming window, in accordance with an example embodiment. In this
figure, time moves along a horizontal axis to the left, so
streaming window 501 (n) is earlier in time than streaming window
502 (n+1). As shown in this figure, a stream (e.g., an edge in the
logical plan) might be a continuous stream of in-order streaming
windows (e.g., 501 and 502), which, in turn, are a continuous
stream of in-order tuples (e.g., tuple 503), which might be thought
of as records, structs, or classes with only data members and no
methods, structured according to a schema or data type. Also shown
in this figure are control tuples 504 (begin window n), 505 (end
window n), 506 (begin window n+1), and 507 (end window n+1), which
might not contain data related to the application, in an example
embodiment.
[0080] FIG. 6 is a diagram showing a logical plan, a physical plan,
and an execution plan, in accordance with an example embodiment.
This figure should be read from top to bottom; that is to say, a
logical plan 601 precedes a physical plan 602, which precedes an
execution plan 603, per the flowchart in FIG. 4. As indicated in
FIG. 6, the application is a financial application whose streaming
data originates in a stock ticker, e.g., Stock Tick Input 1 in each
of the plans. In an example embodiment, a user of the distributed
streaming platform might have input (e.g., provided the location of
the application's files) logical plan 601 through a CLI. The
logical plan includes four operators: (a) Stock Tick Input 1; (b)
Daily Volume 2; (c) Quote 3, and (d) Console 4 (e.g., output to a
display). The distributed streaming platform (e.g., the STRAM)
converts the logical plan 601 into a physical plan 602 by
statically partitioning the operator Daily Volume 2 into three
instances (e.g., per a partition count in the specification):
instance Daily Volume 2_1, instance Daily Volume 2_2, and instance
Daily Volume 2_3, each of which might be a thread. Then the
distributed streaming platform (e.g., the STRAM) connects the three
instances to the upstream instance Stock Input 1 using a stream
duplicator 604 and connects them to the downstream instance Quote 3
using a stream merger (or unifier) 605. The distributed streaming
platform (e.g., the STRAM) then obtains 3 containers (e.g.,
processes) from a resource manager (e.g., a YARN Resource Manager)
and, to complete the execution plan, assigns (a) instance Daily
Volume 2_1 to container 1, (b) instance Stock Tick Input 1,
instance Daily Volume 2_2, instance Quote 3, and instance Quote 4
to container 2, and (c) instance Daily Volume 2_3 to container 3.
Such an execution plan might work well with a stream codec whose
hash function results in a high throughput of tuples through
instance Daily Volume 2_2, since that throughput could avoid
passing through a buffer server's message queue.
[0081] FIG. 7A and 7B are diagrams showing examples of the static
partitioning of operator instances in a physical plan, in
accordance with an example embodiment. As depicted in FIG. 7A, a
logical plan (or DAG) 701 includes four operators: operator 0,
operator 1, operator 2, and operator 3. According to the static
partitioning (e.g., partition counts) in the specification for the
logical plan 701, the distributed streaming platform (e.g., the
STRAM) could partition operator 1 into three instances, 1a, 1b, and
1c, and operator 2 into two instances, 2a and 2b, using one unifier
for the three instances of operator 1 and one unifier for the two
instances of operator 2, when creating the physical plan (or DAG)
702. However, this partition might create a bottleneck at the
unifier for the three instances of operator 1. So the distributed
streaming platform (e.g., the STRAM) instead creates a physical
plan (or DAG) 703, in which there are two unifiers for the three
instances of operator 1. It will be appreciated that such an
approach might also be used for dynamic partitioning based on the
results of the instance monitoring described above.
[0082] FIG. 7B also depicts a logical plan (or DAG) 711. This
logical plan includes five operators: operator 0, operator 1,
operator 2, operator 3, and operator 4. Note that the throughput is
expected to be large on the stream connecting operator 0 to
operator 1 and the stream connecting operator 1 to operator 2, as
indicated by the thickness of the streams. These expected
throughputs might be communicated by the user in the specification,
in an example embodiment. According to the static partitioning
(e.g., partition counts) in the specification for the logical plan
711, the distributed streaming platform (e.g., the STRAM) could
partition operator 1 into two instances, 1a and 1b, using one
unifier for the two instances of operator 1, when creating the
physical plan (or DAG) 712. However, this partition might create a
bottleneck with large throughput from the unifier to operator 2. So
the distributed streaming platform (e.g., the STRAM) instead
creates a physical plan (or DAG) 713 with a parallel partition that
avoids large throughput and bottlenecks. In the parallel partition,
the large throughput out of operator 0 is split between two
branches; one with instances 1a of operator 1, 2a of operator 2,
and 3a of operator 3, and another with instances 1b of operator 1,
2b of operator 2, and 3b of operator 3. The two branches are then
merged through a unifier that outputs a single stream to operator
4. It will be appreciated that such splitting and merging (which is
also referred to as fan-out and fan-in, respectively) might be used
to process a Map-Reduce application that is ultimately written to a
file (e.g., an HDFS file), in an example embodiment. It will also
be appreciated that such an approach might also be used for dynamic
partitioning based on the results of the instance monitoring
described above.
[0083] In an example embodiment, the distributed streaming platform
(e.g., the STRAM) might ignore all or part of the static partition,
treating it as a hint rather than a command. In that event, the
software might issue a diagnostic message (e.g., through the CLI or
to a log file) to the user who submitted the application.
[0084] FIG. 8 is a flowchart diagram that illustrates a process for
recovering from a failed container or server, in accordance with an
example embodiment. It will be appreciated that such recovery
(e.g., failover) is a form as dynamic adjustment based on
monitoring results, as described above. In an example embodiment,
this process might be performed by the STRAM. In an alternative
embodiment, some or all of the operations of this process might be
performed by the STRAM Childs or other software in the distributed
streaming platform.
[0085] As depicted in FIG. 8, the software (e.g., the STRAM)
determines that a container or its server has failed (e.g., based
on notification from YARN Resource Manager, a time out on heartbeat
to slave, an exception repeatedly thrown by an instance in the
container, etc.), in operation 801. In operation 802, the software
obtains a new container (e.g., from the YARN Resource Manager) and
assigns instances of operators to the container per the original
execution plan or per an updated execution plan based on the
monitoring results reported by slaves (e.g., STRAM Childs).Then in
operation 803, the software restarts the application according to
the following recovery policies: (A) if the recovery policy is at
most once (e.g., data loss is acceptable), then the software
instructs (e.g., through the STRAM Child) each instance in the new
container to subscribe to the next streaming window in the upstream
buffer server's message queue (or, alternatively, instructs the
upstream buffer server through a STRAM Child to transmit that
streaming window to each of those instances); (B) if the recovery
policy is at least once (e.g., data loss is not acceptable but
extra computations are), then (1) the software determines the
latest viable checkpoint for each instance in the new container
using streaming window identifiers of checkpoints for that instance
and for downstream instances (e.g., the streaming window identifier
of the latest viable checkpoint is less or older than the streaming
window identifiers of the checkpoints for the downstream
instances); (2) restarts (e.g., through the STRAM Child) each
instance in the new container using the latest viable checkpoint
and restarts (e.g., through the STRAM Childs) the downstream
instances using each of their latest checkpoints (e.g., with each
of the downstream instances being instructed to subscribe to the
streaming window in each of their upstream buffer server's message
queues with a streaming window identifier greater or newer than
each of their latest checkpoints), and (3) instructs (e.g., through
the STRAM Child) each instance in the new container to subscribe to
the streaming window in the upstream buffer server's message queue
with a streaming window identifier that is greater or newer than
the streaming window identifier of the latest viable checkpoint; or
(C) if the recovery policy is exactly once (e.g., data loss is not
acceptable and neither are extra computations), then the software
restarts (e.g., through the STRAM Child) each instance in the new
container using the last checkpoint (its streaming window
identifier will be less or older by one) and instructs (e.g.,
through the STRAM Child) each of those instances to subscribe to
the streaming window in the upstream buffer server's message queue
that was lost (not checkpointed).
[0086] It will be appreciated that the recovery policy of at most
once can be processed faster than a recovery policy of at least
once and exactly once, at the cost of data loss. And while a
recovery policy of exactly once might be processed faster than a
recovery policy of at least once, the former recovery policy might
significantly impact performance of the distributed streaming
platform since it requires the checkpointing of an instance at the
end of every streaming window.
[0087] It will also be appreciated that operations 802 and 803
might also be used during other dynamic adjustments as described
above, e.g., adding new containers to an updated execution plan,
based on monitoring results from the slaves or a command (e.g., a
macro) entered by a user or script at the CLI while the application
is continuously executing.
[0088] FIG. 9 is a diagram showing several stream modes, in
accordance with an example embodiment. Recall that in an example
embodiment, each stream might be connected to one output port for
an upstream operator and one input port for a downstream operator.
When the stream mode is in-line 901 (shown in FIG. 9 as the thick
arrow), the operator instances (each of which might be a single
thread) connected by the stream are in the same container (or
process). And consequently, the streaming windows going between the
output port (of the upstream operator on the stream) and the input
port (if the downstream operator on the stream) need not pass
through the message queue managed by the container's buffer server.
It will be appreciated that such a mode facilitates high throughput
of streaming windows.
