U.S. patent application number 16/501631 was filed with the patent office on 2019-11-21 for system for organizing and fast searching of massive amounts of data.
This patent application is currently assigned to CUMULUS SYSTEMS INC.. The applicant listed for this patent is SANDEEP BELE, AJIT BHAVE, CUMULUS SYSTEMS INC., SAI KRISHNAM NADIMPALLI, ARUN RAMACHANDRAN. Invention is credited to Sandeep Bele, Ajit Bhave, Sai Krishnam Raju Nadimpalli, Arun Ramachandran.
Application Number | 20190354547 16/501631 |
Document ID | / |
Family ID | 48042844 |
Filed Date | 2019-11-21 |
View All Diagrams
United States Patent
Application |
20190354547 |
Kind Code |
A1 |
Bhave; Ajit ; et
al. |
November 21, 2019 |
System for organizing and fast searching of massive amounts of
data
Abstract
A system to collect and store in a special data structure
arranged for rapid searching massive amounts of data. Performance
metric data is one example. The performance metric data is recorded
in time-series measurements, converted into unicode, and arranged
into a special data structure having one directory for every day
which stores all the metric data collected that day. The
performance metric data is collected by one or more probes running
on machines about which data is being collected. The performance
metric data is compressed prior to transmission to a server over
any data path. The data structure at the server where analysis is
done has a subdirectory for every resource type. Each subdirectory
contains text files of performance metric data values measured for
attributes in a group of attributes to which said text file is
dedicated. Each attribute has its own section and the performance
metric data values are recorded in time series as unicode hex
numbers as a comma delimited list. Analysis of the performance
metric data is done using regular expressions. For speed, cache
memory is used. Performance metric data outside the start time and
end time named by the user on a query screen is eliminated before
the regular expression is applied.
Inventors: |
Bhave; Ajit; (Palo Alto,
CA) ; Ramachandran; Arun; (Cupertino, CA) ;
Nadimpalli; Sai Krishnam Raju; (Bangalore, IN) ;
Bele; Sandeep; (Chinchwadgaon, IN) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
BHAVE; AJIT
RAMACHANDRAN; ARUN
NADIMPALLI; SAI KRISHNAM
BELE; SANDEEP
CUMULUS SYSTEMS INC. |
PALO ALTO
PALO ALTO
BANGALORE
CINCHWADGAON
CAMPBELL |
CA
CA
CA |
US
US
IN
IN
US |
|
|
Assignee: |
CUMULUS SYSTEMS INC.
CAMPBELL
CA
|
Family ID: |
48042844 |
Appl. No.: |
16/501631 |
Filed: |
May 13, 2019 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
15183717 |
Jun 15, 2016 |
10387475 |
|
|
16501631 |
|
|
|
|
13853925 |
Mar 29, 2013 |
9396287 |
|
|
15183717 |
|
|
|
|
13200996 |
Oct 5, 2011 |
|
|
|
13853925 |
|
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
G06F 11/3409 20130101;
H04L 43/12 20130101; G06F 40/126 20200101; G06F 40/146 20200101;
G06F 16/2455 20190101; G06F 11/3072 20130101; G06F 16/90335
20190101; H04L 29/06047 20130101; H04L 29/08099 20130101; G06F
16/148 20190101; G06F 21/552 20130101; G06F 16/258 20190101; G06F
16/2468 20190101; G06F 16/334 20190101; G06F 11/3006 20130101; H04L
29/08072 20130101; G06F 16/2452 20190101; G06F 16/31 20190101; H04L
29/06 20130101; H04L 67/42 20130101; G06F 16/185 20190101; G06F
16/3331 20190101; G06F 11/3476 20130101; G06F 16/9024 20190101;
G06F 16/335 20190101; H04L 43/04 20130101; G06F 40/106 20200101;
G06F 2216/03 20130101; G06F 16/2453 20190101; H04L 43/14 20130101;
G06F 16/283 20190101; G06F 16/24522 20190101; G06F 16/245 20190101;
H04L 43/022 20130101; G06F 16/36 20190101; H04L 67/025 20130101;
G06F 16/2423 20190101; G06F 16/2477 20190101; G06F 16/951 20190101;
G06F 16/9535 20190101; G06F 16/86 20190101; G06F 16/90344 20190101;
G06F 16/383 20190101; G06F 40/205 20200101; H04L 69/329 20130101;
G06F 11/3082 20130101; G06F 16/248 20190101; G06F 16/3329 20190101;
