U.S. patent application number 13/908188 was filed with the patent office on 2013-10-10 for methods and systems for real-time continuous updates.
The applicant listed for this patent is Teradata Corporation. Invention is credited to Alan Lee Beck, Jeremy Branscome, Joseph Chamdani, Ravindran Krishnamurthy, Krishnan Meiyyappan, Kapil Surlaker, Hung Tran.
Application Number | 20130268489 13/908188 |
Document ID | / |
Family ID | 41432276 |
Filed Date | 2013-10-10 |
United States Patent
Application |
20130268489 |
Kind Code |
A1 |
Surlaker; Kapil ; et
al. |
October 10, 2013 |
METHODS AND SYSTEMS FOR REAL-TIME CONTINUOUS UPDATES
Abstract
Embodiments of the present invention provide fine grain
concurrency control for transactions in the presence of database
updates. During operations, each transaction is assigned a snapshot
version number or SVN. A SVN refers to a historical snapshot of the
database that can be created periodically or on demand.
Transactions are thus tied to a particular SVN, such as, when the
transaction was created. Queries belonging to the transactions can
access data that is consistent as of a point in time, for example,
corresponding to the latest SVN when the transaction was created.
At various times, data from the database stored in a memory can be
updated using the snapshot data corresponding to a SVN. When a
transaction is committed, a snapshot of the database with a new SVN
is created based on the data modified by the transaction and the
snapshot is synchronized to the memory. When a transaction query
requires data from a version of the database corresponding to a
SVN, the data in the memory may be synchronized with the snapshot
data corresponding to that SVN.
Inventors: |
Surlaker; Kapil; (Sunnyvale,
CA) ; Krishnamurthy; Ravindran; (Redmond, WA)
; Meiyyappan; Krishnan; (Fremont, CA) ; Beck; Alan
Lee; (Santa Clara, CA) ; Tran; Hung;
(Sunnyvale, CA) ; Branscome; Jeremy; (Santa Clara,
CA) ; Chamdani; Joseph; (Santa Clara, CA) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
Teradata Corporation |
Dayton |
OH |
US |
|
|
Family ID: |
41432276 |
Appl. No.: |
13/908188 |
Filed: |
June 3, 2013 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
12144486 |
Jun 23, 2008 |
8458129 |
|
|
13908188 |
|
|
|
|
Current U.S.
Class: |
707/625 |
Current CPC
Class: |
G06F 16/2329 20190101;
G06F 16/2477 20190101 |
Class at
Publication: |
707/625 |
International
Class: |
G06F 17/30 20060101
G06F017/30 |
Claims
1. A method of versioning data in a database maintained within a
host computer system coupled with a hardware accelerator, the host
computer system including a host memory, and the hardware
accelerator including a hardware accelerator memory, said method
comprising: determining snapshots of data in the database at an
interval; identifying differences in the data; determining a set of
changes that takes the data from one snapshot to another snapshot;
and generating a mask comprising the set of changes that is applied
by a machine code database instruction for modifying the data in
pages of the hardware accelerator memory from one snapshot to
another snapshot.
2. The method of claim 1, wherein generating the mask comprises
generating a mask having a fixed width format that is independent
of the data type being modified.
3. The method of claim 1, wherein generating the mask comprises
generating a mask having data that is in a same compressed form as
the data in the database.
4. The method of claim 1, further comprising storing a mask for an
active snapshot in the host memory and storing at least one
additional mask for completed snapshots in the host memory.
5. The method of claim 1, further comprising storing a mask for an
active snapshot in the host memory and storing at least one
additional mask for completed snapshots in the hardware accelerator
memory.
6. The method of claim 1, further comprising setting the interval
based on a configuration.
7. The method of claim 1, further comprising setting the interval
based on a size of accumulated changes.
8. The method of claim 1, further comprising applying the mask to
data in the hardware accelerator memory upon a transaction
commit.
9. The method of claim 1, further comprising applying the mask to
data in the hardware accelerator memory upon demand by a query.
10. A tangible, non-transitory, computer readable medium comprising
program code for performing a method of versioning data in a
database maintained within a host computer system coupled with a
hardware accelerator, the host computer system including a host
memory, and the hardware accelerator including a hardware
accelerator memory, said medium comprising: program code for
determining snapshots of data in the database at an interval;
program code for identifying differences in the data; program code
for determining a set of changes that takes the data from one
snapshot to another snapshot; and program code for generating a
mask comprising the set of changes that is applied by a machine
code database instruction for modifying the data in pages of the
hardware accelerator memory from one snapshot to another
snapshot.
11. The computer readable medium of claim 10, wherein the program
code for generating the mask comprises program code for generating
a mask having a fixed width format that is independent of the data
type being modified.
12. The computer readable medium of claim 10, wherein the program
code for generating the mask comprises program code for generating
a mask having data that is in a same compressed form as the data in
the database.
13. The computer readable medium of claim 10, further comprising
program code for storing a mask for an active snapshot in the host
memory and program code for storing at least one additional mask
for completed snapshots in the host memory.
14. The computer readable medium of claim 10, further comprising
program code for storing a mask for an active snapshot in the host
memory and program code for storing at least one additional mask
for completed snapshots in the hardware accelerator memory.
15. The computer readable medium of claim 10, further comprising
program code for setting the interval based on a configuration.
16. The computer readable medium of claim 10, further comprising
program code for setting the interval based on a size of
accumulated changes.
17. The computer readable medium of claim 10, further comprising
program code for applying the mask to data in the hardware
accelerator memory upon a transaction commit.
18. The computer readable medium of claim 10, further comprising
program code for applying the mask to data in the hardware
accelerator upon demand by a query.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] This application is a divisional application of U.S. patent
application Ser. No. 12/144,486; entitled "METHODS AND SYSTEMS FOR
REAL-TIME CONTINUOUS UPDATES" by Kapil Surlaker, Ravi
Krishnamurthy, Krishnan Meiyyappan, Alan Beck, Hung Tran, Jeremy
Branscome, and Joseph Chamdani; filed on Jun. 23, 2008.
[0002] This application is related to the following U.S. Patent
Applications and Patents, which are herein incorporated by
reference in their entirety: U.S. patent application Ser. No.
11/895,952, filed on Aug. 27, 2007, entitled Methods and Systems
for Hardware Acceleration of Database Operations and Queries, by
Joseph I. Chamdani et al.; U.S. patent application Ser. No.
11/895,998, filed on Aug. 27, 2007, entitled Hardware Acceleration
Reconfigurable Processor for Accelerating Database Operations and
Queries, by Jeremy Branscome et al.; U.S. patent application Ser.
No. 11/895,997, filed on Aug. 27, 2007, entitled Processing
Elements of a Hardware Acceleration Reconfigurable Processor for
Accelerating Database Operations and Queries, by Jeremy Branscome
et al.; U.S. patent application Ser. No. 12/168,821, filed on Jul.
7, 2008, entitled Methods and Systems for Generating Query Plans
that are Compatible for Execution in Hardware, by Ravi
Krishnamurthy et al.; U.S. patent application Ser. No. 12/099,076,
filed on Apr. 7, 2008, entitled Methods and Systems for Run-Time
Scheduling Database Operations that are Executed in Hardware, by
Joseph I. Chamdani et al.; U.S. patent application Ser. No.
12/099,131, filed on Apr. 7, 2008, entitled Accessing Data in a
Column Store Database Based on Hardware Compatible Data Structures,
by Liuxi Yang et al.; U.S. patent application Ser. No. 12/099,133,
filed on Apr. 7, 2008, entitled Accessing Data in a Column Store
Database Based on Hardware Compatible Indexing and Replicated
Reordered Columns, by Krishnan Meiyyappan et al.; and U.S. patent
application Ser. No. 12/144,303, filed on Jun. 23, 2008, entitled
Fast Bulk Loading and Incremental Loading of Data into a Database,
by James Shau et al.
BACKGROUND
[0003] Despite their different uses, applications, and workload
characteristics, most systems run on a common Database Management
System (DBMS) using a standard database programming language, such
as Structured Query Language (SQL). Most modern DBMS
implementations (Oracle, IBM DB2, Microsoft SQL, Sybase, MySQL,
PostgreSQL, Ingress, etc.) are implemented on relational databases,
which are well known to those skilled in the art.
[0004] Typically, a DBMS has a client side where applications or
users submit their queries and a server side that executes the
queries. On the server side, most enterprises employ one or more
general purpose servers. However, although these platforms are
flexible, general purpose servers are not optimized for many
enterprise database applications. In a general purpose database
server, all SQL queries and transactions are eventually mapped to
low level software instructions called assembly instructions, which
are then executed on a general purpose microprocessor (CPU). The
CPU executes the instructions, and its logic is busy as long as the
operand data are available, either in the register file or on-chip
cache. To extract more parallelism from the assembly code and keep
the CPU pipeline busy, known CPUs attempt to predict ahead the
outcome of branch instructions and execute down the SQL code path
speculatively. Execution time is reduced if the speculation is
correct; the success of this speculation, however, is data
dependent. Other state-of-the-art CPUs attempt to increase
performance by employing simultaneous multithreading (SMT) and/or
multi-core chip multiprocessing (CMP). To take advantage of these,
changes have to be made at the application or DBMS source code to
manually create the process/thread parallelism for the SMT or CMP
CPUs. This is generally considered highly undesirable.
[0005] Unfortunately, general purpose CPUs are not efficient for
database applications. Branch prediction is generally not accurate
because database processing involves tree traversing and link list
or pointer chasing that is very data dependent. Known CPUs employ
the well-known code-flow (or Von Neumann) architecture, which uses
a highly pipelined instruction flow (rather than a dataflow where
operand data is pipelined) to operate on data stored in the CPUs
tiny register files. Real database workloads, however, typically
require processing Gigabytes to Terabytes of data, which overwhelms
these tiny registers with loads and reloads. On-chip cache of a
general purpose CPU is not effective since it's relatively too
small for real database workloads. This requires that the database
server frequently retrieve data from its small memory or disk.
Accordingly, known database servers rely heavily on squeezing the
utilization of their small system memory size and disk input/output
(I/O) bandwidth. Those skilled in the art recognize that these
bottlenecks between storage I/O, the CPU, and memory are very
significant performance factors.