[0089] When the stream mode is in-node 902 (shown in FIG. 9 as the
arrow with medium thickness), the streaming windows going between
the output port (of an upstream operator on the stream) and the
input port (of the downstream operator on the stream) pass through
the message queue managed by the container's buffer server. Passing
through the message queue might entail serialization of tuples into
bytes at the output port (of the upstream operator on the stream)
and de-serialization of bytes into tuples at the input port (of the
downstream operator on the stream), according to a stream codec
(e.g., for stream sockets or other network sockets) such as Kryo.
Consequently, the throughput of streaming windows when the
stream-mode is in-node will be lower than the throughput of
streaming windows when the stream-mode is in-line.
[0090] When the stream mode is in-rack 903 (shown in FIG. 9 as the
thin arrow), the streaming windows going between the output port
(for an upstream operator on the stream) and the input port (for
the downstream operator on the stream) pass through both the
message queue managed by the container's buffer server and a
network interface controller (NIC) or other hardware component that
connects one server (or other physical computation unit) with
another. Consequently, the throughput of streaming windows when the
stream-mode is in-rack will be significantly lower than the
throughput of streaming windows when the stream-mode is in-line or
in-node. And when the steam mode is "other" (not shown in FIG. 9),
the throughput of streaming windows might be significantly lower
than the throughput of streaming windows when the stream-mode is
in-line, in-node, and in-rack, in an example embodiment.
[0091] In an example embodiment, the software might ignore some or
all of the stream modes, treating them as a hints rather than
commands. For example, a user or code might submit a specification
in which all streams are specified as in-line in order to obtain
fast throughput of streaming windows, though such an approach would
result in a process that exceeded the capacity of a container. In
that event, the software might issue a diagnostic message (e.g.,
through the CLI or to a log file) to the user who submitted the
application.
[0092] FIG. 10 is flowchart diagram that illustrates a process for
dynamically partitioning operator instances, in accordance with an
example embodiment. In an example embodiment, this process might be
performed by the STRAM. In an alternative embodiment, some or all
of the operations of this process might be performed by the STRAM
Childs or other software in the distributed streaming platform.
[0093] As depicted in FIG. 10, the software (e.g., the STRAM)
determines that an instance of an operator in a container is
creating a bottleneck, based on monitoring results received from
container's slave, in operation 1001. For example, the instance
might be an upstream instance to a downstream instance with two
input ports. And the input port connected to the upstream instance
might have significantly lower throughput than the other input
port. In operation 1002, the software pauses the instance of the
operator, e.g., after its last checkpointing (e.g., serialization
to HDFS). The software then assigns multiple instances of the
operator to the container and connects the instances to the
upstream operators, e.g., using a stream duplicator, and to the
downstream operator, e.g., using a stream merger or a unifier, in
operation 1003. In operation 1004, the software starts the
instances of the operator (e.g., through the slave), using the last
checkpoint and a recovery policy (e.g., at most once, at least
once, or exactly once), as explained in detail above with respect
to operation 803 in FIG. 8.
[0094] In an example embodiment, a stream codec might be used to
split the tuples in a streaming window between multiple instances
of the same operator that result from static partitioning in the
specification or dynamic partitioning at run-time. For example, a
hash function might be applied to a tuple to obtain a hash code and
the lower bits of the hash code might determine which instance of
the operator receives the tuple (e.g., if there are two instances
the lower bit of the hash code would be sufficient to split the
tuples between the instances). It will be appreciated that such an
approach (which might be referred to as "sticky key") differs from
a round-robin approach, where the first tuple would go to the first
instance, the second tuple would go to the second instance, the
third tuple would go to the first instance, the fourth tuple would
go to the second instance, etc.
[0095] In such an example embodiment, the "sticky key" approach
might result in a skewed distribution, where one instance of the
same operator receives many more tuples than the other instances,
e.g., 4 tuples received by the one instance to every 1 tuple
received by each of the other instances. In that event, the STRAM
might lessen the skew by applying at runtime a ratio (e.g., 2:1) of
maximum load to minimum load, as configured by the user of the
application or the application itself (e.g., through the CLI). In
that event, the one instance receiving more tuples would receive at
most 2 tuples for every 1 instance received by each of the other
instances.
[0096] FIG. 11A is a diagram showing the use of dynamic
partitioning of instances to lessen skew resulting from "sticky
key" assignment of tuples, in accordance with an example
embodiment. In this example embodiment, the logical plan 1101 is
translated into an execution plan 1102, in which there are three
instances of operator 1, namely, 1a, 1b, and 1c. Because of the
"sticky key" assignment of tuples, instance 1c is emitting 60% of
the tuples received by the lone instance of operator 2 and
instances 1a and 1b are each emitting 20% of the total number of
tuples received by the lone instance of operator 2. Such a skew
might be the result of spiking caused by a hash function tied to
directly or indirectly to a geographic location (e.g., the IP
address of users) in another time zone; e.g., a spike that results
from Internet users waking up in the morning in Russia. If the skew
rule (e.g., as configured by the user) is that no instance shall
emit more than 50% of the total number of tuples received by
another instance, the STRAM might enforce the skew rule through an
execution plan 1103 that merges instance 1a and 1b into a single
instance (1a+1b) and splits instance 1c into two instances, 1ca and
1cb. It will be appreciated that this execution plan preserves the
partition count (e.g., three instances of operator 1) in the
original execution plan 1102.
[0097] FIG. 11B is a diagram showing the use of a unifier instance
to lessen skew resulting from "sticky key" assignment of tuples, in
accordance with an example embodiment. Here again, the logical plan
1111 is translated into an execution plan 1112, in which there are
three instances of operator 1, namely, 1a, 1b, and 1c. Because of
the "sticky key" assignment of tuples, instance 1c is emitting 60%
of the tuples received by the lone instance of operator 2 and
instances 1a and 1b are each emitting 20% of the total number of
tuples received by the lone instance of operator 2. If the skew
rule is again that no instance shall emit more than 50% of the
total number of tuples received by another instance, the STRAM
might enforce the skew rule through an execution plan 1113 that
splits instance 1c into two instances, 1ca and 1cb, and later
merges the streams from those instances using a unifier that might
include special code for handling spikes (e.g., a leaky bucket
algorithm). It will be appreciated that this execution plan does
not preserve the partition count (e.g., three instances of operator
1) in the original execution plan 1112.
[0098] FIG. 11C is a diagram showing the use of cascading unifiers
for more linear scaling, in accordance with an example embodiment.
In an example, the logical plan 1121 is translated into an
execution plan 1122, in which there are four instances of the
upstream operator (uopr1, uopr2, uopr3, and uopr4) and one instance
of the downstream operator (dopr), where N is the number of
instances of the upstream operator and M is the number of instances
of the downstream operator. However, in execution plan 1122, all
four of the upstream operators emit streams that pass through a NIC
to another container, where a unifier with special code merges the
streams for the downstream operator. Such an execution plan might
result in a bottleneck forming at the NIC and/or at the container
in which the unifier runs. To prevent such a bottleneck, the STRAM
might use cascading unifiers consisting of two or more levels of
unifiers. In execution plan 1123, there are two levels (K equals
2); the first level contains two containers, each with its own
unifier, and the second level contains one container with one
unifier.
[0099] FIG. 12 is a diagram illustrating a stream in a message
queue managed by a container's buffer server, in accordance with an
example embodiment. In an example embodiment, each container might
be a multi-threaded process with one thread for the slave, one
thread for each instance of an operator, and one thread for each
buffer server. As depicted in FIG. 12, the message queue 1201 is a
FIFO (not priority) queue, where the oldest complete streaming
window in the stream is window n at the bottom of the queue and the
newest complete streaming window in the stream is window n+2
towards the top. In an example embodiment, message queue 1201 might
be based on a publisher-subscriber model, where one output port
(writer port) writes streaming windows for a stream into the
message queue and multiple input ports (read port1, read port2, and
read port3) read from the message queue by subscribing to the
stream from a particular streaming window "onwards" (e.g., in term
of n incrementing). Thus, in message queue 1201, read port3 might
be subscribing to the stream from streaming window n onwards (e.g.,
n, n+1, n+2, n+3, etc.), read port2 might be subscribing to the
stream from streaming window n+1 onwards (e.g., n+1, n +2, n+3,
etc.), and read port1 might be subscribing to the stream from
streaming window n+3 onwards.
[0100] In an example embodiment, security for the distributed
streaming platform might be provided by Kerberos, where the access
points are the STRAM and each of the buffer servers. In that
embodiment, the STRAM might obtain a security token and pass it to
the STRAM Child (e.g., a thread), which, in turn, passes it to the
buffer server (e.g., also a thread) that it monitors and controls
in their shared container (or process). The buffer server could
then use the security token to verify the security of any new
connection to the container. Also, in an example embodiment,
security for the distributed streaming platform's graphical user
interfaces (GUIs) might be provided by Simple and Protected GSSAPI
Negotiation Mechanism (SPNEGO).