G06F 16/284 20190101; G06F 16/338 20190101; G06F 16/9032 20190101;
H04L 12/4625 20130101; G06F 16/43 20190101; G05B 23/0272 20130101;
G06F 16/3337 20190101 |
International
Class: |
G06F 16/36 20060101
G06F016/36; G06F 16/903 20060101 G06F016/903; G06F 11/30 20060101
G06F011/30; G06F 11/34 20060101 G06F011/34; H04L 12/26 20060101
H04L012/26; G06F 16/2452 20060101 G06F016/2452; G06F 16/9535
20060101 G06F016/9535; G06F 16/9032 20060101 G06F016/9032; G06F
16/901 20060101 G06F016/901; G06F 16/33 20060101 G06F016/33; G06F
16/2458 20060101 G06F016/2458; G06F 16/2455 20060101 G06F016/2455;
G06F 16/2453 20060101 G06F016/2453; G06F 16/242 20060101
G06F016/242; G06F 16/951 20060101 G06F016/951; G06F 16/335 20060101
G06F016/335; G06F 16/28 20060101 G06F016/28; G06F 16/25 20060101
G06F016/25; G06F 16/245 20060101 G06F016/245; G06F 16/185 20060101
G06F016/185; G06F 16/14 20060101 G06F016/14; G06F 16/84 20060101
G06F016/84; G06F 16/43 20060101 G06F016/43; G06F 17/27 20060101
G06F017/27; G06F 21/55 20060101 G06F021/55; H04L 29/08 20060101
H04L029/08; H04L 29/06 20060101 H04L029/06; G05B 23/02 20060101
G05B023/02; G06F 16/31 20060101 G06F016/31; G06F 16/332 20060101
G06F016/332; G06F 16/338 20060101 G06F016/338; G06F 16/383 20060101
G06F016/383; H04L 12/46 20060101 H04L012/46 |
Claims
1. An apparatus comprising: a memory storing performance metric
data of one of more resource types, each resource type having one
or more attributes, said performance metric data having been
converted to Unicode prior to storage; a processor programmed with
at least an operating system and a search program for searching
said memory; said search program having the capability to receive
one or more partial syntaxes, each comprised of a resource type, an
attribute and a regular expression giving search conditions; and
said search program capable of receiving two or more user-defined
partial syntax expressions, one or more of said partial syntax
expression for resource types which are related to the resource
type of a first partial syntax, and capable of searching the first
partial syntax for results which qualify, and applying said results
as the starting point for evaluating said one or more related
resource type partial syntaxes for results which qualify, and
merging the evaluation results of all the partial syntaxes.
2. [user defined start time and stop time] The apparatus of claim 2
wherein said search program is capable of reading one or more time
range boxes which contain a user supplied start time and stop time
for said search and restricting data accessed from said memory to
data gathered between and including said start time and said stop
time.
3. [reverse mapping] The apparatus of claim 2 wherein said search
program is capable of recognizing performance metric data in the
form of numbers in said regular expression expressing filter
conditions in each partial syntax and converting each said
performance metric number to Unicode before carrying out said
search.
4. [template search] The apparatus of claim 1 where said search
program is capable of recognizing identification codes in said
regular expression representing prearranged searches which are
stored in advance and is capable of generating the proper syntax
for an identification code to cause said search program to
implement the appropriate prearranged search.
5. [user defined variables in template search] The apparatus of
claim 4 wherein said search program is capable of recognizing user
defined variables for one or more variables in a template search
identified by an identification code and substituting said user
defined variable(s) into said template search before running said
search.
6. [plain language search] The apparatus of claim 1 wherein said
search program is capable of accepting plain language searches and
converting them to the proper syntax for the partial syntaxes
before executing said search.