[0006] However, overcoming these bottlenecks is a complex task
because typical database systems consist of several layers of
hardware, software, etc., that influence the overall performance of
the system. These layers comprise, for example, the application
software, the DBMS software, operating system (OS), server
processor systems, such as its CPU, memory, and disk I/O and
infrastructure. Traditionally, performance has been optimized in a
database system horizontally, i.e., within a particular layer. For
example, many solutions attempt to optimize various solutions for
the DBMS query processing, caching, the disk I/O, etc. These
solutions employ a generic, narrow approach that still fails to
truly optimize the large performance potentials of the database
system, especially for relational database systems having complex
read-intensive applications. In addition, in order to support
updates, the conventional solutions typically employ complex
mechanisms to provide versioning control and concurrent access.
BRIEF DESCRIPTION OF THE DRAWINGS
[0007] The accompanying drawings, which are incorporated in and
constitute a part of this specification, illustrate embodiments of
the invention and together with the description, serve to explain
the principles of the invention. In the figures:
[0008] FIG. 1 illustrates an exemplary system that is consistent
with the principles of the present invention;
[0009] FIG. 2 illustrates exemplary system topologies that are
consistent with the principles of the present invention;
[0010] FIG. 3A illustrates a database system and FIG. 3B
illustrates some of the optimizations of the present
disclosure;
[0011] FIG. 4 illustrates a functional architecture of the custom
computing (C2) software of the present invention;
[0012] FIG. 5 illustrates a protocol stack employed by the C2
software and a Hardware Acceleration Reconfigurable Processor
(HARP) of the present invention;
[0013] FIG. 6 illustrates an exemplary architecture of a HARP;
[0014] FIG. 7 illustrates a column store database and associated
data structures employed by some embodiments of the present
invention;
[0015] FIG. 8 illustrates a table column layout and associated data
structures employed by some embodiments of the present
invention;
[0016] FIG. 9 illustrates an exemplary machine code database
instruction flow for a SQL query that is consistent with the
principles of the present invention;
[0017] FIG. 10 illustrates an exemplary dataflow for a SQL query
through processing elements in the HARP in accordance with the
principles of the present invention;
[0018] FIG. 11 illustrates an exemplary logic flow of the
addressing schemes employed in the present invention;
[0019] FIG. 12 illustrates a structure of the global database
virtual address scheme of the present invention;
[0020] FIG. 13 illustrates an exemplary timeline of transactions
and snapshots to support multi-versioning and continuous
updates;
[0021] FIG. 14 illustrates exemplary synching of pages between
different snapshots;
[0022] FIG. 15 illustrates an example of synching a page from one
snapshot to a subsequent snapshot;
[0023] FIG. 16 illustrates an example of synching a page from one
snapshot to a previous snapshot; and
[0024] FIGS. 17 and 18 illustrate associative accumulation machine
code database instructions that may be employed to change page
contents from one snapshot version to another.
DETAILED DESCRIPTION
[0025] Embodiments of the present invention provide fine grain
concurrency control for transactions in the presence of database
updates. During operations, each transaction is assigned a snapshot
version number or SVN. A SVN refers to a historical snapshot of the
database that can be created periodically or on demand.
Transactions are thus tied to a particular SVN, such as, when the
transaction was created. Queries belonging to the transactions can
access data that is consistent as of a point in time, for example,
corresponding to the latest SVN when the transaction was
created.
[0026] At various times, data from the database stored in a memory
can be updated using the snapshot data corresponding to a SVN. For
example, when a transaction is committed, a snapshot of the
database with a new SVN is created based on the data modified by
the transaction and the snapshot is synchronized to the memory. As
another example, when a transaction query requires data from a
version of the database corresponding to a SVN, the data in the
memory may be synchronized with the snapshot data corresponding to
that SVN. This feature, in essence, creates multiple versions of
the database and allows read-intensive database applications to
operate in the presence of database updates.
[0027] Due to the comprehensive nature of the present inventions in
the C2 solution, the figures are presented generally from a high
level of detail and progress to a low level of detail. For example,
FIGS. 1-3 illustrate exemplary systems and topologies enabled by
the present invention. FIG. 4-5 illustrate the architecture of the
C2 software. FIG. 6 illustrates the architecture of a HARP module.
FIGS. 7-8 illustrate the database format and data structures
employed by the C2 solution of the present invention. FIGS. 9-10
illustrates an example execution of a SQL query by the C2 solution
of the present invention. Reference will now be made in detail to
the exemplary embodiments of the invention, which are illustrated
in the accompanying drawings. FIG. 11 illustrates an exemplary
logic flow of the addressing schemes employed. FIG. 12 illustrates
a structure of the global database virtual address scheme. FIG. 13
illustrates an exemplary timeline of transactions and snapshots to
support multi-versioning and continuous updates. FIG. 14
illustrates exemplary synching of pages between different
snapshots. FIG. 15 illustrates an example of synching a page from
one snapshot to a subsequent snapshot. FIG. 16 illustrates an
example of synching a page from one snapshot to a previous
snapshot. FIGS. 17 and 18 illustrate associative accumulation
machine code database instructions that may be employed to change
page contents from one snapshot version to another. Wherever
possible, the same reference numbers will be used throughout the
drawings to refer to the same or like parts.
[0028] FIG. 1--An Exemplary C2 System
[0029] The present invention employs a custom computing (C2)
solution that provides a significant gain in performance for
enterprise database applications. In the C2 solution, a node or
appliance may comprise the host (or base) system that is combined
with hardware acceleration reconfigurable processors (HARP). These
HARPs are specially designed to optimize the performance of
database systems and its applications, especially relational
database systems and read-intensive applications.
[0030] A host system may be any standard or pre-existing DBMS
system. In general, such systems will comprise a standard general
purpose CPU, a system memory, I/O interfaces, etc.
[0031] The HARPs are coupled to the host system and are designed to
offload repetitive database operations from the DBMS running on the
host system. The HARPs utilize dataflow architecture processing
elements that execute machine code instructions that are defined
for various database operations. The C2 solution may employ a node
that is scalable to include one HARP, or multiple HARPs. In
addition, the C2 solution may use a federated architecture
comprising multiple nodes, i.e., multiple DBMS servers that are
enhanced with the C2 solution.
[0032] In some embodiments, the C2 solution employs an open
architecture and co-processor approach so that the C2 hardware can
be easily integrated into existing database systems. Of note, the
hardware acceleration of the C2 solution utilizes novel machine
code database instructions to execute certain fragments of a query
in a dataflow and using parallel, pipelined execution.
[0033] In the present invention, the C2 solution also comprises
software that orchestrates the operations of the DBMS running on
the host system and the HARPs. The C2 software is configured with a
flexible, layered architecture to make it hardware and database
system agnostic. Thus, the C2 software is capable of seamlessly
working with existing DBMSs based on this open architecture.
[0034] In general, the C2 software receives the query from the DBMS
and breaks the query down into query fragments. The C2 software
then decides which of these query fragments can be appropriately
handled in software (in the C2 software itself or back in the
originating DBMS) or, ideally, with hardware acceleration in the
HARPs. All or part of the query may be processed by the C2 software
and HARPs.
[0035] In addition, in order to maximize the efficiency of the
hardware acceleration, the C2 solution stores its databases in
compressed, column-store format and utilizes various
hardware-friendly data structures. The C2 solution may employ
various compression techniques to minimize or reduce the storage
footprint of its databases. The column-store format and
hardware-friendly data structures allow the HARPs or C2 software to
operate directly on the compressed data in the column-store
database. The column-store database may employ columns and column
groups that are arranged based on an implicit row identifier (RID)
scheme and RID to primary column to allow for easy processing by
the HARPs. The hardware-friendly data structures also allow for
efficient indexing, data manipulation, etc. by the HARPs.
[0036] For example, the C2 solution utilizes a global virtual
address space for the entire database to greatly simplify and
maximize efficiency of create, read, update, and delete operations
of data in a database. In some embodiments, the columns and column
groups are configured with a fixed width to allow for arithmetic
memory addressing and translation from a virtual address to a
physical memory address. On-demand and speculative prefetching may
also be utilized by the C2 solution to hide I/O bandwidth latency
and maximize HARP utilization.
[0037] Referring now to FIG. 1, an exemplary system 1f the C2
solution is illustrated. As shown, system 100 may comprise an
application 102 that is running on a client 104, such as a personal
computer or other system. Application 102 interfaces a DBMS 106
across a network 108, such as the Internet, local area network,
etc. DBMS 106 may further interface one or more databases stored in
storage infrastructure 112. For purposes of explanation, DBMS 106
and its components may be collectively referred to in this
disclosure as a node of system 100. Although FIG. 1 shows a single
node, system 100 may of course comprise multiple nodes. The various
components of FIG. 1 will now be further described.
[0038] Application 102 may be any computer software that requests
the services of DBMS 106. Such applications are well known to those
skilled in the art. For example, application 102 may be a web
browser in which a user is submitting various search requests. Of
course, application 102 may be another system or software that is
consuming the services of DBMS 106 and submitting queries to DBMS
106.
[0039] Client 104 represents the hardware and software that
supports the execution of application 102. Such clients are well
known to those skilled in the art. For example, client 104 may be a
personal computer or another server.
[0040] DBMS 106 is any computer software that manages databases. In
general, DBMS 106 controls the organization, storage, management,
retrieval of data in a database. As is well known, these types of
systems are common for supporting various SQL queries on relational
databases (and thus may also be known as a RDBMS). Due to its open
architecture, various DBMS systems may be employed by the present
invention. Typical examples of DBMSs include Oracle, DB2, Microsoft
Access, Microsoft SQL Server, PostgreSQL, and MySQL.
[0041] In some embodiments, and for purposes of explanation, DBMS
106 is shown comprising C2 software 110 interfacing MySQL software
114 via an API 116. MySQL software 114 is open source software that
is sponsored and provided by MySQL AB and is well known to those
skilled in the art. Of course, any DBMS software, such as those
noted above, may be employed in the present invention.
[0042] C2 software 110 orchestrates the execution of a query
forwarded from DBMS 106, and thus, operates in conjunction with
MySQL software 114. For example, in the C2 software 110, SQL
queries are broken down into query fragments and then routed to the
most appropriate resource. A query fragment may be handled in C2
hardware, i.e., HARP module 204. (HARP module 204 is further
described with reference to FIG. 2.) The query fragment may also be
processed in the C2 software itself, or returned for handling by
MySQL software 114.
[0043] In general, C2 software 110 utilizes a flexible, layered
architecture to make it hardware and database system agnostic. For
example, C2 software 110 may operate as a storage engine of MySQL
software 114. As is well known, MySQL software 114 may provide an
API 116 for storage engines, which C2 software 110 may plug in to.