[0101] In an example embodiment, a reservoir buffer (e.g., a
thread) associated with an instance of an operator might be used to
synchronize streaming windows for operator instances with multiple
input ports. In an example embodiment, the reservoir buffer might
monitor the input ports to determine when a begin window tuple
(e.g., a control tuple) with a new window identifier has been
received by one of the input ports. The reservoir buffer might then
emit a begin window tuple with that window identifier on the output
port for the instance (e.g., using the container's message queue or
another FIFO queue), in an example embodiment. But the reservoir
buffer might emit an end window tuple with that window identifier
on the output port for the instance (e.g., using the container's
message queue or another FIFO queue) only after the reservoir
buffer determines that an end window tuple with that identifier has
been received by all of the input ports for the instance. It will
be appreciated that in such an example embodiment, the instance's
propagation of a begin window tuple (and the processing of the data
tuples that follow the begin window tuple) is non-blocking with
respect to the instance, whereas the instance's propagation of an
end window tuple is blocking with respect to the instance (except,
in an example embodiment, when performing the operations in a
recovery policy as described below). Further, the completion of a
streaming window in an instance of an operator (e.g., propagation
of an end window through an instance, in an example embodiment)
only occurs after all upstream instances have finished processing
the streaming window. That is, the window identifier of the
streaming window in an upstream instance is greater than or equal
to the window identifier of the streaming window in the instance
and the window identifier of the streaming window in a downstream
instance is less than or equal to the window identifier of the
streaming window in the instance. Also, in an example embodiment,
the reservoir buffer might merge data tuples received on multiple
input ports from different instances of the same upstream operator
into a single queue (e.g., using the container's message queue or
another FIFO queue), through a "first come-first served" approach
to aggregation. This is illustrated in the following figure.
[0102] FIG. 13 is a diagram illustrating the flow of tuples in the
streams of an operator instance with two input ports and one output
port, in accordance with an example embodiment. As shown in this
figure, the first input port (i1) receives a stream1 (1301) whose
earliest tuples are bw1 (begin window 1) and t11 and whose latest
tuple is ew1 (e.g., end window 1). The tuples bw1 and t11 arrive
before any tuples from stream2 (1302), which is received by the
second input port (i2). The earliest tuples in stream2 (1302) are
bw2 (begin window 2) and t21 and the latest tuple in this stream is
ew2 (end word 2). In an example embodiment, the operator processes
tuples using a "first come-first served" approach resulting in a
FIFO message queue 1303 for the output port (o1) where all of the
tuples from the second stream are enqueued (e.g., processed) before
the last three tuples of the first stream, even though it arrived
first. Note that during the processing, the operator removed
control tuples bw1, bw2, ew1, and ew2 and inserted in their stead
control tuples bw and ew. In an example embodiment, the operator in
this figure might be used for aggregating (or merging) streams from
two instances of the same operator.
[0103] FIG. 14A is diagram showing the interactions between a STRAM
and a STRAM Child, in an example embodiment. As shown in this
figure, the STRAM 1401 is the master of the STRAM Child 1402, its
slave. In turn, the STRAM Child 1402 is the master of the
instances, namely, instance 1403 (Instance A) and instance 1404
(Instance B). Additionally, the STRAM Child 1402 is the master of
the Buffer Server 1405, which manages a message queue (FIFO, not
priority) of streaming windows. Each slave (e.g., STRAM Child 1402
or instance 1404, respectively) reports monitoring data to its
master (e.g., STRAM 1401 or STRAM Child 1402, respectively), which
then makes dynamic adjustments which might be effected by the slave
(e.g., STRAM Child 1402 or instance 1404, respectively). Also, as
shown in this figure, STRAM Child 1402, instance 1403 (Instance A)
and instance 1404 (Instance B), and Buffer Server 1405 might each
be a single thread executing in a multithreaded process (or
container), in an example embodiment.
[0104] FIG. 14B is a sequence diagram showing the initiation of a
streaming application, in accordance with an example embodiment. As
depicted in this figure, a streaming application is initiated by a
command (e.g., from a user or a script identifying the location of
the application's files in a file system such as HDFS) received at
a command line interface (CLI), which might be a wrapper for a web
service, in an example embodiment. The CLI communicates the command
to a resource manager (RM), which might be a YARN Resource Manager,
in operation 1. The RM then launches a streaming application
manager (STRAM) in a container in operation 2, to start the
application. The STRAM compiles the application (which includes a
logical plan), produces a physical plan (from the logical plan)
that partitions the operators into instances, obtains the
containers for the physical plan from the RM, and determines an
execution plan based on the physical plan and containers obtained.
The STRAM then executes the execution plan, which creates one Stram
Child in each container and which assigns the instances of the
operators to the containers, in operation 3. Two input adapters
then start the streams that feed the instances in the containers.
Some of the streams are inline streams between instances in the
same container. These streams avoid the overhead of a message queue
managed by a buffer server. Some of the streams are node-local
streams between instances in different containers on the same
server. These streams pass though a message queue managed by a
buffer server, but avoid the overhead of server NICs. And some of
the streams are between instances on different servers, which incur
the overhead of both the message queue managed by buffer server and
a server NIC. The computations performed by the instances culminate
in a stream that is output by a single output adapter, e.g., to a
GUI displayed by a console or a file in a file system (e.g.,
HDFS).
[0105] FIG. 14C is a diagram showing the ongoing execution of a
streaming application, in accordance with an example embodiment. As
shown in this figure, an application 1411 communicates through a
CLI with a STRAM 1413, that monitors and controls threes slaves
(e.g., STRAM Child 1414, STRAM Child 1415, and STRAM Child 1416).
In an example embodiment, each of the slaves executes in its
container. As shown in the figure, a server in a networked cluster
might have multiple containers. In order to make dynamic
adjustments to the application through the slaves, the STRAM 1413
obtains resources (e.g., containers) from Resource Manager 1412.
STRAM Child 1414 monitors and controls an input adapter, which
receives a stream of streaming data from a source over the Internet
and inserts control tuples into the streaming data, in an example
embodiment. STRAM Child 1415 monitors and controls instances that
perform computations on the streaming data. And STRAM Child 1416
monitors and controls an output adapter that removes the inserted
control tuples and outputs the resultant streaming data to the
application's display, in an example embodiment.
[0106] FIG. 15A is a logical plan for a streaming application that
originates in a stock ticker, in accordance with an example
embodiment. Noted again for clarity, stock ticker data is only one
example and other type of data that is not financial in nature can
also be analyzed. Now, as shown in this figure, the logical plan
includes operator 1 (Stock Tick Input) that inputs streams of data
that include time, price, and volume into the application from an
external source, e.g., a website such as Yahoo! Finance. Operator 1
(Stock Tick Input) transmits the volume stream to operator 2 (Daily
Volume), which computes a stream of daily-volume data and transmits
it to operator 3 (Quote), which also receives the time stream and
the price stream from operator 1 (Stock Tick Input). Operator 3
(Quote) computes a stream of quote data and transmits it to
operator 4 (Console), e.g., for display in a GUI. The GUI can be of
any type of device, such as a desktop computer, a laptop computer,
a portable device, a smartphone, a tablet computer, or any device
that can present, display or render the data for the GUI.
[0107] Operator 1 (Stock Tick Input) also transmits the price
stream to operator 5 (High Low), which computes a stream of
high-low price data and transmits it to operator 7 (Chart).
Operator 1 (Stock Tick Input) also transmits the volume stream to
operator 6 (Minute Vol), which computes a stream of
volume-per-minute data and transmits it to operator 7 (Chart).
Operator 7 (Chart) computes a stream of chart data and transmits it
to operator 8 (Console), e.g., for display in a GUI. Operator 1
(Stock Tick Input) also transmits the price stream to operator 9
(SMA or simple moving average), which computes a stream of
sma-price data and transmits it to operator 10 (Console), e.g., for
display in a GUI.
[0108] FIG. 15B is an execution plan for a streaming application
that originates in a stock ticker, in accordance with an example
embodiment. This figure shows an execution plan for the logical
plan described in FIG. 15A. As depicted in FIG. 15B, the STRAM is
operating in its own container at the bottom left of the figure.
Pursuant to the execution plan, the STRAM has assigned instance 1
(Stock Tick Input) to its own container, where it is monitored by a
STRAM Child, which reports its results to the STRAM for dynamic
adjustment (e.g., through the STRAM Child, which, in turn, controls
the instances and the buffer server). Also pursuant to the
execution plan, the STRAM has assigned instance 2 (Daily Volume),
instance 3 (Quote), and instance 4 (Console) to one container,
where they are jointly monitored by a STRAM Child, which reports
its results to the STRAM for dynamic adjustment. The STRAM has
assigned instance 5 (High Low), instance 6 (Minute Vol), instance 7
(Chart), and instance 8 (Console), where they are jointly monitored
by a STRAM Child, which reports its results to the STRAM for
dynamic adjustment. And the STRAM has instance 9 (SMA), and
instance 10 (Console) to one container, where they are jointly
monitored by a STRAM Child, which reports its results to the STRAM
for dynamic adjustment. It will be appreciated that the execution
plan shown in FIG. 15B makes use of few inter-container streams,
since such streams incur overhead associated with the transmission
of the stream through the upstream container's message queue
managed (e.g., managed by a buffer server). Inter-server streams
incur an even greater overhead associated with the transmission of
the stream through NICs.