7. [data in said specially defined file system] The apparatus of
claim 1 wherein said memory stores said Unicode in a file system
organized with one top level directory per unit of time where a
unit of time is the time over which performance metric data is
gathered for one attribute, one performance metric number gathered
per one subunit of time, said top level directory storing one or
more subdirectories, one per resource type, each subdirectory
storing one or more files of Unicode data gathered from said one or
more attributes of said resource type to which said subdirectory is
devoted, each said file having a one or more sections, each for
storing the data from one of the related attributes of said
resource in a plurality of time slots, each for storing one Unicode
letter and each of which corresponds to the timeslot during which
the underlying performance number was gathered.
8. [metadata for each performance number stored in the file system
itself] The apparatus of claim 7 wherein each item of performance
metric data has metadata comprising at least the time of day during
which it was gathered, the attribute from which it was gathered,
the date on which it was gathered, and wherein the metadata of each
performance metric number is stored in the file system itself.
9. An apparatus comprising: a memory storing performance metric
data of one of more resource types, each resource type having one
or more attributes, said performance metric data having been
converted to Unicode prior to storage; a processor programmed with
at least an operating system and a search program for searching
said memory; said search program having the capability to receive
one or more partial syntaxes, each comprised of a resource type, an
attribute and a regular expression giving search conditions; and
said search program capable of receiving first, second, third,
fourth and fifth or more user-defined partial syntax expressions,
said second and fourth partial syntax expression for resource types
which are related to the resource type of said first partial
syntax, and third and fifth partial syntaxes being for resource
types which are related to said second and fourth partial syntax,
respectively, said search program being capable of searching the
first partial syntax for results which qualify, and applying said
results as the starting point for evaluating whether said second
and fourth partial syntaxes have results which qualify, said search
program being capable of using the results which qualify by using
said second and fourth partial syntaxes and applying said results
as the starting point for evaluating whether said third and fifth
partial syntaxes have results which qualify, and merging the
qualifying results of all the partial syntaxes.
10. [user defined start time and stop time] The apparatus of claim
9 wherein said search program is capable of reading one or more
time range boxes which contain a user supplied start time and stop
time for said search and restricting data accessed from said memory
to data gathered between and including said start time and said
stop time.
11. [reverse mapping] The apparatus of claim 9 wherein said search
program is capable of recognizing performance metric data in the
form of numbers in said regular expression expressing filter
conditions in each partial syntax and converting each said
performance metric number to Unicode before carrying out said
search.
12. [template search] The apparatus of claim 9 where said search
program is capable of recognizing identification codes in said
regular expression representing prearranged searches which are
stored in advance and is capable of generating the proper syntax
for an identification code to cause said search program to
implement the appropriate prearranged search.
13. [user defined variables in template search] The apparatus of
claim 12 wherein said search program is capable of recognizing user
defined variables for one or more variables in a template search
identified by an identification code and substituting said user
defined variable(s) into said template search before running said
search.
14. [plain language search] The apparatus of claim 9 wherein said
search program is capable of accepting plain language searches and
converting them to the proper syntax for the partial syntaxes
before executing said search.
15. [data in said specially defined file system] The apparatus of
claim 9 wherein said memory stores said Unicode in a file system
organized with one top level directory per unit of time where a
unit of time is the time over which performance metric data is
gathered for one attribute, one performance metric number gathered
per one subunit of time, said top level directory storing one or
more subdirectories, one per resource type, each subdirectory
storing one or more files of Unicode data gathered from said one or
more attributes of said resource type to which said subdirectory is
devoted, each said file having a one or more sections, each for
storing the data from one of the related attributes of said
resource in a plurality of time slots, each for storing one Unicode
letter and each of which corresponds to the timeslot during which
the underlying performance number was gathered.
16. [metadata for each performance number stored in the file system
itself] The apparatus of claim 15 wherein each item of performance
metric data has metadata comprising at least the time of day during
which it was gathered, the attribute from which it was gathered,
the date on which it was gathered, and wherein the metadata of each
performance metric number is stored in the file system itself.