API 116 comprises the software that specifies how the C2 software
110 and MySQL software 114 will interact, how they will request
services from each other, such as SQL queries and results.
[0044] As a storage engine, C2 software 110 may employ the MySQL
API 116 to provide various storage mechanisms, indexing facilities,
locking levels, and ultimately provide a range of different
functions and capabilities that are transparent to MySQL software
114. As noted above, this is one aspect of how the present
invention overcomes the generic approach in known solutions without
having to sacrifice performance for functionality, or fine tune the
database. Of note, although FIG. 1 shows a single storage engine,
MySQL software 114 may be coupled to multiple storage engines (not
shown) in addition to C2 software 110. C2 software 110 is also
described in further detail with reference to FIGS. 4-5.
[0045] Network 108 represents the communication infrastructure that
couples application 102 and DBMS 106. For example, network 108 may
be the Internet. Of course, any network, such as a local area
network, wide area network, etc., may be employed by the present
invention.
[0046] Storage infrastructure 112 comprises the computer storage
devices, such as disk arrays, tape libraries, and optical drives
that serve as the storage for the databases of system 100. Storage
infrastructure 112 may employ various architectures, such as a
storage area network, network attached storage, etc., which are
known to those skilled in the art.
[0047] In some embodiments, the C2 solution stores its databases in
storage infrastructure 112 in column-store format. Column-store
format is where data is stored in columns or groups of columns.
Column-store format is advantageous for data fetching, scanning,
searching, and data compression. The column-store format may employ
fixed width columns and column groups with implicit RIDs and a RID
to primary key column to allow for arithmetic memory addressing and
translation. This allows HARPs 204 to utilize hardware processing
for database processing, such as column hopping, and to operate
directly on the compressed data in the columns.
[0048] In contrast, in typical DBMS environments, data is stored in
row-store format. Row-store format is sometimes considered by those
skilled in the art for having better performance in data updates
and record retrieval; thus, it is sometimes considered to have
better functionality over column-store databases in most
applications with a high ratio of updates over reads. In the
present invention, however, the C2 solution achieves better
performance by using hardware acceleration with a column-store
database, yet it still delivers the functionality and benefits of
row-store databases. The column store format used by the C2
solution of the present invention is further described with
reference to FIGS. 7-8.
[0049] FIG. 2--System Topologies
[0050] FIG. 2 illustrates exemplary system topologies that are
consistent with the principles of the present invention. As shown,
FIG. 2 illustrates a basic C2 node topology, a scale up C2 node
topology, and a scale out topology. These various topologies may be
utilized to customize the C2 solution for various sizes of
databases and desired performance. In addition, these topologies
are provided to illustrate that the C2 solution can be easily
scaled up to virtually any size of database or performance.
[0051] First, the basic C2 node will be explained, which comprises
a single host system 202 and a single HARP module 204. Variations
of this basic node will then be explained to show how the basic
node can be scaled up and how multiple nodes can be employed in a
federated architecture.
[0052] The basic C2 node topology may comprise a host system 202
and a hardware acceleration reconfigurable processor (HARP) module
204. Collectively, host 202 and HARP module 204 may be referred to
as a node or appliance. In some embodiments, host system 202 and
HARP module 204 are coupled together over a known communications
interface, such as a PCIe or hypertransport (HT) interface. In
terms of packaging, host system 202 and HARP module 204 may be
built on one or more cards or blades that are bundled together in a
common chassis or merely wired together. In the C2 solution, host
system 202 and HARP module 204 may be flexibly packaged using a
modular form factor for ease of installation and scaling.
[0053] The host system 202 may comprise a general purpose CPU, such
as a Xeon x86 processor by the Intel Corporation, and a memory,
such as a dynamic random access memory. Such types of host systems
are well known to those skilled in the art. In general, in the C2
solution, host system 202 will be used to process parts of a query
that are less time consuming (i.e., slow path portion), such as
server-client connection, authentication, SQL parsing, logging,
etc. However, in order to optimize performance, the bulk of query
execution (i.e., the fast path portion) is offloaded to the HARP
module 204.
[0054] Host system 202 may run MySQL software 114 and also run C2
software 110 that orchestrates query processing between MySQL 114
and HARP 204. In particular, C2 software 110 will decompose a query
into a set of query fragments. Each fragment comprises various
tasks, which may have certain dependencies. C2 software 110 will
determine which fragments and tasks are part of the fast path
portion and offload them to the HARP module 204. Appropriate tasks
for the selected query fragments are sent to HARP module 204 with
information on the database operation dependency graph. Within the
HARP module 204, tasks are further broken down into
parallel/pipelined machine code operations (known as MOPs) and
executed in hardware.
[0055] HARP module 204 comprises processing logic (HARP logic 302)
and a relatively large memory (HARP memory 304) for hardware
accelerating database operations of the node. In some embodiments,
HARP module 204 is configured to handle various repetitive database
tasks, such as table scanning, indexing, etc. In the C2 solution,
HARP module 204 can receive high-level database query tasks (not
just low-level read/write or primitive computation tasks as is
typically for a general purpose processor) in the form of machine
code database instructions.
[0056] HARP logic 302 is the hardware that executes machine code
database instructions for the database tasks being handled by HARP
module 204. To adapt to application requirement changes, the HARP
logic 302 is designed to have hardware re-configurability.
Accordingly, in some embodiments, HARP logic 302 is implemented
using field programmable gate arrays (FPGAs). However, any type of
custom integrated circuit, such as application specific integrated
circuits (ASICs), may be implemented as HARP logic 302.
[0057] HARP memory 304 serves as the memory of HARP module 204. In
order to maximize the efficiency of the HARP logic 302, the HARP
memory 304 may be implemented using relatively large amounts of
memory. For example, in some embodiments, the HARP memory 304 in a
HARP module 204 may comprise 256 Giga-Bytes or more of RAM or DRAM.
Of course, even larger amounts of memory may be installed in HARP
module 204. HARP logic 302 and HARP memory 304 are further
described with reference to FIG. 6.
[0058] In addition to the basic C2 node, a scale up C2 node
topology may be used as an extension of the basic C2 node. As
shown, host system 202 may now be coupled to a plurality or array
of 1-N HARP modules 204. In this type of node, a PCIe switch or
other suitable switching fabric may couple these components
together with storage infrastructure 112. Of course, other internal
arrangements for a scale up C2 node may be utilized in the present
invention.
[0059] Going further, a scale out topology can be used for multiple
C2 nodes. As shown, the scale out topology may comprise various
combinations of either the basic or scale up C2 nodes. For example,
as shown, the scale out topology may comprise Nodes 1-M, which are
coupled to storage infrastructure 112. In FIG. 2, Node 1 is shown
as a basic C2 node, while Node M is shown as a scale up node. A
control node 206 is also shown and manages the operations of Nodes
1-M. Control node 206 is shown as a separate node; however, those
skilled in the art will recognize the role of control node 206 by
any of Nodes 1-M. Other variations in node hierarchy and management
are within the scope of the present invention. Of course, this
topology may also comprise a variety of combinations of nodes.
[0060] FIGS. 3A and 3B--Some Advantages of the Present
Invention
[0061] FIG. 3A illustrates a prior art database system and FIG. 3B
illustrates an exemplary implementation of the C2 solution for the
present invention. In FIG. 3A, a typical prior art database system
is shown. An SQL query is submitted to a DBMS (e.g., MySQL), which
runs on top of a typical operating system. The CPU attempts to then
execute the SQL query. However, because the CPU is a general
purpose CPU it executes this query based on software, which has
several limitations.
[0062] In contrast, as shown in FIG. 3B, the SQL query may
submitted to a C2 system having a DBMS that comprises a top layer
DBMS software (i.e., MySQL) 114 and C2 software 110. C2 software
110 interfaces with the DBMS software 114 to orchestrate and
optimize processing of the SQL query.
[0063] In particular, C2 software 110 may identify portions of the
query, i.e., the fast path portion, which is better handled in
hardware, such as HARP module 204. Such portions may be those
fragments of the query that are repetitive in nature, such as
scanning, indexing, etc. In the prior art system, the DBMS is
limited by its own programming, the operating system, and the
general purpose CPU. The present invention avoids these bottlenecks
by offloading fast path portions of a query to HARP module 204.
[0064] As shown, HARP module 204 comprises HARP logic 302 and a
HARP memory 304 to accelerate the processing of SQL queries. In
order maximize the use of HARP module 204, the present invention
may also utilize column store databases. Whereas the prior art
system is hindered by the limitations of a standard row store
database. These features also allow the present invention to
maximize the performance of the I/O between the operating system
and storage.
[0065] For ease of implementation, C2 software 110 may be
implemented on well-known operating systems. The operating system
will continue to be used to perform basic tasks such as controlling
and allocating memory, prioritizing system requests, controlling
input and output devices, facilitating networking, and managing
files and data in storage infrastructure 112. In some embodiments,
various operating systems, such as Linux, UNIX, and Microsoft
Windows, may be implemented.
[0066] FIGS. 3A and 3B are provided to illustrate some of the
differences between the present invention and the prior art and
advantages of the present invention. Those skilled in the art will
also recognize that other advantages and benefits may be achieved
by the embodiments of the present invention. For purposes of
explanation, the present disclosure will now describe the C2
software, hardware, data structures, and some operations in further
detail.
[0067] FIG. 4--C2 Software Architecture
[0068] As noted, C2 software 110 orchestrates the processing a
query between MySQL software 114 and HARP module 204. In some
embodiments, C2 software 110 runs as an application on host system
202 and as a storage engine of MySQL software 114. FIG. 4
illustrates an architecture of the C2 software 110. As shown, C2
software 110 comprises a query and plan manager 402, a query
reduction/rewrite module 404, an optimizer 406, a post optimizer
module 408, a query plan generator 410, an execution engine 412, a
buffer manager 414, a task manager 416, a memory manager 418, a
storage manager 420, an answer manager 422, an update manager 424,
shared utilities 426, and a HARP manager 428. Each of these
components will now be briefly described.
[0069] Query and plan manager 402 analyzes and represents the query
received from the MySQL software 114, annotates the query, and
provides a representation of the query plan. Query
reduction/rewrite module 404 breaks the query into query fragments
and rewrites the query fragments into tasks. Rewrites may be needed
for compressed domain rewrites and machine code database
instruction operator rewrites. Optimizer 406 performs cost-based
optimization to be done using cost model of resources available to
C2 software 110, i.e., HARP module 204, resources of C2 software
110 itself using software operations, or MySQL software 114.