[0109] In an example embodiment, operator instances in physical
plans and execution plans might be identified as integers, rather
than strings, chars, etc., for purposes of performance
efficiency.
[0110] FIGS. 16A to 16E illustrate an application dashboard in a
graphical user interface (GUI) for a distributed streaming
platform, in accordance with an example embodiment. As depicted in
FIG. 16A, a dashboard displayed by the distributed streaming
platform might include a GUI view 1601 that includes a list 1602 of
the application instances being run by the distributed streaming
platform. It will be appreciated that the distributed streaming
platform might support multiple tenants (e.g., application
instances), in an example embodiment. GUI view 1601 also includes
GUI controls 1603, 1604, and 1605. If a user checks the checkbox
next to one of the application instances in list 1602 and clicks
control 1603 (labeled "inspect"), the application dashboard might
display a view that shows data for that application instance, such
as the view shown in following figure. If a user checks the
checkbox next to one of the application instances in list 1602 and
clicks control 1604 (labeled "kill"), the application dashboard
might stop the application instance, if it is running. And if a
user checks the checkbox next to one of the application instances
in list 1602 and clicks control 1605 (labeled "relaunch"), the
application dashboard might re-launch the application instance, if
it has been killed.
[0111] FIG. 16B shows a GUI view 1611 that might be displayed when
a user enters a command to inspect an application instance, e.g.,
using control 1603 in FIG. 16A. The GUI view 1611 displays data for
an application instance identified by application name 1612 (whose
value is
"com/mailhartech/demos/chart/YahooFinanceApplication.class"). The
data displayed for the application includes the last window
identifier 1613 (whose value is 8860), the number 1614 of
containers (whose value is 5) and the period 1615 of time that the
application has been running (3 days, 10 hours, 3 minutes). GUI
view 1611 also displays a list 1616 of the 5 containers used by the
application, a chart 1617 of the operators used by the application,
and graph 1618 of metrics related to the application. As indicated
by the GUI controls in the toolbar 1619 in GUI view 1611, the
application dashboard can be customized by the user through the
addition of new widgets to an existing dashboard or the creation of
a new dashboard. FIG. 16C shows a close-up diagram of the chart
1617 and the graph 1618 in FIG. 16B.
[0112] FIG. 16D shows a GUI view 1621 that might be displayed when
a user enters a command to inspect an operator instance, e.g., by
clicking on an operator instance in FIG. 16B. GUI view 1621
includes table 1622 and a graph 1627. The table 1622 might include
a name 1623 for the operator instance (whose value is
"UniqueURLCounter"), a container identifier 1624 (whose value is
7), a current window 1625 (whose value is 1046), and a recovery
window 1626 (whose value is 999). The graph 1627 displays
performance metrics for the operator instance, e.g., Emitted/sec,
Processed/sec, Percentage of CPU, and Latency. FIG. 16E is a
close-up diagram of table 1622. The table 1622 includes a list of
ports which shows an input port 1628 with tuples named "data" and
an output port 1629 with tuples named "count". The table 1622 also
includes a recording table 1630, which facilitates debugging as
explained further below.
[0113] FIGS. 17A to 17C illustrate GUI views for debugging an
application running on a distributed streaming platform, in
accordance with an example embodiment. FIG. 17A includes a GUI view
1701 that might be displayed when a user selects an operator from
the list 1617 of operators shown in FIG. 16C. As shown in FIG. 17A,
the selection of an operator instance results in the display of two
GUI controls, a control 1702 labeled "inspect" that allows a user
to see further data regarding the selected instance and a control
1703 labeled "start recording" that allows a user to record the
data related to the processing of tuples by the selected
instance.
[0114] FIG. 17A also includes a GUI view 1704 which might be
displayed (e.g., as part of table 22 in FIG. 16E) when a user
clicks control 1703 in GUI view 1701. GUI view 104 includes three
controls: (a) a control 1705 labeled "view tuples"; (b) a control
1706 labeled "stop recording"; and (c) a control 1707 labeled
"refresh list". If a user clicks on the control labeled 1705, the
distributing steaming platform might display the GUI view 1708
shown in FIG. 17B.
[0115] GUI view 1708 is a tuple viewer associated with a recording
whose name 1709 is
"container_1370542662205_0007_01_000007_3_1370896978891". GUI view
1708 also includes a window identifier 1710 (whose value is 8816)
that shows the streaming window that is the source of the tuples
shown in the GUI view and processed by the selected instance (e.g.,
"operator 3" as shown at the top left of the view). The tuples
themselves are shown as a scrollable stream 1711, with the earliest
tuple at the top of the stream and the latest tuple at the bottom
of the stream. Tuple 1712 is an input tuple ("data" in terms of
table 1622) whose identifying number is 5804 and whose value 1713
is a URL, namely, "http://twcm.me/MLwbd". Tuple 1714 is an output
tuple ("count" in terms of table 1622) whose identifying number is
5808 and whose value 1715 includes "http://twcm.me/MILwdb": "100",
which is a count of the number of times that the URL
"http://twcm.me/MLwbd" has been seen. FIG. 17C shows GUI view 1704
after a user has clicked control 1706 to stop the recording of data
related to tuples processed by the selected instance.
[0116] FIG. 18 is a flowchart diagram that illustrates a process
for combining two operator instances connected by a stream in a
container, in accordance with an example embodiment. In an example
embodiment, this process might be performed by the Streaming
Application Master (STRAM). In the same or an alternative
embodiment, some or all of the operations of this process might be
performed by a STRAM child (e.g., a slave) or other software in the
distributed streaming platform.
[0117] As depicted in FIG. 18, the software (e.g., the STRAM)
launches a real-time streaming application that runs on a
distributed streaming platform, in operation 1801. In an example
embodiment, the real-time streaming application might be structured
as a directed acyclic graph (DAG) with instances of operators as
nodes and streams as edges between nodes, as described earlier. In
operation 1802, the software receives an indication that a stream
is I/O bound, where a stream connects one operator instance to
another operator instance in a single container provided by the
distributed streaming platform. In an example embodiment, the
indication is a platform measurement as to resource use, e.g.,
throughput of streaming windows or tuples for the stream. Or the
indication might be a platform measurement as to a pre-defined
hint, as described in further detail below. Then in operation 1803,
the software transmits the platform measurement to the real-time
streaming application (e.g., through an API exposed by the
distributed streaming platform) and receives a request (e.g., a
request setting an attribute of a stream) from the real-time
streaming application to combine the operator instances connected
by the stream into a single operator instance. And in operation
1804, the software creates a single operator instance (e.g.,
through compiler in-lining) and re-initiates the stream (e.g.,
using a recovery policy such as at least once, at most once,
exactly once, etc.).
[0118] In an example embodiment, each of the operations in this
process might be executed in real time or near real time rather
than offline. Moreover, some of the operations described in the
process (e.g., the operations following the original launch of the
real-time streaming application) might be continuous or non-stop
operations, in an example embodiment.
[0119] In operation 1802, the software receives an indication that
a stream is I/O bound. Recall that in an example embodiment, an
operator instance might be a thread and a server buffer managing a
message queue for stream might be a thread. It will be appreciated
that the term "I/O bound" is a term of art in the field of computer
science and refers to a condition where the time period to complete
a computation (e.g., the operations performed by a thread) is
primarily determined by time spent waiting for read/write
operations to be completed, e.g., reading/writing to storage (e.g.,
to volatile storage such as main memory or to persistent storage
such as a hard disk or flash memory). By contrast, the term "CPU
bound" refers to a condition where the time period to complete a
computation (e.g., the operations performed by a thread) is
primarily determined by time spent processing data. It will be
appreciated that a stream might become I/O bound, for example, if
the streaming windows of tuples processed by a thread involved in
the stream (e.g., an operator instance or a server buffer) spill
over from main memory to persistent storage. In an example
embodiment, the software might determine that a stream is I/O bound
if throughput increases as a result of a partition of one or both
of the stream's operator instances, e.g., when the software
performs automated sensitivity analysis with respect to
partitioning and throughput. It will be appreciated that such
partitioning might be avoided by combining the operator instances
connected by a stream into a single operator instance, which, in
turn, might free up resources such as containers.
[0120] In operation 1803, the software from the real-time streaming
application receives a request to combine the operator instances
connected by the stream into a single operator instance. In an
example embodiment, the request might involve the setting of an
attribute on the stream, e.g., STREAM_LOCALITY=ThreadLocal. In the
same or other embodiments, the request might take the form of a
change to the stream mode as described earlier, e.g., from in-node
to in-line or from in-rack to in-line. In FIG. 18, the request
occurs at run-time, e.g., after the launch of the real-time
streaming application. Alternatively, the request might occur prior
to launch, in the logical plan (e.g., the specification) for the
real-time streaming application.