17. An apparatus comprising: a memory storing performance metric
data of one of more resource types, each resource type having one
or more attributes, said performance metric data having been
converted to Unicode prior to storage; a processor programmed with
at least an operating system and a search program for searching
said memory; said search program having the capability to receive
one or more partial syntaxes, each comprised of a resource type, an
attribute and a regular expression giving search conditions; and
said search program capable of receiving first, second, third,
fourth, fifth, sixth and seventh or more user-defined partial
syntax expressions, said second partial syntax expression for a
resource type which is related to the resource type of said first
partial syntax, and third partial syntax being for a resource type
which is related to said resource type of second partial syntax,
said fourth partial syntax expression being for a resource type
which is related to the resource type of said third partial syntax,
said fifth partial syntax expression being for a resource which is
related to the resource type of said second partial syntax, said
sixth partial syntax expression being for a resource type which is
related to the resource type of said first partial syntax, said
seventh partial syntax expression being for a resource type which
is related to the resource type of said sixth partial syntax, said
search program being capable of searching the first partial syntax
for results which qualify, and applying said results as the
starting point for evaluating whether said second and sixth partial
syntaxes have results which qualify, said search program being
capable of using the results which qualify by using said sixth
partial syntaxes and applying said results as the starting point
for evaluating whether said seventh partial syntax have results
which qualify, said search program being capable of using the
results which qualify when using said second partial syntax as the
starting point for evaluating whether said third and fifth partial
syntaxes have results which qualify, said search program being
capable of using the results which qualify when using said third
partial syntax as the starting point for evaluating whether said
fourth partial syntax has results which qualify, and merging the
qualifying results of evaluation of said: first, second, third and
fourth partial syntaxes, and first, second and fifth partial
syntaxes, and first, sixth and seventh partial syntaxes.
18. [user defined start time and stop time] The apparatus of claim
17 wherein said search program is capable of reading one or more
time range boxes which contain a user supplied start time and stop
time for said search and restricting data accessed from said memory
to data gathered between and including said start time and said
stop time.
19. [reverse mapping] The apparatus of claim 17 wherein said search
program is capable of recognizing performance metric data in the
form of numbers in said regular expression expressing filter
conditions in each partial syntax and converting each said
performance metric number to Unicode before carrying out said
search.
20. [template search] The apparatus of claim 17 where said search
program is capable of recognizing identification codes in said
regular expression representing prearranged searches which are
stored in advance and is capable of generating the proper syntax
for an identification code to cause said search program to
implement the appropriate prearranged search.
21. [user defined variables in template search] The apparatus of
claim 17 wherein said search program is capable of recognizing user
defined variables for one or more variables in a template search
identified by an identification code and substituting said user
defined variable(s) into said template search before running said
search.
22. [data in said specially defined file system] The apparatus of
claim 17 wherein said memory stores said Unicode in a file system
organized with one top level directory per unit of time where a
unit of time is the time over which performance metric data is
gathered for one attribute, one performance metric number gathered
per one subunit of time, said top level directory storing one or
more subdirectories, one per resource type, each subdirectory
storing one or more files of Unicode data gathered from said one or
more attributes of said resource type to which said subdirectory is
devoted, each said file having a one or more sections, each for
storing the data from one of the related attributes of said
resource in a plurality of time slots, each for storing one Unicode
letter and each of which corresponds to the timeslot during which
the underlying performance number was gathered.
23. [metadata for each performance number stored in the file system
itself] The apparatus of claim 22 wherein each item of performance
metric data has metadata comprising at least the time of day during
which it was gathered, the attribute from which it was gathered,
the date on which it was gathered, and wherein the metadata of each
performance metric number is stored in the file system itself.
Description
BACKGROUND OF THE INVENTION
[0001] In the management of IT systems and other systems where
large amounts of performance data is generated, there is a need to
be able to gather, organize and store large amounts of performance
data and rapidly search it to evaluate management issues. For
example, server virtualization systems have many virtual servers
running simultaneously. Management of these virtual servers is
challenging since tools to gather, organize, store and analyze data
about them are not well adapted to the task.
[0002] One prior art method for remote monitoring of servers, be
they virtual servers or otherwise, is to establish a virtual
private network between the remote machine and the server to be
monitored. The remote machine to be used for monitoring can then
connect to the monitored server and observe performance data. The
advantage to this method is that no change to the monitored server
hardware or software is necessary. The disadvantage of this method
is the need for a reliable high bandwidth connection over which the
virtual private network sends its data. If the monitored server
runs software which generates rich graphics, the bandwidth
requirements go up. This can be a problem and expensive especially
where the monitored server is overseas in a data center in, for
example, India or China, and the monitoring computer is in the U.S.
or elsewhere far away from the server being monitored.