[0070] These modules interact with each other to determine how to
execute a query, such as a SQL query from MySQL software 114. The
data structures output by the query plan generator 410 will be the
same data structure that the optimizer 406 and the rewrite module
404 will operate on. Once a parsed SQL query has been represented
in this data structure (converted, for example, from MySQL),
manager 402 rewrites the query such that each fragment of the query
can be done entirely in MySQL software 114, in C2 software 110, or
in HARP module 204. Once the final query representation is
available, the rewrite module 404 goes through and breaks the graph
into query fragments.
[0071] Post optimizer module 408 is an optional component that
rewrites after the optimizer 406 for coalescing improvements found
by optimizer 406. Query plan generator 410 generates an
annotations-based, template-driven plan generation for the query
tasks. Execution engine 412 executes the query fragments that are
to be handled by software or supervises the query execution in HARP
module 204 via HARP manager 428.
[0072] Buffer manager 414 manages the buffers of data held in the
memory of host 202 and for the software execution tasks handled by
host 202. Task manager 416 orchestrates the execution of all the
tasks in HARP module 204 and software, i.e., in execution engine
412 or MySQL software 114.
[0073] Memory manager 416 manages the virtual address and physical
address space employed by C2 software 110 and HARP module 204 in
HARP memory 304. In some embodiments, memory manager 416 utilizes a
50 bit VA addressing (i.e., in excess of 1 petabyte). This allows
C2 software 110 to globally address an entire database and optimize
hardware execution of the query tasks.
[0074] Storage manager 420 is responsible for managing transfers of
data from HARP memory 304 to/from storage infrastructure 112.
Answer manager 422 is responsible for compiling the results of the
query fragments and providing the result to MySQL software 114 via
the API 116.
[0075] Update manager 424 is responsible for updating any data in
the database stored in storage infrastructure 112. Shared utilities
426 provide various utilities for the components of C2 software
110. For example, these shared utilities may include a performance
monitor, a metadata manager, an exception handler, a compression
library, a logging and recovery manager, and a data loader.
[0076] HARP manager 428 controls execution of the tasks in HARP
module 204 by setting up the machine code database instructions and
handles all interrupts from any of the hardware in HARP module 204.
In some embodiments, HARP manager 428 employs a function library
known as a Hardware Acceleration Function Library (HAFL) in order
to make its function calls to HARP module 204.
[0077] FIG. 5--Protocol Stack of C2 Software
[0078] As shown, a SQL query is received in the RDBMS layer, i.e.,
MySQL software 114. MySQL software 114 then passes the SQL query
via API 116 to C2 software 110. In C2 software 110, the SQL query
is processed and executed. At this layer, C2 software 110 also
manages retrieving data for the SQL query, if necessary, from
storage infrastructure 112 or from host system 202.
[0079] In order to communicate with HARP module 204, HARP manager
428 employs the HAFL layer in order to make its function calls to
HARP module 204. In order to allow for variances in hardware that
may exist in HARP module 204, the protocol stack may also comprise
a hardware abstraction layer. Information is then passed from C2
software 110 to HARP module 204 in the form of machine code
database instructions via an interconnect layer. As noted, this
interconnect layer may be in accordance with the well-known PCIe or
HT standards.
[0080] Within HARP module 204, the machine code database
instructions are parsed and forwarded to HARP logic 302. These
instructions may relate to a variety of tasks and operations. For
example, as shown, the protocol stack provides for systems
management, task coordination, and direct memory access to HARP
memory 304. In HARP logic 302, machine code database instructions
are interpreted for the various types of processing elements (PE).
HARP logic 302 may interface with HARP memory 304, i.e., direct
memory access by utilizing the memory management layer.
[0081] FIG. 6--HARP Logic
[0082] FIG. 6 illustrates an exemplary architecture of the HARP
logic 302. As shown, HARP logic 302 may comprise a set of
processing cores 602, 604, 606, and 608, and switching fabric 610.
Processing core 602 (as well as cores 604, 606, and 608) may
comprise a set of processing elements (PEs) 620. In the embodiment
shown, processing cores 602, 604, 606, and 608 each comprise two
PEs; of course, each processing core may comprise any number of
PEs.
[0083] In addition to its PEs, processing core 602 may comprise a
task processor 612, a memory manager 614, a buffer cache 616, and
an interconnect 618. One or more these components may be duplicated
or removed from the other processing cores 604, 606, and 608. For
example, as shown, core 602 may be the sole core that includes task
processor 612 and an interconnect 618. This architecture may be
employed because cores 602, 604, 606, and 608 are connected via
switching fabric 610 and may operate logically as a single
processor or processor core. Of course, one skilled in the art will
recognize that various redundancies may be employed in these
processing cores as desired.
[0084] Task processor 612 is the hardware that supervises the
operations of the processing cores 602, 604, 606, and 608. Task
Processor 612 is a master scheduling and control processing
element, disconnected from the direct dataflow of the execution
process for a query. Task processor 612 maintains a running
schedule of machine code database instructions which have
completed, are in progress, or are yet to execute, and their
accompanying dependencies, the Task Processor 612 may also dispatch
machine code database instructions for execution and monitor their
progress. Dependencies can be implicit, or explicit in terms of
strong intra- or inter-processor release criteria. Machine code
database instructions stalled for software-assist can be
context-switched by the Task Processor 612, which can begin or
continue execution of other independent query tasks, to optimize
utilization of execution resources in HARP logic 302.
[0085] Memory manager 614 is the hardware that interfaces HARP
memory 304. For example, memory manager 614 may employ well known
memory addressing techniques, such as translation look-aside
buffers to map the global database virtual address space to a
physical address in HARP memory 304 to access data stored in HARP
memory 304.
[0086] Buffer cache 616 serves as a small cache for a processing
core. For example, temporary results or other meta-data may be held
in buffer cache 616.
[0087] PCIe interconnect 618 is the hardware that interfaces with
host system 202. As noted, interconnect 618 may be a PCIe or HT
interconnect.
[0088] PEs 620 represent units of the hardware and circuitry of
HARP logic 302. As noted, PEs 620 utilize a novel dataflow
architecture to accomplish the query processing requested of HARP
logic 302. In particular, PEs 620 implement execution of an
assortment of machine code database instructions that are known as
Macro Ops (MOPs) and Micro Ops (UOPs). MOPs and UOPs are programmed
and executed by the PEs 620 to realize some distinct phase of data
processing needed to complete a query. MOPs and UOPs are just
example embodiments of machine code database instructions; other
types of instruction sets for high level database operations of
course may be used by the C2 solution.
[0089] PEs 620 pass logical intermediate MOP results among one
another through a variable-length dataflow of dataflow tokens,
carried across an interconnect data structure (which is a physical
data structure and not a software data structure) termed an
Inter-Macro Op Communication (IMC) path. Of note, the IMC paths and
self routing fabric 610 allow HARP module 204 to utilize a minimal
amount of reads/writes to HARP memory 304 by keeping most
intermediate results flowing through the IMCs in a pipelined,
parallel fashion. IMC may be temporarily stored in buffer caches
616 and interconnect fabric 610; however, IMCs can also be
dispatched out through interconnect 618 to other PEs 620 on another
HARP module.
[0090] In the dataflow concept, each execution step, as implemented
by a MOP and its accompanying UOP program, can apply symmetrically
and independently to a prescribed tuple of input data to produce
some tuple of result. Given the independence and symmetry, any
number of these tuples may then be combined into a list, matrix, or
more sophisticated structure to be propagated and executed in
pipelined fashion, for optimal execution system throughput. These
lists of tuples, comprised fundamentally of dataflow tokens, are
the intermediate and final results passed dynamically among the
MOPs via IMC.
[0091] Although the dataflow travels over physical links of
potentially fixed dimension, the logical structure of the contents
can be multi-dimensional, produced and interpreted in one of two
different ways: either with or without inherent, internal
formatting information. Carrying explicit internal formatting
information allows compression of otherwise extensive join
relationships into nested sub list structures which can require
less link bandwidth from fabric 610 and intermediate storage in
buffer cache 616, at the cost of the extra formatting delimiters,
increased interpretation complexity and the restriction of fixing
the interpretation globally among all consumers. Without inherent
formatting, a logical dataflow may be interpreted by the consumer
as any n-dimensional structure having an arbitrary but consistent
number of columns of arbitrary but consistent length and width. It
should be noted that the non-formatted form can be beneficial not
only in its structural simplicity, but in the freedom with which
consumer MOPs may interpret, or reinterpret, its contents depending
upon the purpose of the execution step a consumer is
implementing.
[0092] The dataflow used in realizing a given query execution can
be described by a directed acyclic graph (DAG) with one intervening
MOP at each point of flow convergence and bifurcation, one MOP at
each starting and ending point, as well as any point necessary in
between (i.e. single input & output MOP). The DAG must have at
least one starting and one ending point, although any larger number
may be necessary to realize a query. MOPs which serve as the
starting point are designed to begin the dataflow by consuming and
processing large amounts of data from local storage. Ending point
MOPs may terminate the dataflow back into local storage, or to a
link which deposits the collected dataflow (result table list) into
host CPU memory. An example of a DAG for a well-known TPC-H query
is shown in FIG. 9.
[0093] As mentioned above, MOP DAGs can physically and logically
converge or bifurcate, programmatically. The physical convergence
is accomplished with a multi-input MOPs which relate inputs in some
logical fashion to produce an output comprised of all inputs (e.g.
composition, merge, etc.). The physical bifurcation is accomplished
by means of multicast technology in the IMC fabric, which
dynamically copies an intermediate result list to multiple consumer
MOPs. These mechanisms work together to allow realization of any
desired DAG of MOP execution flow.
[0094] In the present invention, each MOP is configured to operate
directly on the compressed data in the column-store database and
realizes some fundamental step in query processing. MOPs are
physically implemented and executed by PEs 620 which, depending on
specific type, will realize a distinct subset of all MOP types.
MOPs work systematically on individual tuples extracted either from
local database storage in HARP memory 304 or the IMC dataflow,
producing output tuples which may be interpreted by one or more MOP
processes downstream.
[0095] UOPs are the low-level data manipulators which may be
combined into a MOP-specific UOP program accompanying a MOP, to
perform analysis and/or transformation of each tuple the MOP
extracts. MOPs which utilize UOP programs are aware of the
dependency, distributing selected portions of each tuple to the
underlying UOP engine, extant within all PEs 620 supporting such
MOPs. For each set of inputs from each tuple, the UOP program
produces a set of outputs, which the MOP may use in various ways to
realize its function.