[0121] FIG. 19 is a sequence diagram that shows the components
involved in combining two operator instances connected by a stream
in a container, in accordance with an example embodiment. It will
be appreciated that this figure involves operations similar to
those depicted in FIG. 18. As shown in FIG. 19, there might be four
components involved in combining two operator instances connected
by a stream in a container: (a) a real-time streaming application;
(b) a STRAM; (c) a STRAM child; and (d) the operator instances. In
operation 1, which is earliest in time, the STRAM sets the stream
mode between two operator instances connected by a stream in a
container to be in-node, e.g., per the specification for the
real-time streaming application. In operation 2, the STRAM assigns
the operator instances to the containers, e.g., per a partition in
the specification, and launches the application. In operations 3, a
STRAM child monitors a throughput statistic (e.g., throughput of
streaming windows or tuples) for a stream between two in-node
operator instances in a single container. In operation 4, the STRAM
receives the results of the monitoring (including the throughput
statistic) from the STRAM child and, in operation 5, the real-time
streaming application obtains the results using an API exposed by
the distributed streaming platform. Based on the throughput
statistic for the stream (e.g., which indicates that the stream is
I/O bound), the real-time streaming application uses the API to
request that the stream mode be made in-line (e.g., that the
stream's locality be made "ThreadLocal"), in operation 6. In an
example embodiment, such a stream mode will result in the two
operator instances connected by the stream being combined into a
single operator instance. Then in operation 7, the STRAM makes the
stream in-line (e.g., through compiler in-lining) and re-initiates
the stream, e.g., using a recovery policy.
[0122] It will be appreciated that in FIG. 19, a throughput
statistic provides the indication that a stream is I/O bound.
However, in the same or an alternative embodiment, the indication
might involve (s) a platform measurement of another statistic or
(b) a platform measurement as to a pre-defined hint, e.g., number
of hash table/map keys.
[0123] FIG. 20 is a diagram showing the combination of two operator
instances connected by a stream in a container, in accordance with
an example embodiment. As depicted in this figure, operator
instance 2002 (labeled Operator A) and operator instance 2003
(labeled Operator B) are connected by a stream in a single
container 2001 on the left side of the arrow before combination.
The stream is associated with a message queue 2004 (a main-memory
buffer backed by persistent storage) managed by a buffer server
(which might be a thread in an example embodiment) that receives
streaming windows of tuples from Operator A and transmits them in
order to Operator B. Following combination, single container 2001,
on the right side of the arrow, contains only one operator instance
2005 (labeled Operator A-Operator B), which does not use the
message queue 2004. In an example embodiment, the combination of
the two operator instances into a single operator instance might be
done by pausing the stream between the two operator instances,
in-lining the program instructions for the two operator instances,
recompiling, and re-initiating the stream using a recovery
policy.
[0124] FIG. 21 is a diagram showing the call stacks for a combined
operator instance resulting from the combination of two operator
instances connected by a stream in a container, in accordance with
an example embodiment. As depicted in this figure, call stack 2101A
is a call stack for a container with a STRAM child, a buffer
server, an operator instance 2102 (labeled Operator A), and an
operator instance 2103 (labeled Operator B) before combination of
the operator instances. In an example embodiment, the container
might be a process, the STRAM child might be a thread, the buffer
server might be a thread, and each operator instance might be a
thread. In an example embodiment, call stack 2101A might include
container code 2104, container data 2105, and container files 2106.
Additionally, call stack 2101A might also include thread registers,
a thread stack, and thread code for each of the four threads: the
STRAM child, the buffer server, operator instance 2102, and
operator instance 2103. Following combination of operator instance
2102 and operator instance 2103, call stack 2101B includes thread
registers, a thread stack, and thread code for a single operator
instance, operator instance 2107 (labeled Operator A-B). It will be
appreciated that the overhead (e.g., call and/or return time for
each execution of a thread) associated with call stack 2101B is
less than the overhead associated with call stack 2101A, since
there is one less operator instance in the former call stack.
[0125] FIG. 22 is a flowchart diagram that illustrates a process
for creating a dynamic partition using a pre-defined hint, in
accordance with an example embodiment. In an example embodiment,
this process might be performed by the Streaming Application Master
(STRAM). In the same or an alternative embodiment, some or all of
the operations of this process might be performed by a STRAM child
(e.g., a slave) or other software in the distributed streaming
platform.
[0126] As depicted in FIG. 22, the software (e.g., the STRAM)
receives a real-time streaming application that runs on a
distributed streaming platform, in operation 2201. In an example
embodiment, the real-time streaming application is structured as a
directed acyclic graph (DAG) with instances of operators as nodes
and streams as edges between nodes. The real-time streaming
application is associated with a pre-defined hint that is a
key-value pair (e.g., the number of entries in a hash table/map).
In operation 2202, the software launches the real-time streaming
application by initiating streams and assigning instances of
operators to containers provided by the distributed streaming
platform. In operation 2203, the software reads the value for a
pre-defined hint (e.g., the number of entries in hash table/map)
and transmits the value to the real-time streaming application
through an application programming interface (API) exposed by the
distributed streaming platform. Then in operation 2204, the
software receives a request from the real-time streaming
application through the API to make a dynamic adjustment (e.g.,
partition an instance of an operator reading from and/or writing to
a hash table/map into multiple instances of the operator). And in
operation 2205, the software makes the dynamic adjustment and
initiates/re-initiates the corresponding streams using a recovery
policy, as described above.
[0127] Here again, each of the operations in this process might be
executed in real time or near real time rather than offline, in an
example embodiment. Moreover, some of the operations described in
the process (e.g., the operations following the original launch of
the real-time streaming application) might be continuous or
non-stop operations, in an example embodiment.
[0128] The process described above uses as an example a key-value
pair that is number of entries in a hash table/map. This example is
not meant to be limiting. For example, in the same or other
embodiments, the key-value pair might be a count associated with
another data structure, e.g., the number of entries in a linked
list.
[0129] In operation 2205, the software makes a dynamic adjustment,
e.g., partitioning an instance of an operator reading from and/or
writing to a hash table/map into multiple instances (e.g., 3
instances) of the operator. In an example embodiment, this dynamic
partitioning might result in the creation of multiple instances
(e.g., 3 instances) of the hash table/map, for example, if the hash
table/map is local variable (e.g., on the thread stack) for the
operator instance. Or the dynamic partitioning might result in no
new additional instances of the hash table/map, for example, if the
hash table/map is a global variable (e.g., container data) for the
operator instance. In an example embodiment, the dynamic adjustment
might not involve partitioning an operator instance. Rather, the
dynamic adjustment might involve setting/unsetting an operator
attribute (e.g., a customization of the operational behavior of an
operator). Or the dynamic adjustment might involve setting or
unsetting an operator property (e.g., a customization of the formal
definition of an operator). Also, in an example embodiment, the
dynamic adjustment might involve setting the locality attribute of
a stream to ThreadLocal to combine two operators into one. Or the
dynamic adjustment might involve flushing a hash table/map to
disk.
[0130] FIG. 23 is a sequence diagram that shows the components
involved in creating a dynamic partition using a pre-defined hint,
in accordance with an example embodiment. It will be appreciated
that this figure involves operations similar to those depicted in
FIG. 22. As shown in FIG. 23, there might be four components
involved in creating a dynamic partition using a pre- defined hint:
(a) a real-time streaming application; (b) a STRAM; (c) a STRAM
child; and (d) the operator instances. In operation 1, which is
earliest in time, the STRAM receives a specification for a
real-time streaming application which defines a hint that is
key-value pair, e.g., the number of keys in a hash table/map used
by an operator. In operation 2, the STRAM assigns operator
instances to containers, e.g., per a static partition in the
specification, and launches the real-time streaming application. In
operation 3, a STRAM child monitors values of the hint, e.g., the
number of keys in hash table/map used by operator instance. In
operation 4, the STRAM receives the results of the monitoring
(including the hint, e.g., the number of keys in hash table/map)
from the STRAM child and, in operation 5, the real-time streaming
application obtains the results using an API exposed by the
distributed streaming platform. Based on the hint (e.g., the number
of keys in hash table/map), the real-time streaming application
uses the API to request that a dynamic partition to increase the
number of operator instances (and the number of hash tables/maps,
in an example embodiment), in operation 6. For example, the
real-time streaming application might include logic (e.g., code)
that requests a dynamic partition to increase the number of
instances of an operator if the number of keys exceeds one million.
Then in operation 7, the STRAM receives the request and assigns
more instances of the operator to the original or a new container
and initiates/re-initiates the corresponding streams as discussed
above.
[0131] Here again, in an example embodiment, the dynamic
partitioning in operation 7 might result in the creation of
multiple instances (e.g., 3 instances) of the hash table/map, for
example, if the hash table/map is local variable (e.g., on the
thread stack) for the operator instance. Or the dynamic
partitioning might result in no new additional instances of the
hash table/map, for example, if the hash table/map is a global
variable (e.g., container data) for the operator instance.
[0132] FIG. 24 is a flowchart diagram that illustrates a process
using a scalable local cache in a container and a pre-defined hint,
in accordance with an example embodiment. In an example embodiment,
this process might be performed by the Streaming Application Master
(STRAM). In the same or an alternative embodiment, some or all of
the operations of this process might be performed by a STRAM child
(e.g., a slave) or other software in the distributed streaming
platform.