[0096] For example, one manner a MOP may use UOP output is to
evaluate each tuple of a list of tuples for a set of predicating
conditions, where the MOP decides either to retain or to drop each
tuple based on the UOP result. Another manner is for the UOP to
perform an arithmetic transformation of each input tuple, where the
MOP either appends the UOP result to form a larger logical tuple,
or replaces some portion of the input tuple to form the output
tuple.
[0097] Given a finite number of execution resources in PEs 620, the
full MOP dataflow DAG needed to execute a query may be partitioned
into segments of connected MOPs called tasks. These tasks are then
scheduled by task processor 612 for execution in a sequential
fashion, as MOP execution resources become available in PEs 620.
Significant in this process is the propagation of the execution
dataflow among these tasks, such that the entire query result is
accurately and consistently computed, regardless of how each task
is apportioned and regardless of the latency between scheduling
each task.
[0098] One method that may be employed in HARP logic 302 is to
treat each task atomically and independently, terminating the
dataflow back into local storage in HARP memory 304 at the end of
each task and restarting that dataflow at the beginning of the
subsequent task by reloading it from HARP memory 304. In some
embodiments, a more efficient method may be employed to pipeline
tasks at their finer, constituent MOP granularity, where at least
one MOP of a new task may begin execution before all MOPs of the
previous task have finished. This fine-grained method is referred
to as Task Pipelining.
[0099] Keeping the dataflow alive over task boundaries is a key to
realizing the extra efficiency of Task Pipelining. To accomplish
this in the C2 solution, IMCs may include the ability to
dynamically spill, or send their dataflow to an elastic buffer
backed by HARP memory 304, pending the awakening of a consumer MOP
which will continue the dataflow. On scheduling the consumer MOP,
IMCs are able to fill dynamically, reading from the elastic buffer
in HARP memory 304 as necessary to continue execution, pulling out
any slack that may have built up in the dataflow while waiting for
the scheduling opportunity. Task Pipelining with these mechanisms
then may provide a more efficient use of execution resources, down
to the MOP granularity, such that a query may be processed as
quickly as possible.
[0100] High-latency, low-bandwidth, non-volatile storage in storage
infrastructure 112 often holds the contents of a query, due to the
sheer volume of data involved. Because execution rates can outstrip
the bandwidth available to read from such storage, tasks requiring
latent data can shorten execution time by starting and progressing
their dataflow execution at the rate the data arrives, instead of
waiting for an entire prefetch to complete before beginning
execution. This shortcut is referred to as Prefetch Pipelining. The
C2 solution may employ both on-demand prefetching and speculative
prefetching. On-demand prefetching is where data is prefetched
based on the progress of the dataflow. Speculative prefetching is
where data is prefetched based on an algorithm or heuristic that
estimates the data is likely to be requested as part of a
dataflow.
[0101] In the present invention, realizing Prefetch Pipelining is
accomplished by having one or more MOPs beginning a task's dataflow
are capable of accepting data progressively as it is read from slow
storage in storage infrastructure 112. IMCs are capable of filling
progressively as data arrives, as are all MOPs already designed to
read from local storage in HARP memory 304. Given that support,
MOPs can satisfy the requirement of executing progressively at the
rate of the inbound dataflow and accomplish efficient Prefetch
Pipelining.
[0102] As shown, processing core 602 may comprise scanning/indexing
PE 622 and XCAM PE 624 as its set of PEs 620. As noted, PEs 620 are
the physical entities responsible for executing MOPs, with their
underlying UOPs, and for realizing other sophisticated control
mechanisms. Various incarnations of processing elements are
described herein, where each incarnation supports a distinct subset
of the MOP and control space, providing different and distinct
functionality from the perspective of query execution. Each of the
different PE forms is now addressed where those which support MOPs
employing UOP programs implicitly contain a UOP processing
engine.
[0103] Scanning/Indexing PE 622 implements MOPs which analyze
database column groups stored in local memory, performing parallel
field extraction and comparison, to generate row pointers (row ids
or RIDs) referencing those rows whose value(s) satisfy the applied
predicate. For some MOP forms, a metadata Value List (which is an
abstract term for a logical tuple list flowing through an IMC)
containing a column of potentially sparse row pointers may be given
as input, in which case the scan occurs over a sparse subset of the
database. For other forms, scanning occurs sequentially over a
selected range of rows.
[0104] The selection predicate is stipulated through a micro-op
(UOP) program of finite length and complexity. For conjunctive
predicates which span columns in different column groups, scanning
may be done either iteratively or concurrently in dataflow
progression through multiple MOPs to produce the final, fully
selected row pointer list.
[0105] Inasmuch as the Scanning/Indexing PE 622 optimizes scanning
parallelism and is capable of constructing and interpreting
compacted what are known as bitmap bundles of row pointers (which
are a compressed representation of row pointers, sparse or dense,
that can be packed into logical tuples flowing through an IMC), it
operates most efficiently for highly selective predicates,
amplifying the benefits thereof. Regardless, its MOP support
locates specific database content.
[0106] Scanning/Indexing PE 622 also implements MOPs which project
database column groups from HARP memory 304, search and join index
structures, and manipulate in-flight metadata flows, composing,
merging, reducing, and modifying multi-dimensional lists of
intermediate and final results. Depending on the MOP, input is one
or more Value Lists whose content may be interpreted in a one- or
two-dimensional manner, where two-dimensional lists may have an
arbitrary number of columns (which may have arbitrary logical
width).
[0107] In the context of list reduction, a UOP program of finite
length and complexity is stipulated as a predicate function, to
qualify one or more components of the input Value List elements,
eliminating tuples which do not qualify. List composition involves
the combining of related lists into a single output format which
explicitly relates the input elements by list locality, while list
merging involves intermingling input tuples of like size in an
unrelated order. Modification of lists involves a UOP which can
generate data-dependent computations, to replace component(s) of
each input tuple.
[0108] The Scanning/Indexing PE 622 may also be used for joins with
indexes, like a Group Index, which involves the association of each
input tuple with potentially many related data components, in a
one-to-many mapping, as given by referencing the index via a row
pointer component contained in each input tuple. MOPs implemented
by the Scanning/Indexing PE 622 may thus relate elements of a
relational database in by query-specific criteria, which is useful
for any query of moderate to advanced complexity.
[0109] XCAM PE 624 implements MOPs which perform associative
operations, like accumulation and aggregation, sieving, sorting and
associative joins. Input is in the form of a two-dimensional
metadata Value List which can be interpreted as containing at least
two columns related by list locality: key and associated value.
[0110] Accumulation occurs over all data of like keys
(associatively), applying one of several possible aggregation
functions, like Summation or an atomic compare and exchange of the
current accumulator value with the input value component. A direct
map mode exists which maps the keys directly into HARP memory 304,
employing a small cache (not shown) to minimize memory access
penalties. A local mode of accumulation exists, as well, to realize
zero memory access penalties by opportunistically employing the
cache, at the risk of incomplete aggregation.
[0111] Sieving involves the progressive capture of keys qualifying
as most extreme, according to a programmable sieving function,
generating a result list of the original input keys and values such
that the last N tuples' keys are the most extreme of all keys in
the original input. Iterative application of Sieve can converge on
a sorted output, over groups of some small granularity.
[0112] Sorting can also be accomplished through construction and
traversal of either hashes or B-Trees, which are constructed to
relate each input key to its associated value with a structure that
is efficient to search and join with.
[0113] Within each of PEs 620 thus may be a UOP Processing Engine
(not shown). Whereas PEs 620 execute MOPs in a dataflow fashion at
the higher levels, embedded UOP Processing Engines in PEs 620
realize the execution of UOPs, which embed within their logical MOP
parent to serve its low-level data manipulation and analysis needs.
In some embodiments, the UOP processing engine is code-flow logic,
where a UOP program is executed repetitively by a parent Processing
Element at MOP-imposed boundaries, given MOP-extracted input data,
to produce results interpreted by the parent MOP.
[0114] Considering the code-flow nature, each UOP engine has its
own program storage, persistent register set and execution
resources. It is capable, through appropriate UOP instructions, to
accept data selected from the parent MOP and to simultaneously
execute specified data manipulation or analysis thereon, in
combination with some stored register state. In this manner, this
tiny code-flow processor is able to fit seamlessly into the
dataflow as a variable-latency element which, at the cost of
increased latency, is capable of performing any of the most complex
low-level data manipulation and analysis functions on the dataflow
pouring through. The capability of the MOP to select and present
only those data required for UOP processing, at a fine granularity,
minimizes the latency imposed by the UOP code flow, maximizing
overall dataflow throughput.
[0115] FIG. 7--C2 Data Structures
[0116] The C2 solution utilizes various hardware-friendly data
structures to assist in hardware accelerating database operations
by HARP modules 204. As described below in Table 1, the C2 solution
may employ three different columns. These columns enable efficient
utilization of HARP logic 302 and HARP memory 304, and bandwidth
and disk bandwidth with storage infrastructure 112.
TABLE-US-00001 TABLE 1 Column Formats Name Description Packed All
rows in the column are packed consecutively. All rows Column have a
uniform size in each column. A packed column can be disk-resident
or memory-resident. The packed Column can be sorted with respect to
the primary key. Sorted- This is for the date columns. The date is
stored in the Compressed form of running length. Each date has an
entry for re- Column cording its starting count and ending count.
The offset could be zero if the date does not have any item in the
table. It companion column can be accessed with staring and ending
counts. Companion The companion table links the Sorted-Compressed
Column Column to other columns.
[0117] In general, hot columns (i.e., columns having active or
frequent access) stay in the HARP memory 304 so that they can be
accessed randomly fast. Warm Columns (i.e., columns having less
active access) also stay in the HARP memory 304; but occasionally,
they may be evicted to a disk in storage infrastructure 112. Cold
columns usually be held in storage infrastructure 112, but may be
partially brought into HARP memory 304, e.g., for one time usage.
In some embodiments, date columns in the Sorted-Compressed format
will be held in the memory of host system 202 and accessed by the
software running on host 202.
[0118] In general, there is a single entry point for HARP module
204 to identify all the database columns. In particular, as shown
in FIG. 7, a root table 702 points to all the available table
descriptors 704. The table descriptors 704 in turn point to their
respective table columns 706. Each table stores multiple columns in
the VA memory space. Each of these tables will now be further
described.
[0119] As noted, root table 702 identifies all the tables accessed
by HARP module 204. In some embodiments, each entry in the table
takes 8 bytes. When needed, multiple Root Table blocks can be
chained by a next pointer. The Descriptor Pointers in the root
table 702 points to the individual table descriptors. The indices
of the Descriptor Pointers also serve as the table ID. To simplify
the hardware design, a CSR (Control Status Register) may be
employed to store the Root Table information as long as the
hardware accessible Table IDs and Descriptors' information is
retained in HARP module 204.
[0120] Each database defined table has a table descriptor 704. All
the table descriptors 704 may reside in the HARP memory 304. A
table descriptor 704 may comprise different groups of data. A group
may contain one or more columns. Within a group, the data is
organized as rows. A group of data resides in a memory plane which
is allocated to it. A data element in a particular plane has direct
reference to its corresponding element in another plane. The
relationship of the addresses among all the element pairs is the
same arithmetical computation. The table descriptor is portable
because the present invention utilizes a global virtual address
space. In other words, when copying the table descriptor from one
virtual memory location to another, all the information in the
table is still valid.
[0121] In the C2 solution, the data structures of the database are
architected to optimize database data processing in HARP hardware.
All table columns/column groups, indices and meta-data are defined
in a global database virtual address space (DBVA). A reserved DBVA
section is allocated for table descriptors 704 as part of the
meta-data. Table descriptors 704 include information about a table,
such as the table name, number of rows, number of columns/column
groups, column names, width(s) within a column group, etc. In
addition to the information of data layout and access information
in the VA space, the table descriptors 704 also have information
about the compression types/algorithms used for each individual
column. In the present invention, hardware can directly use this
information to accomplish database queries and table element
insertion, update, and deletion.
[0122] FIG. 8--Table Column Layout
[0123] FIG. 8 is now provided to provide further detail on the
structure of a table in column-store format as employed by the C2
solution of the present invention. As shown, each database table is
broken into multiple columns or column groups having a fixed width.
Variable width columns are also supported using column hopping or a
column heap structure with linked lists. In the C2 solution, a
column group can have one or more columns packed together. Because
of the simple arithmetic mapping or the single indirection in the
companion column, the hardware and software of the present
invention can easily access rows across the columns without any
degradation in performance; thus, the C2 solution can provide the
same functionality and benefits as known row store databases. Table
and column descriptors may also be embedded in the MOPs and query
tasks.
[0124] Of note, in the present invention, the columns or column
groups possess an implicit row id (RID). A RID is considered
implicit because it is not materialized as a part of a column or
column group. Instead, each column and column group is designated a
starting RID, which corresponds to an address in the global
database virtual address space, which is then mapped to a physical
address in HARP memory 304. Since each column and column group is a
fixed width, the RID can provide the basis for arithmetically
calculating the memory address of any data in the column or column
group.
[0125] In some embodiments, all columns are packed together in the
single DBVA. In addition, a meta-data structure may be employed to
facilitate certain column accesses. For example, as shown, a row
pointer primary key index may comprise a sorted list of primary
keys and their associated row id (RID) in a column or column group.
Of course, a B-tree index may be used as an alternative to this
type of index.
[0126] In the present invention, two active sets of database
regions are maintained, i.e., a main database region and an augment
region for newly added data. Query processing operates on both
regions and is accelerated by the HARP module 204. The augment
region is utilized to hold new inserted items. Optionally, the
augment region may be rolled into the main region. For example, as
shown in FIG. 8, RIDs 1-n are the main region, while RIDs n+1, etc.
comprise the augment region.
[0127] Deletion updates may be committed into the main region right
away. To alleviate the drastic changes across all the columns in a
table, the present invention may allocate a valid or invalid bit. A
row deletion in a table, therefore, becomes a trivial task of
setting the appropriate bit in every column group in the table.
[0128] FIG. 9--Example of a SQL Query
[0129] FIG. 9 shows one of the 22 TPC-H queries, query #3, and how
it would be executed using the machine code database instructions.
TPC-H queries are published by the Transaction Processing
Performance Council (TPC), which is a non-profit organization to
define benchmarks and to disseminate objective, verifiable TPC
performance data to the industry. TPC benchmarks are widely used
today in evaluating the performance of computer systems. This
particular query is a shipping priority query to find the potential
revenue and shipping priority of the orders having the largest
revenue among those that had not been shipped of a given date. The
market segment and date are randomly generated from the prescribed
range, and "BUILDING" and Mar. 15, 1995 are the example here. This
query is a complex multiple table join of three tables, CUSTOMER,
ORDERS, and LINEITEM tables.
[0130] C2 Software 110 will decompose this query into 24 MOPs to
send to HARP module 204, along with their dependency information,
which establishes the topology of the dataflow from MOP to MOP. All
MOPs are started and hardware processing begins in pipelined
fashion, with each MOP's results being fed to one or more
downstream consumers over one or more dedicated logical IMC
connections.
[0131] The responsibility of the first MOP, ScanCol(0), is to
reference HARP memory 304 to find all the customers in the CUSTOMER
table who belong to the `BUILDING` market segment, producing into
IMCO all matching CUSTOMER references in the form of one RID per
qualified row. Revindex(1) then traverses a reverse index residing
in 304, pre-built to relate customers to their one or more orders
residing in the ORDERS table, outputting references to all orders
made by the given customers. Because the CUSTOMER references are no
longer necessary and to boost performance by reducing utilization
of IMC transmission resources over IMC2, the ListProject(2) removes
the original customer references after the reverse index join,
leaving only the ORDER references. The ScanRPL(3) MOP then scans
these orders' 0 ORDERDATE column, retaining ORDER references only
to those orders whose order date occurs before the date
`1995-03-15`.
[0132] Progressing onward through IMC3, the dataflow entering
Revindex(4) consists of ORDER table references (RIDs) which have
satisfied all criteria mentioned thus far: each order was placed by
a customer in the `BUILDING` market segment before the date Mar.
15, 1995. To finish evaluating the "WHERE" clause of the
illustrated SQL query statement, these orders must be qualified in
terms of certain properties of their related line items.
[0133] The purpose of the Revindex(4) MOP is then to associate each
of the qualifying orders to its one or more constituent line items
from the LINEITEM table, returning appropriate references thereto.
At this point, the flow contains a two-column tuple list relating
ORDER references (RIDs) to LINEITEM RIDs, multicasting identical
copies of these tuples into IMC4 and IMC5. ListProject(5) extracts
only the LINEITEM RID column from the dataflow in preparation for
ProjRpl(6), which extracts each line item's L_SHIPDATE column
value, feeding these ship dates to IMC7. ListCompose(7) consumes
IMC7 along with IMC5, executing a composition of the input lists to
create a three-column tuple list where each tuple contains an ORDER
RID, an associated LINEITEM RID and its ship date. ListSelect(8)
consumes the composed list from IMC 8 and selects only those tuples
having ship date older than `1995-03-15`, thus completing the
"WHERE" clause requirements.
[0134] Again, at the output of ListSelect(8), the dataflow still
logically appears as a three-column tuple list where each tuple
relates an ORDER RID to one of its associated LINEITEM RIDs and
that line item's ship date. It should be noted in this flow that
multiple distinct LINEITEM RIDs may appear (in different tuples)
with an identical ORDER RID, a definite possibility here since a
single order may be comprised of an arbitrary number of line items
in the target database and this query specifically requests only
those line items satisfying the ship date criteria. The redundancy
of ORDER RIDs in the list suggests an aggregation step will be
needed to realize the SUM of the SQL select statement, but before
that, some more data must be gathered and calculations done.
[0135] IMC9 and IMC10 both carry the output of ListSelect(8),
identically. ListProject(9) extracts only the LINEITEM RID column
from IMC9, passing that on to both ProjRpl(12) and ProjRpl(11),
which fetch each referenced LINEITEM's L_EXTENDEDPRICE and L
DISCOUNT, respectively. Those procured extended price and discount
data are then composed together by ListCompose(13) to form a
two-column tuple to be carried via IMC17. ListTupleArith(14)
implements the arithmetic process of computing
(L_EXTENDEDPRICE*(1-L_DISCOUNT)) on a per-tuple basis before
sending this arithmetic result to ListCompose(15). In the meantime,
ListProject(10) extracts the ORDER RID column from the output of
ListSelect(8), such that ListCompose(15) can make a two-column
composition relating, within each tuple, an ORDER RID to its line
item's arithmetic product.
[0136] The final hardware step to complete the query involves fully
evaluating the SELECT clause, including its SUM aggregation
function. The remainder of the MOP flow of FIG. 9, beginning with
the output of ListCompose(15), is dedicated to this process.
[0137] AssocAccumSum(16) receives from IMC19 with each of the
two-column tuples relating an ORDER RID to one of its line item's
(L_EXTENDEDPRICE*(1-L DISCOUNT))product, computing a summation of
these values independently for each distinct ORDER RID. For
example, a given ORDER RID may appear twice in IMC19 (once in two
different tuples), having two distinct LINEITEMs which satisfied
all criteria thus far. Each of these LINEITEMs would have generated
its own product in ListTupleArith(14), such that the aggregation
process of AssocAccumSum(16) must sum them together. The result is
a distinct sum of products over each distinct ORDER RID, realizing
the SQL SUM aggregation function, here named REVENUE within the
query.
[0138] Once the aggregation has completed for a given ORDER RID,
ListProject(17) extracts the ORDER RID itself, passing it to
ProjRpl(18), ProjRpl(19) and ProjRpl(20). These MOPs gather in
parallel the referenced orders' 0 ORDERDATE, O_SHIPPRIORITY, and 0
ORDERKEY, repectively, while ListCompose(21) forms a two-column
tuple consisting of 0_SHIPPRIORITY and 0 ORDERKEY. ListCompose(22)
meanwhile forms a two-column tuple comprised of O_ORDERKEY and
REVENUE. The final MOP, ListCompose(23), composes the two
two-column tuple lists into a final four-column tuple list which
satisfies the SQL query and its SELECT statement.
[0139] It should be noted in this example that the SQL query SELECT
actually stipulates L_ORDERKEY. But an optimization may be applied
here, knowing that O_ORDERKEY is functionally equivalent when used
in this manner, thus avoiding the need to carry any LINEITEM RIDs
beyond IMC11 or IMC12.
[0140] FIG. 9--Example of a Dataflow through the HARP
[0141] In FIG. 9 we have described how an SQL statement gets mapped
into a logical MOP DAG (directed acyclic graph) which gets executed
in a dataflow fashion with IMC chaining between MOPs. FIG. 10
illustrates an exemplary dataflow through PEs 620 in HARP logic 302
for the same TPC-H SQL #3 query shown in FIG. 9. As noted, C2
Software 110 will decompose this query task into 10 PE stages to
send to HARP module 204, along with their MOP and UOP instructions
and dependency information.