[0133] As depicted in FIG. 24, the software (e.g., the STRAM)
receives a real-time streaming application that runs on a
distributed streaming platform, in operation 2401. The real-time
streaming application is structured as a directed acyclic graph
(DAG) with instances of operators as nodes and streams as edges
between nodes. In an example embodiment, multiple (e.g.,
partitioned) instances of the same operator access the same
database. In operation 2402, the software receives a pre-defined
hint associated with the real-time steaming application, where the
pre-defined hint sets a maximum period of time (e.g., approximately
8 hours) for local caching of a result from a query of the database
by each of the multiple instances. Alternatively, the maximum
period of time for local caching might be a default setting for the
real-time streaming application. In operation 2403, the software
launches the real-time streaming application by assigning instances
of operators to containers provided by the distributed streaming
platform and initiating streams. In an example embodiment, each
container is associated with a local cache which is resizable
upward or downward (e.g., from approximately 4 GB). Then in
operation 2404, the software monitors a performance statistic
(e.g., throughput such as number of bytes read per ms, number of
bytes written per ms, number of keys read per ms, number of keys
written per ms, etc.) for accesses to the database by each of the
multiple instances and transmits the results of the monitoring to
the real-time streaming application through an application
programming interface (API) exposed by the distributed streaming
platform. In an example embodiment, the number of keys read/written
per ms might pertain to the keys in a key-value table/map (e.g., a
social security number or SSN used as a key for a person's
address). In operation 2405, the software receives a request from
the real-time streaming application through the API to make a
dynamic adjustment that increases the maximum period of time (e.g.,
to approximately 24 hours or more) for local caching of a result
from a query of the database by each of multiple instances. And in
operation 2406, the software makes the dynamic adjustment.
[0134] Here again, each of the operations in this process might be
executed in real time or near real time rather than offline, in an
example embodiment. Moreover, some of the operations described in
the process (e.g., the operations following the original launch of
the real-time streaming application) might be continuous or
non-stop operations, in an example embodiment.
[0135] In an example embodiment, YARN might be used to create the
local cache in each container, on an as-needed basis. It will be
appreciated that such a local cache (e.g., a compute-local memory
cache) differs from the traditional implementation of Memcached,
which employs a client-server relationship with a relatively small
number of servers providing memory caching services to a relatively
large number of clients.
[0136] FIG. 25 is a sequence diagram that shows the components
involved in a process using a scalable local cache in a container
and a pre-defined hint, in accordance with an example embodiment.
It will be appreciated that this figure involves operations similar
to those depicted in FIG. 24. As shown in FIG. 25, there might be
four components involved in a process using a scalable local cache
in a container and a pre-defined hint: (a) a real-time streaming
application; (b) a STRAM; (c) a STRAM child; and (d) the operator
instances. In operation 1, which is earliest in time, the STRAM
receives a specification which defines a hint, e.g., maximum time
in local cache. In operation 2, the STRAM assigns operator
instances to containers, e.g., per a static partition in the
specification, and launches the real-time streaming application. In
operation 3, a STRAM child monitors a throughput statistic (e.g.,
number of keys read/written per ms) by an operator instance that
reads/writes data to persistent storage (e.g., a database shared
with other partitioned instances of the operator) using a local
cache controlled by the maximum time. In operation 4, the STRAM
receives the results of the monitoring (including a throughput
statistic, e.g., number of keys read/written per ms) from the STRAM
child and, in operation 5, the real-time streaming application
obtains the results using an API exposed by the distributed
streaming platform. Based on the hint (including the throughput
statistic, e.g., number of keys read/written per ms), the real-time
streaming application uses the API to request an increased maximum
time in local cache, in operation 6. For example, the real-time
streaming application might include logic (e.g., code) that
requests an increased time in local cache if the number of keys
read/written per ms (e.g., from the database) is greater than ten
thousand keys per ms. Then in operation 7, the STRAM receives the
request and increases the maximum time in local cache.
[0137] FIG. 26 is a diagram showing a scalable local cache in a
container, in accordance with an example embodiment. As depicted in
this figure, a YARN container 2601 includes two allocations of
volatile memory (e.g., main memory): a local cache 2602 created
with YARN and a message queue 2602 controlled by a buffer server
(e.g., a thread). The container also includes a stream with two
partitioned instances of an operator, Instance A1 and A2, which
receive streaming windows from an input adapter and which transmit
streaming windows to a unifier. In an example embodiment, each of
the operator instances might read/write data to a shared persistent
database (not shown) using local cache 2602. It will be appreciated
that the cache coherency for the local cache is related to the
maximum time in local cache. That is to say, as the maximum time in
local cache increases, so does the probability of a stale value
being in the local cache.
[0138] FIG. 27 is a flowchart diagram that illustrates a process
using partitionable unifiers, in accordance with an example
embodiment. In an example embodiment, this process might be
performed by the Streaming Application Master (STRAM). In the same
or an alternative embodiment, some or all of the operations of this
process might be performed by a STRAM child (e.g., a slave) or
other software in the distributed streaming platform.
[0139] As depicted in FIG. 27, the software (e.g., the STRAM)
receives a real-time streaming application that runs on distributed
streaming platform, in operation 2701. The real-time streaming
application is structured as directed acyclic graph (DAG) with
operators as nodes and streams as edges between nodes. In an
example embodiment, the real-time streaming application includes an
operator that receives values (e.g., values in key-value
table/map), counts values that are unique (or distinct), and emits
the unique values in a stream. In operation 2702, the software
partitions (e.g., statically in a specification for the real-time
streaming application or dynamically based on monitoring) the
operator into multiple partitioned instances. In an example
embodiment, the partitioning results in a round-robin distribution
of values to the multiple partitioned instances. In operation 2703,
the software assigns each unique value emitting from a partitioned
instance to one of a group of unifiers according to a pre-defined
scheme (e.g., alphabetic/ASCII or numeric). In an example
embodiment, each partitionable unifier creates a count of the
unique values received by that partitionable unifier. Then in
operation 2704, the software transmits the counts from each of the
partitionable unifiers to a downstream instance of an operator that
aggregates the counts into a sum of unique values. And in operation
2705, the software causes the sum of unique values to be displayed
in a graphical user interface (GUI) such as a dashboard.
[0140] Here again, each of the operations in this process might be
executed in real time or near real time rather than offline, in an
example embodiment. Moreover, some of the operations described in
the process (e.g., the operations following the original launch of
the real-time streaming application) might be continuous or
non-stop operations, in an example embodiment.
[0141] FIG. 28A is a diagram showing three use cases that do not
use partitionable unifiers, in accordance with example embodiments.
In use case 1 shown in this figure, no static or dynamic partition
is used. So all of the values (e.g., alphabetic identifiers labeled
"A-Z" in the figure, such as a series of surnames: Adams, Smith,
Jones, Smith, Williams, Jones, . . . , Smith) stream from an input
(e.g., an input adapter labeled "Input") to a single operator
instance, labeled Inst. It will be appreciated that an accurate
count of unique (or distinct) values can be obtained by a single
operator instance. However, there is no parallelism (e.g., multiple
instances of the same operator executing in parallel). And
consequently, use case 1 does not scale to handle large series of
values.
[0142] Use case 2 shown in FIG. 28A solves the scalability problem
by using multiple instances (labeled "Inst 1", "Inst 2", and "Inst
3") of the same operator executing in parallel which transmit
counts of unique values to a downstream operator (e.g., a unifier
labeled "Unif") that sums the counts. In an example embodiment, the
values are distributed between the three operator instances using a
round-robin scheme that results in each operator instance receiving
approximately 33.3% of the values. It will be appreciated that an
accurate sum of unique values might not be obtained in this use
case. For example, if the values include three values of the
surname Smith and each of these values is distributed by the
round-robin scheme to a different operator instance, this value
will be reported as unique (or distinct) even though it is not.
[0143] Use case 3 shown in FIG. 28A also solves the scalability
problem by using multiple instances (Inst 1, Inst 2, and Inst 3) of
the same operator executing in parallel. In an example embodiment,
the values are distributed between the three operator instances
using a sticky-key partition based on the first letter (e.g., an
ASCII char value or char) in a surname (e.g., a string/array of
ASCII char values or chars). So surnames whose first letter is
between A-H inclusive go to operator instance Inst 1, surnames
whose first letter is between I-P inclusive go to operator instance
Inst 2, and surnames whose first letter is between R-Z inclusive go
to operator instance Inst 3. It will be appreciated that an
accurate sum of unique values can be obtained in this use case.
However, skewing might occur in this use case, as shown by the bold
arrows going from the input to Inst 1 and from Inst 1 to the
unifier, if the values being input show a skewed distribution
(e.g., most of the input surnames have first letters between A-H
inclusive).
[0144] FIG. 28B is a diagram showing a use case that uses
partitionable unifiers, in accordance with an example embodiment.
It will be appreciated that this figure involves operations similar
to those depicted in FIG. 27. To some extent, the functionality
described in this use case combines the load-balancing of the
functionality described in use case 2 with the sticky-key
functionality described in use case 3. In use case 4 depicted in
this figure, the values are distributed between the three operator
instances (Inst 1, Inst 2, and Inst 3) using a round-robin scheme
that results in each operator instance receiving approximately
33.3% of the values. Each of the operator instances keeps a count
of unique values and emits those unique values to a group of
partitionable unifiers (labeled "P_Unif 1", "P_Unif 2", and "P_Unif
3") using a sticky-key partition based on the first letter (e.g.,
an ASCII char value or char) in a surname (e.g., a string/array of
ASCII char values or chars). So as shown in the topmost box labeled
2801, operator instances Inst 1, Inst 2, and Inst 3 each emit
unique values of surnames beginning with a letter between A-H
inclusive to partitionable unifier P_Unif 1. As shown in the middle
box labeled 2802, operator instances Inst 1, Inst 2, and Inst 3
each emit unique values of surnames beginning with a letter between
I-P inclusive to partitionable unifier P_Unif 2. And as shown in
the bottommost box labeled 2803, operator instances Inst 1, Inst 2,
and Inst 3 each emit unique values of surnames beginning with a
letter between R-Z inclusive to partitionable unifier P_Unif 3.