[0142] Stage 1 is performed by Scanning PE 1002 is to find all the
customers in CUSTOMER table that is in "BUILDING" market segment
and passes the results (C_RIDs of matching customer records) in an
IMC to Indexing PE 1004.
[0143] Stage 2 is a join operation of C_CUSTKEY=O_CUSTKEY performed
by Indexing PE 1004 using a reverse index method. Each C_RID of
Stage 1's matching customer records corresponds to an O_RID hitlist
of ORDER table records, given a customer may place multiple orders.
The results (O_RIDs) are passed in an IMC to Scanning PE 1006.
[0144] Stage 3 is performed by Scanning PE 1006 to read the 0
ORDERDATE field of all the matching orders (ORIDs) that Stage 2
outputs, compare for "<`1995-03-15", and passes the results
(O_RIDs) in an IMC to Indexing PE 1008.
[0145] Stage 4 is a join operation of O_ORDERKEY=LORDERKEY
performed by Indexing PE 1008 using a reverse index method. Each
O_RID of Stage 3's matching order records corresponds to an L_RID
hitlist of LINEITEM table records, given an order may have multiple
line items. The results (L_RIDs) are passed in an IMC to Scanning
PE 1010.
[0146] Stage 5 is performed by Scanning PE 1010 to read the
L_SHIPDATE field of all matching line items (L_RIDs) that Stage 4
outputs, compare for ">`1995-03-15", and passes the results
(L_RIDs) in 3 IMCs to Indexing PE 1012, 1014, and 1016.
[0147] Stage 6 is a column extraction/projection operation done by
Indexing PE 1012, 1014, and 1016 to get L_ORDERKEY,
L_EXTENDEDPRICE, and L_DISCOUNT column.
[0148] Stage 7 is a list merge operation of 2 columns
(L_EXTENDEDPRICE and L_DISCOUNT) done by Indexing PE 1018.
[0149] Stage 8 is an aggregation operation of REVENUE of each
L_ORDERKEY group, done by XCAM PE 1020 based on outputs of Indexing
PE 1012 and 1018. As the SQL statement defines, REVENUE is
calculated as the sum of (L_EXTENDEDPRICE*(1-L_DISCOUNT)). Note
that even though the GROUP BY defines the group key as
concatenation of L_ORDERKEY, O_ORDERDATE, O_SHIPPRIORITY, the group
key is simplified to L_ORDERKEY since it is already a unique
identifier. The output of XCAM PE 1020 is a pair list of group key
(L_ORDERKEY) with its REVENUE.
[0150] Stage 9, done by Indexing PE 1022 and 1024, is a column
extraction of O_ORDERDATE based on L_ORDERKEY output of XCAM PE
1020.
[0151] Stage 10, done by XCAM PE 1026, is a sieve (ORDER BY)
operation of REVENUE, O_ORDERDATE to output top N groups with
largest REVENUEs. These outputs are placed at a "result" buffer
area in HARP memory 304, ready to be retrieved by DBMS software
114.
[0152] FIG. 10--Example of a Dataflow through the HARP
[0153] FIG. 10 illustrates an exemplary dataflow through PEs 620 in
HARP logic 302 for the same TPC-H SQL #3 query shown in FIG. 9. As
noted, C2 Software 110 will decompose this query into 7 query tasks
to send to HARP module 204, along with their task command and
dependency information.
[0154] Task 1 is performed by a Scanning PE is to find all the
customers in CUSTOMER table that is in BUILDING market segment and
passes the results in an IMC to an Indexing PE.
[0155] Task 2 is performed by an Indexing PE to extract the
C_CUSTKEY field (column) of Task 1's matching customer records and
search against the 0 CUSTKEY index table and passes the results in
an IMC to a Scanning PE.
[0156] Task 3 is performed by the Scanning PE to read the
O_ORDERDATE field of all the matching orders that Task 2 returns
and passes the results in an IMC. Given a customer could have
placed multiple orders, Task 2's CUSTKEY search could result in
multi-record hit lists. An Indexing PE then extracts the
appropriate records and passes the results in an IMC to a Scanning
PE.
[0157] Task 4 is performed by a Scanning PE to compare the 0
ORDERDATE of each order record from Task 3 and returns only those
records with dates before Mar. 15, 1995 in an IMC to a set of
Indexing PEs.
[0158] Task 5 is performed by a set of Indexing PEs and then an
XCAM PE to extract the 0 ORDERKEY field of Task 4's qualified
orders and search against the L_ORDERKEY index table.
[0159] Task 6 is performed by a set of Indexing PEs to extract the
L SHIPDATE field of all the matching line items that Task 5
returns.
[0160] Task 7 is then performed by an XCAM PE to compare the L
SHIPDATE of each line item record from Task 6 and sort those
records with shipping date later than Mar. 15, 1995. The sorted
line item records are placed at a result buffer area in HARP memory
304, ready to be retrieved by DBMS software 114.
[0161] Explanation of the Global Database Virtual Addressing
Scheme.
[0162] As noted, the C2 solution of the present invention can
employ an arbitrarily large virtual address space (DBVA) to
globally address an entire database. This addressing scheme is
utilized to assist hardware acceleration, because it allows HARP
module 204 and C2 software 110 to arithmetically calculate the
location of any data in a database. FIG. 11 illustrates an
exemplary logic flow of the addressing schemes employed and FIG. 12
illustrates a structure of the global database virtual address
scheme of the present invention.
[0163] Referring now to FIG. 11, when a query is received from
MySQL software 114, it is eventually submitted to query execution
engine 412 for processing. At this level, execution engine 412
operates on the database using what are known as logical addresses.
A logical address is utilizes the naming convention of the SQL
query. For example, a logical address may consist of a database
name, a table name, a column group name, and a column name which is
then converted to a DBVA. Since it is a virtual address, the DBVA
may be arbitrarily large, e.g., in excess of a petabyte.
[0164] The DBVA may then be converted to physical addresses of the
data in storage 112 or in the memory of host system 202. For
example, as shown, a storage manager 420 may have a mapping that
converts a DBVA to a storage physical address (SPA). This allows
for retrieval of data from storage 112. In addition, buffer manager
414 may convert the DBVA to a physical address in the memory of
host system 202, for example, using a standard memory manager in
the operating system running on host system 202. Such conversions
are well known to those skilled in the art.
[0165] However, of note, the present invention may employ an
extensible addressing scheme for one or more HARP modules 204. As
noted, HARP modules 204 may comprise a relatively large memory,
i.e., in excess of 256 gigabytes or more. In addition, the C2
solution may comprise nodes that have multiple HARP modules 204 and
may also support nodes. Accordingly, the present invention may
comprise a novel, layered virtual addressing scheme.
[0166] In one embodiment, each HARP module 204 translates the DBVA
to a physical address (HPA) of HARP memory 304. For example, HARP
module 204 may utilize a translation-lookaside buffer (TLB) for
this translation.
[0167] Alternatively, especially in a multi-HARP module environment
or federated system, each HARP module 204 may employ a secondary
virtual address space (called an HVA) that is underlies the DBVA.
Referring now to FIG. 12, in this embodiment, memory manager 418
may thus include a DBVA-HVA service that maps a DBVA address into
yet another virtual address in the HVA. Memory manager 418 may then
translate the HVA into a HPA using, for example, a TLB for the
respective HARP memory 304.
[0168] This feature allows the present invention to continue
utilizing a single global virtual address space for a database even
where system 100 includes multiple HARP modules 204 or multiple
nodes of HARP modules. Accordingly, systems of the C2 solution can
continue to employ hardware-friendly addressing for hardware
acceleration of the database operations regardless of the number of
HARP modules 204 or the size of the database.
[0169] Multi-Versioning and Continuous Updates
[0170] Embodiments of the present invention can provide fine grain
concurrency control in the presence of database updates. For
read-only transactions, historical snapshots of the data in pages
in HARP memory 304 are created. In addition, historical snapshots
can be created for pages in the memory of host 202 and in storage
112. For illustrative purposes, the following description is
directed towards multi-versioning and continuous updates of data in
various pages of HARP memory 304. However, one skilled in the art
will recognize how the concepts of the present invention for
multi-versioning and continuous updates can be extended to host 202
and storage 112.
[0171] The following description begins by generally explaining
various concepts of multi-versioning and continuous updates that
may be implemented by embodiments of the present invention. FIGS.
11 and 13 illustrate an example of how multi-versioning and
continuous updates may be accomplished.
[0172] Embodiments of the present invention provide fine grain
concurrency control for transactions in the presence of database
updates. During operations, each transaction is assigned a snapshot
version number or SVN. A SVN refers to a historical snapshot of the
database that can be created periodically or on demand.
Transactions are thus tied to a particular SVN, such as, when the
transaction was created. Queries belonging to the transactions can
access data that is consistent as of a point in time, for example,
corresponding to the latest SVN when the transaction was
created.
[0173] At various times, data from the database stored in a memory
can be updated using the snapshot data corresponding to a SVN. For
example, when a transaction is committed, a snapshot of the
database with a new SVN is created based on the data modified by
the transaction and the snapshot is synchronized to the memory. As
another example, when a transaction query requires data from a
version of the database corresponding to a SVN, the data in the
memory may be synchronized with the snapshot data corresponding to
that SVN. This feature, in essence, creates multiple versions of
the database and allows read-intensive database applications to
operate in the presence of database updates.
[0174] A query can be isolated to a consistent historical snapshot.
Changes made by committed update transactions may be visible to
other transactions unless they are operating on a historical
snapshot. Queries and data manipulation are executed within a
transaction boundary. In other words, a transaction serves to
provide a consistent read for its related queries and query tasks.
When multiple queries are executed within the same transaction,
these queries will also operate on the same version. Two modes of
transactions, serializable and snapshot serializable may be
supported.
[0175] A serializable mode transaction is exposed to the latest
committed changes. Snapshot serializable transactions are exposed
only to the snapshot version they were assigned when they started.
If a transaction does not exist when the query/data manipulation is
executed, C2 software 110 may start a transaction automatically and
auto commit when the statement is processed.
[0176] For locking, C2 software 110 may rely on MySQL software 114
for table locking. Internally, memory manager 418 can support page
level locking in HARP memory 304 which can be used to provide more
concurrency if required. For serializable transactions, write locks
are issued so that no other serializable transaction can access the
same data. For snapshot serializable transactions, transactions are
not blocked on account of the serializable transactions.