Each of the partitionable unifiers creates its own count of unique
values and then transmits the count to a unifier (labeled "Unif")
that sums the counts. In an example embodiment, the sum might
replace the counts initially made by operator instance Inst 1,
operator Inst 2, and operator Inst 3.
[0145] In FIG. 28B, the example uses surnames (e.g., a string/array
of ASCII char values or chars) as the values input to the
partitioned operator instances. This example is not meant to be
limiting. In the same or an alternative embodiment, the values
input to the partitioned operator instances might be integers,
floating point numbers, enumerated values, or any other ordinal
values, whether individually or in a collection such as an array or
list. It will be appreciated that partitionable unifiers, as
described in FIGS. 27 and 28B, allow the computation of unique (or
distinct) values in a scalable and distributed fashion.
[0146] In an example embodiment, the operations described in FIGS.
27 and 28B might be implemented as platform functionality. In such
an embodiment, the sticky-key partition amongst the partitionable
unifiers might be performed using a function call, e.g.,
Get_Partionable_Unifier_Key.
[0147] Though some of the embodiments described above have involved
a stock ticker, they are intended as illustrative rather than
limiting. In another example embodiment, some or all of the
operations described above might be used with online machine
learning (including online active learning) where predictions are
compared with subsequent feedback received from a data stream
(e.g., a stock ticker) or a human classifier/labeler. Or some or
all of the operations described above might be used to provide
pricing in real time to a stock exchange, an advertising exchange,
or other online market. Also, some or all of the operations
described above might be used for analyzing websites, targeting
ads, recommending goods or services, providing search results or
other responses to queries, geo-positioning including geo-location,
inventory analysis, online gaming including social gaming, network
routing including routing in wireless networks, etc. Or some or all
of the operations described above might be used for security
including fraud detection, outage detection (e.g., in a data
center), or other analyses of event data (including sensor data),
in real time.
[0148] In an example embodiment, advertising models that use
bidding mechanisms can also benefit from near real-time performance
analysis, which enables buyers and sellers to make faster changes
to ad pricing or inventory adjustments.
[0149] Returning to FIG. 1, personal computer 102 and the servers
in website 104 and website 105 might include (1) hardware
consisting of one or more microprocessors (e.g., from the x86
family, the PowerPC family, the ARM family, etc.), volatile storage
(e.g., RAM), and persistent storage (e.g., a hard disk or
solid-state drive), and (2) an operating system (e.g., Linux,
Windows Server, Mac OS Server, Windows, Mac OS, etc.) that runs on
the hardware. Similarly, in an example embodiment, mobile device
103 might include (1) hardware consisting of one or more
microprocessors (e.g., from the ARM family, the x86 family, etc.),
volatile storage (e.g., RAM), and persistent storage (e.g., flash
memory such as microSD) and (2) an operating system (e.g., Symbian
OS, RIM BlackBerry OS, iPhone OS, Palm webOS, Windows Mobile,
Android, Linux, etc.) that runs on the hardware.
[0150] Also in an example embodiment, personal computer 102 and
mobile device 103 might each include a browser as an application
program or as part of an operating system. Examples of browsers
that might execute on personal computer 102 include Internet
Explorer, Mozilla Firefox, Safari, and Google Chrome. Examples of
browsers that might execute on mobile device 103 include Safari,
Mozilla Firefox, Android Browser, and Palm webOS Browser. It will
be appreciated that users of personal computer 102 and mobile
device 103 might use browsers to communicate (e.g., through a
graphical user interface or GUI) with website software running on
the servers at website 104. Alternatively, a users of personal
computer 102 and mobile device 103 might communicate with website
104 directly or indirectly (e.g., using a script) through a command
line interface (CLI).
[0151] It will be appreciated that the above example embodiments
include functionality that (1) enables a slave (e.g., STRAM Child)
to monitor operator instances in the slave's container and
effectuate dynamic adjustments ordered by the STRAM; (2) generates
streaming windows using control tuples inserted by an input adapter
that creates data tuples from an external data stream through the
application of a schema; (3) displays data from data tuples in a
GUI view using an output adapter that removes control tuples; and
(4) supports checkpointing on streaming window boundaries using
checkpoint tuples inserted by an input adapter.
[0152] With the above embodiments in mind, it should be understood
that the inventions might employ various computer-implemented
operations involving data stored in computer systems. Any of the
operations described herein that form part of the inventions are
useful machine operations. The inventions also relate to a device
or an apparatus for performing these operations. The apparatus may
be a general purpose computer selectively activated or configured
by a computer program stored in the computer. In particular,
various general purpose machines may be used with computer programs
written in accordance with the teachings herein, or it may be more
convenient to construct a more specialized apparatus to perform the
required operations.
[0153] The inventions can also be embodied as computer readable
code on a computer readable medium. The computer readable medium is
any data storage device that can store data, which can thereafter
be read by a computer system. Examples of the computer readable
medium include hard drives, network attached storage (NAS),
read-only memory, random-access memory, CD-ROMs, CD-Rs, CD-RWs,
DVDs, Flash, magnetic tapes, and other optical and non-optical data
storage devices. The computer readable medium can also be
distributed over a network coupled computer systems so that the
computer readable code is stored and executed in a distributed
fashion.
[0154] Although example embodiments of the inventions have been
described in some detail for purposes of clarity of understanding,
it will be apparent that certain changes and modifications can be
practiced within the scope of the following claims. For example,
some or all of the processes described above might be used with
streaming media such as streaming audio or streaming video. Or the
hardware for the distributed streaming platform might include a
quantum computer (e.g., D-Wave System's quantum computer), along
with or instead of traditional servers (e.g., in the x86 or ARM
families). Moreover, the operations described above can be ordered,
modularized, and/or distributed in any suitable way. Accordingly,
the present embodiments are to be considered as illustrative and
not restrictive, and the inventions are not to be limited to the
details given herein, but may be modified within the scope and
equivalents of the following claims. In the following claims,
elements and/or steps do not imply any particular order of
operation, unless explicitly stated in the claims or implicitly
required by the disclosure.
[0155] The example embodiments described in this disclosure include
the following: (1) a method, comprising the operations of:
launching an application that runs on a streaming platform, wherein
the application is structured as a directed acyclic graph (DAG)
with instances of operators as nodes and streams as edges between
nodes; receiving an indication that a first instance of an operator
is I/O bound, wherein a stream connects the first instance of an
operator to a second instance of another operator in a single
container provided by the streaming platform; transmitting the
indication to the application and receiving a request to combine
the first instance with the second instance into a single third
instance of an operator; and creating the third instance and re-
initiating the stream using a recovery policy, wherein each of the
operations is executed by one or more processors in real time or
near real time rather than offline; (2) a method as in (1), wherein
the instances of operators are program instructions and the streams
are unbound sequences of streaming windows that are ordered in
terms of time; (3) a method as in (2), wherein each streaming
window is an atomic microbatch of sequential structured tuples; (4)
a method as in (1), wherein the indication includes a platform
measurement as to resource use; (5) a method as in (4), wherein the
platform measurement is throughput of streaming windows; (6) a
method as in (4), wherein the platform measurement is throughput of
tuples; (7) a method as in (1), wherein the third instance is
created at least in part through compiler in-lining; (8) a method
as in (1), wherein the request sets an attribute of the stream; (9)
one or more computer-readable media persistently storing one or
more programs, wherein the one or more programs, when executed,
instruct one or more processors to perform the following
operations: launch an application that runs on a streaming
platform, wherein the application is structured as a directed
acyclic graph (DAG) with instances of operators as nodes and
streams as edges between nodes; receive an indication that a first
operator instance is I/O bound, wherein a stream connects the first
operator instance to a second operator instance in a single
container provided by the streaming platform; transmit the
indication to the application and receive a request to combine the
first operator with the second operator instance into a single
third operator instance; and create the third operator instance and
re-initiate the stream using a recovery policy, wherein each of the
operations is executed in real time or near real time rather than
offline; (10) computer-readable media as in (9), wherein the
operators are program instructions and the streams are unbound
sequences of streaming windows that are ordered in terms of time;
(11) computer-readable media as in (10), wherein each streaming
window is an atomic microbatch of sequential structured tuples;
(12) computer-readable media as in (9), wherein the indication
includes a platform measurement as to resource use; (13)
computer-readable media as in (12), wherein the platform
measurement is throughput of streaming windows; (14)
computer-readable media as in (12), wherein the platform
measurement is throughput of tuples; (15) computer-readable media
as in (9), wherein the third instance is created at least in part
through compiler in-lining; (16) computer-readable media as in (9),
wherein the request sets an attribute of the stream; (17) a method,
comprising the operations of: launching an application that runs on
a streaming platform, wherein the application is structured as a
directed acyclic graph (DAG) with instances of operators as nodes
and streams as edges between nodes; receiving an indication that a
first operator instance is I/O bound, wherein a stream connects the
first operator instance to a second operator instance in a single
container provided by the streaming platform; transmitting the
indication to the application and receiving a request to combine
the first operator with the second operator instance into a single
third operator instance, wherein the request sets an attribute on a
stream; and creating the third operator instance and re-initiate
the stream, wherein each of the operations is executed by one or
more processors in real time or near real time rather than offline;
(18) a method as in (17), wherein the indication includes a
platform measurement as to resource use; (19) a method as in (18),
wherein the platform measurement is throughput of tuples; (20) a
method as in (17), wherein the third instance is created at least
in part through compiler in-lining.
[0156] The example embodiments described in this disclosure further
include the following: (1) a method, comprising the operations of:
receiving an application that runs on a streaming platform, wherein
the application is structured as a directed acyclic graph (DAG)
with instances of operators as nodes and streams as edges between
nodes and wherein the application is associated with a pre-defined
hint that is a key-value pair; launching the application by
assigning the instances of operators to containers provided by
streaming platform and initiating the streams; reading a value for
the pre-defined hint and transmitting the value to the application
through an application programming interface (API) exposed by the
streaming platform; receiving a request from the application
through the API to make a dynamic adjustment; and making the
dynamic adjustment and re-launching the application using a
recovery policy, wherein each of the operations is executed by one
or more processors in real time or near real time rather than
offline; (2) a method as in (1), wherein the instances of operators
are program instructions and the streams are unbound sequences of
streaming windows that are ordered in terms of time; (3) a method
as in (2), wherein each streaming window is an atomic microbatch of
sequential structured tuples; (4) a method as in (1), wherein the
dynamic adjustment includes a dynamic partition of an instance of
an operator into a plurality of instances; (5) a method as in (4),
wherein the instance of the operator reads from and/or writes to a
hash table or hash map associated with the instance; (6) a method
as in (5), wherein the pre-defined hint is a count of entries in
the hash table or hash map; (7) a method as in (6), wherein the
application makes the request upon determining that the count
exceeds a limit that is pre-defined; (8) a method as in (6),
wherein the application makes the request upon determining that the
count exceeds a limit that is dynamically determined; (9) one or
more computer-readable media persistently storing one or more
programs, wherein the one or more programs, when executed, instruct
one or more processors to perform the following operations: receive
an application that runs on a streaming platform, wherein the
application is structured as a directed acyclic graph (DAG) with
instances of operators as nodes and streams as edges between nodes
and wherein the application is associated with a pre-defined hint
that is a key-value pair; launch the application by assigning the
instances of operators to containers provided by streaming platform
and initiate the streams; read a value for the pre-defined hint and
transmitting the value to the application through an application
programming interface (API) exposed by the streaming platform;
receive a request from the application through the API to make a
dynamic adjustment; and make the dynamic adjustment and
re-launching the application using a recovery policy, wherein each
of the operations is executed in real time or near real time rather
than offline; (10) computer-readable media as in (9), wherein the
instances of operators are program instructions and the streams are
unbound sequences of streaming windows that are ordered in terms of
time; (11) computer-readable media as in (10), wherein each
streaming window is an atomic microbatch of sequential structured
tuples; (12) computer-readable media as in (9), wherein the dynamic
adjustment includes a dynamic partition of an instance of an
operator into a plurality of instances; (13) computer-readable
media as in (12), wherein the instance of the operator reads from
and/or writes to a hash table or hash map associated with the
instance; (14) computer-readable media as in (13), wherein the
pre-defined hint is a count of entries in the hash table or hash
map; (15) computer-readable media as in (14), wherein the
application makes the request upon determining that the count
exceeds a limit that is pre-defined; (16) computer-readable media
as in (14), wherein the application makes the request upon
determining that the count exceeds a limit that is dynamically
determined; (17) a method, comprising the operations of: receiving
an application that runs on a streaming platform, wherein the
application is structured as a directed acyclic graph (DAG) with
instances of operators as nodes and streams as edges between nodes
and wherein the application is associated with a pre-defined hint;
launching the application by assigning the instances of operators
to containers provided by streaming platform and initiating the
streams; reading a value for the pre-defined hint and transmitting
the value to the application through an application programming
interface (API) exposed by the streaming platform; receiving a
request from the application through the API to make a dynamic
adjustment; and making the dynamic adjustment and re-launching the
application, wherein the dynamic adjustment includes a dynamic
partition of an instance of an operator into a plurality of
instances and wherein each of the operations is executed by one or
more processors in real time or near real time rather than offline;
(18) a method as in (17), wherein the instance of the operator
reads from and/or writes to a hash table or hash map associated
with the instance; (19) a method as in 18, wherein the pre-defined
hint is a count of entries in the hash table or hash map; and (20)
a method as in 19, wherein the application makes the request upon
determining that the count exceeds a limit that is pre-defined.
[0157] And the example embodiments described in this disclosure
include the following: (1) a method, comprising the operations of:
receiving an application that runs on a streaming platform, wherein
the application is structured as a directed acyclic graph (DAG)
with operators as nodes and streams as edges between nodes and
wherein the application includes an operator that receives a
plurality of values, counts the values that are unique, and emits
the unique values in a stream; partitioning the operator into at
least two partitioned instances; assigning a unique value emitting
from a partitioned instance to one of a plurality of unifiers
according to a pre-defined scheme, wherein each unifier creates a
count of the unique values received by the unifier; transmitting
the counts from each of the unifiers to a downstream instance of an
operator that aggregates the counts into a sum; and displaying the
sum in a graphical user interface (GUI), wherein each of the
operations is executed by one or more processors in real time or
near real time rather than offline; (2) a method as in (1), wherein
the instances of operators are program instructions and the streams
are unbound sequences of streaming windows that are ordered in
terms of time; (3) a method as in (2), wherein each streaming
window is an atomic microbatch of sequential structured tuples; (4)
a method as in (1), wherein the partitioning is static partitioning
in a specification for the application; (5) a method as in (1),
wherein the partitioning is dynamic partitioning based on results
of monitoring the partitioned instances; (6) a method as in (1),
wherein the pre-defined scheme is ASCII; (7) a method as in (1),
wherein the pre-defined scheme is numerical; (8) a method as in
(1), wherein the partitioning results in a round-robin distribution
of values to the partitioned instances; (9) a method as in (1),
wherein the values are values in a key-value table or a key-value
map; (10) one or more computer-readable media persistently storing
one or more programs, wherein the one or more programs, when
executed, instruct one or more processors to perform the following
operations: receive an application that runs on a streaming
platform, wherein the application is structured as a directed
acyclic graph (DAG) with operators as nodes and streams as edges
between nodes and wherein the application includes an operator that
receives a plurality of values, counts the values that are unique,
and emits the unique values in a stream; partition the operator
into at least two partitioned instances; assign a unique value
emitting from a partitioned instance to one of a plurality of
unifiers according to a pre-defined scheme, wherein each unifier
creates a count of the unique values received by the unifier;
transmit the counts from each of the unifiers to a downstream
instance of an operator that aggregates the counts into a sum; and
display the sum in a graphical user interface (GUI), wherein each
of the operations is executed in real time or near real time rather
than offline; (11) computer-readable media as in (10), wherein the
instances of operators are program instructions and the streams are
unbound sequences of streaming windows that are ordered in terms of
time; (12) computer-readable media as in (11), wherein each
streaming window is an atomic microbatch of sequential structured
tuples; (13) computer-readable media as in (10), wherein the
partitioning is static partitioning in a specification for the
application; (14) computer-readable media as in (10), wherein the
partitioning is dynamic partitioning based on results of monitoring
the partitioned instances; (15) computer-readable media as in (10),
wherein the pre-defined scheme is ASCII; (16) computer-readable
media as in (10), wherein the pre-defined scheme is numerical; (17)
computer-readable media as in (10), wherein the partitioning
results in a round-robin distribution of values to the partitioned
instances; (18) computer-readable media as in (10), wherein the
values are values in a key-value table or a key-value map; (19) a
method, comprising the operations of: receiving an application that
runs on a streaming platform, wherein the application is structured
as a directed acyclic graph (DAG) with operators as nodes and
streams as edges between nodes and wherein the application includes
an operator that receives a plurality of values, counts the values
that are unique, and emits the unique values in a stream;
partitioning the operator into at least two partitioned instances,
wherein the partitioning is static partitioning in a specification
for the application and the partitioning results in a round-robin
distribution of values to the partitioned instances; assigning a
unique value emitting from a partitioned instance to one of a
plurality of unifiers according to a pre-defined scheme, wherein
each unifier creates a count of the unique values received by the
unifier; transmitting the counts from each of the unifiers to a
downstream instance of an operator that aggregates the counts into
a sum; and displaying the sum in a graphical user interface (GUI),
wherein each of the operations is executed by one or more
processors in real time or near real time rather than offline; (20)
a method as in (19), wherein the pre-defined scheme is ASCII.
* * * * *
References