[0177] In the present invention, there are several embodiments for
handling updates during the data manipulation (DML) processing. In
one approach, changes made during the DML processing are not made
to the data inline, but stored in offline auxiliary structures.
This allows queries to run interrupted.
[0178] Alternatively, a second approach may be employed that
involves keeping shadow copies of pages touched by transactions,
but maintaining a different version until the changes are synced.
If the pages need to be flushed to storage 112 after changing in
the memory of host system 202, the storage manager 420 may maintain
two versions of the pages, one for the system SVN and the next
version. In this approach, the queries are relatively uninterrupted
since the version needed by them is not affected by the
transactions.
[0179] Another approach can utilize a change log. In some
embodiments, the change log can be in the form of undo or redo
operations. In this approach, the changes are made directly to the
pages in the buffer cache in the same page (rather than a shadow
copy). These changes are made in the buffer cache and may get
flushed to storage 112 but are not made in HARP memory 304 until
prefetched for query processing. A status map may be kept by C2
software 110 to indicate the status of data on memory in host
system 202, storage 112, and HARP memory 304. Thus, when a prefetch
request is made for certain data, this map is consulted to ensure
that data is present in the version required and if not, a change
log can be used to construct it.
[0180] Non-Blocking Continuous Updates by the C2 Solution
[0181] Referring now to FIGS. 13-14, an embodiment of continuous
updating by the C2 solution is conceptually illustrated. As shown
in FIG. 13, a timeline of query transactions X, Y, and Z are shown.
As time progresses, snapshots are generated periodically (or at
commit time for sync on commit).
[0182] Snapshots may be managed by a snapshot manager in utilities
426 and, as noted, are identified by their SVN. In some
embodiments, the snapshots may be managed by update manager 424.
Snapshots can be created in various ways. In snapshot on commit, a
snapshot is created from the data modified by a transaction at
commit time. In a timed snapshot, several transactions' committed
data is buffered and associated with a snapshot based on a timer,
which can be user specified.
[0183] A serializable mode transaction does not bind to an SVN. A
snapshot mode transaction binds to an SVN. Pages are associated
with a specific SVN. A newly created snapshot transaction is
associated with the current SVN. A query task inherits the SVN from
the transaction it belongs to. All the tasks belonging to a query
will inherit that SVN too.
[0184] Synching is the process of bringing a page to the required
SVN. Pages can be on different SVNs. In order for a task to
execute, it's required that all the pages touched by that task are
present at the correct SVN. For example, as shown, for various
pages in HARP memory 304, update manager 424 manages a snapshot
version number (SVN), e.g., SVN2, SVN3, and SVN4 and a set of undo
and redo tables for each SVN. Each transaction is associated with
the SVN of the current snapshot at the time it is created. This SVN
is used for all queries and tasks executed for that transaction.
Snapshots provide approximate up-to-date data for queries while
preserving performance since many queries and tasks may work on the
same snapshot of data. In some embodiments, the snapshot data
structures may be stored in B+ trees. In these data structures,
uncommitted data of active transactions and committed data since
the last snapshot are tracked.
[0185] In some embodiments, the SVN may simply be a value that is
monotonically increasing and assigned by update manager 424. SVNs
can be recycled when the system restarts. In some embodiments,
there is a single global snapshot version number maintained by
update manager 424 that all snapshot serializable queries can
access. Serializable transactions see the version of data for a
page in HARP memory 204 that includes committed changes that have
occurred since the last sync. When the committed changes are
synced, the SVN is incremented.
[0186] When a transaction starts, it is assigned a snapshot version
(the current system SVN). This is the version of the data that all
queries within this transaction will see for the duration of the
transaction. Hence, as shown in FIG. 13, query transaction X is
assigned SVN1 since it started during SVN1. Likewise, query
transactions Y and Z are assigned SVN3 and SVN4, respectively. Of
note, as shown, query transactions may span across multiple SVNs,
but may be referenced according to the SVN in which they
started.
[0187] A query executes on the snapshot with the transaction's SVN.
For example, as shown in FIG. 13, the three shown query tasks in
query transaction X operate on data from SVN 1. The query tasks in
transaction Y operate on data from SVN2, and query tasks in
transaction Z will operate on data from SVN4. The transaction's SVN
is determined at transaction start time. If a transaction does not
exist then one may be automatically created that lasts for the
duration of the query.
[0188] However, as shown, some tasks may be submitted at a
different SVN from the SVN assigned to their transaction. For
example, transaction X may include a task that starts in SVN2
rather than SVN1. Transaction Y may include a task that starts in
SVN3 rather than SVN2. These tasks may thus need to have one or
more pages synched to a different SVN, which will now be explained
with reference to FIGS. 14-16.
[0189] Referring now to FIG. 14, transitions between SVNs may be
mapped using redo and undo tables that are logged in the snapshot
data structures. As shown, a page from HARP memory 304 may be
synched between SVN2 and SVN3, SVN3 and SVN4, and SVN2 and SVN4.
The redo table is used to convert the data in an older snapshot to
a newer snapshot. The undo table is used to convert the data from a
newer snapshot to an older snapshot. The redo table from a snapshot
can also be maintained to the next snapshot and the undo table from
a given snapshot to the prior snapshot. This data thus allows for
reconstruction of snapshots with minimal overhead.
[0190] In some embodiments, these tables are implemented in 32 byte
chunks. These tables may map a VA to a 32 byte chunk of data. The
VA is a 32 byte-aligned address. The VAs may be in sorted
order.
[0191] The task manager will execute a prefetch task to load the
data required by the query or task for its SVN. The prefetch task
is submitted as part of the query execution plan generated by query
plan generator 410. The memory manager 418 maintains page to SVN
mapping. If a page that needs to be prefetched is missing or at the
wrong SVN, then the page may be updated to the correct SVN.
[0192] During operation, tasks for queries that are part of a
transaction are submitted for execution on HARP module 204. These
tasks are submitted with the SVN required for that task so that the
task manager 416 can handle the task appropriately when it reorders
this task across sync barriers. In addition, elsewhere, queries and
query tasks in these transactions may request blocks of this SVN
from the memory in host system 202.
[0193] A snapshot serializable transaction starting at SVN2 will be
shown those HARP memory pages having a SVN2. Meanwhile, other
transactions may make changes to data after this point and the
pages touched by these transactions will be marked with a different
SVN, such as SVN3 or SVN4, when they commit.
[0194] A transaction starting in serializable mode may be able to
see these changes that have not been synced yet. Since all
committed changes for multiple transactions go into the same SVN, a
transaction in serializable mode when it starts must see a
consistent read throughout its lifetime. For example, if a query
task starts in serializable mode and can see committed changes
which are part of SVN2, another query task may come along and make
more changes into SVN2 but these will not be visible to that first
query task.
[0195] In the present invention, a way to solve this is for query
tasks to acquire its table level lock in write mode even if it just
wants to do queries. This does not impact snapshot serializable
queries which only lock in read mode and can always progress.
Alternately, the locks held by memory manager 418 can be held for
transaction duration which will achieve the same effect.
[0196] Then pages are being changed, their state across the
different components may need to be tracked. As noted, in some
embodiments, C2 software maintains track of the SVNs of pages in
storage 112, host system 202, and HARP memory 204. In some
embodiments, a page map is implemented in the form of a hash table
having the address prefix of each page.
[0197] In some embodiments, an isolation level may specify that all
transactions occur in a completely isolated fashion; i.e., as if
all transactions in the system had executed serially, one after the
other. Read committed isolation can be supported by releasing a
read lock on a table at the completion of the read.
[0198] FIG. 15 illustrates an example of converting a page of HARP
memory 304 from SVN2 to SVN3 based on information in a redo table.
As shown, HARP logic 302 may utilize one of its XCAM PEs 624 to
transition a page of HARP memory 304 between SVN2 to SVN3. In
particular, a redo table may be fed into XCAM PE 624, which then
performs an associative accumulate function to change values in the
page of HARP memory 304. Further details of this function are shown
with reference to FIGS. 17 and 18.
[0199] FIG. 16 illustrates an example of converting a page of HARP
memory 304 from SVN3 to SVN2 based on information in an undo table.
As shown, HARP logic 302 may utilize one of its XCAM PEs 624 to
transition a page of HARP memory 304 between SVN3 to SVN2. In
particular, an undo table may be fed into XCAM PE 624, which then
performs an associative accumulate function to change values in the
page of HARP memory 304. Further details of this function are shown
with reference to FIGS. 17 and 18.
[0200] FIGS. 17-18 illustrate associative accumulation machine code
database instructions that may be employed to convert pages between
SVNs. In particular, FIG. 17 illustrates a full replace function,
while FIG. 18 illustrates a selective replace function.
[0201] In general, an associative operation implemented by XCAM PE
624 involves the writing of each input value, mapped by its
associated key value, to a corresponding location in HARP memory
304. Since the input keys may vary arbitrarily in their order and
value, this process can implement a scattering of writes to desired
portions of HARP memory 304. The scattering writes may be performed
in a manner particularly useful for making anything from large,
sweeping to small, incremental updates of database column contents.
As noted, at least two approaches may be available: full replace
shown in FIG. 17 and selective replace shown in FIG. 18.
[0202] The full replace function (or coalescing form) shown in FIG.
17 first collects writes to adjacent memory locations over a window
of address space, before dispatching them to HARP memory 304. While
avoiding reads of original memory contents, coalescing takes
advantage of the potential for spatial locality in the input key
stream, thereby minimizing the write and nullifying the read
bandwidth required, to provide a high performance solution for
executing updates of memory 304. Since reads are avoided and since
the granularity of the update size can be much less than that of
the coalescing window, any key location not addressed within the
window may have its associated memory contents cleared to a
constant value or assigned otherwise undefined data. A small cache
may be utilized to perform the coalescing.
[0203] Referring now to FIG. 18, the selective replace function is
shown. This function may also be considered a non-coalescing form
and may be more precise and selective in updating HARP memory 304,
preserving data in adjacent locations, but may utilize higher
memory read and write bandwidth requirements. This approach is
especially useful for small update granularity (e.g. 1-bit) where
spatial locality cannot be guaranteed among the updates requested
in the input stream and preservation of surrounding data is
necessary. Again, a small cache may be used to mitigate unnecessary
memory accesses, in an opportunistic manner.
[0204] Other embodiments of the invention will be apparent to those
skilled in the art from consideration of the specification and
practice of the invention disclosed herein. It is intended that the
specification and examples be considered as exemplary only, with a
true scope and spirit of the invention being indicated by the
following claims.
* * * * *