[0003] Another method of monitoring a remote server's performance
is to put an agent program on it which gathers performance data and
forwards the gathered data to the remote monitoring server. This
method also suffers from the need for a high bandwidth data link
between the monitored and monitoring servers. This high bandwidth
requirement means that the number of remote servers that can be
supported and monitored is a smaller number. Scalability is also an
issue.
[0004] Other non IT systems generate large amount of data that
needs to be gathered, organized, stored and searched in order to
evaluate various issues. For example, a bridge may have thousands
of stress and strain sensors attached to it which are generating
stress and strain readings constantly. Evaluation of these readings
by engineers is important to managing safety issues and in
designing new bridges or retrofitting existing bridges.
[0005] Once performance data has been gathered, if there is a huge
volume of it, analyzing it for patterns is a problem. Prior art
systems such as performance tools and event log tools use
relational databases (tables to store data that is matched by
common characteristics found in the dataset) to store the gathered
data. These are data warehousing techniques. SQL queries are used
to search the tables of time-series performance data in the
relational database.
[0006] Several limitations result from using relational databases
and SQL queries. First, there is a ripple that affects all the
other rows of existing data as new indexes are computed. Another
disadvantage is the amount of storage that is required to store
performance metric data gathered by the minute regarding multiple
attributes of one or more servers or other resources. Storing
performance data in a relational database engenders an overhead
cost not only in time but also money in both storing it and storing
it in an indexed way so that it can be searched since large
commercial databases can be required if the amount of data to be
stored is large.
[0007] Furthermore, SQL queries are efficient when joining rows
across tables using key columns from the tables. But SQL queries
are not good when the need is to check for patterns in values of
columns in a series of adjacent rows. This requires custom
programming in the form of "stored procedures" which extract the
desired information programmatically. This is burdensome, time
consuming and expensive to have to write a custom program each time
a search for a pattern is needed. As the pattern being searched for
becomes more complex, the complexity of the stored procedure
program also becomes more complex.
[0008] The other way of searching for a pattern requires joining
the table with itself M-1 number of times and using a complex join
clause. This becomes impractical as the number of joins exceeds 2
or 3.
[0009] As noted earlier, the problems compound as the amount of
performance data becomes large. This can happen when, for example,
receiving performance data every minute from a high number of
sensors or from a large number of agents monitoring different
performance characteristics of numerous monitored servers. The
dataset can also become very large when, for example, there is a
need to store several years of data. Large amounts of data require
expensive, complex, powerful commercial databases such as
Oracle.
[0010] There is at least one prior art method for doing analysis of
performance metric data that does not use databases. It is
popularized by the technology called Hadoop. In this prior art
method, the data is stored in file systems and manipulated. The
primary goal of Hadoop based algorithms is to partition the data
set so that the data values can be processed independent of each
other potentially on different machines thereby bring scalability
to the approach. Hadoop technique references are ambiguous about
the actual processes that are used to process the data.
[0011] Therefore, a need has arisen for an apparatus and method to
reduce the amount of performance data that is gathered so that more
sensors or servers can be remotely monitored with a data link of a
given bandwidth. There is also a need to organize and store the
data without using a relational database and to be able to search
the data for patterns without having to write stored procedure
programs, or do table joins and write complex join clauses.
BRIEF DESCRIPTION OF THE DRAWINGS
[0012] FIG. 1 is a block diagram of a typical server on which the
processes described herein for organizing, storing and searching
performance data can run.
[0013] FIG. 2 is an example of a directory structure storing one
day's performance data on a resource the performance of which is
being monitored remotely.
[0014] FIG. 3 is another example of a file system containing a
separate directory for storing performance metric data for three
different days for three different resources, each resource having
two groups of attributes.
[0015] FIG. 4 is a diagram of the directory structure of an example
of data collected by a probe.
[0016] FIG. 5 is a flowchart of the high level process the
monitoring server performs to receive probe data and stored it in
the directory structure for search and analysis.
[0017] FIG. 6 is a template for a regular expression used to
explain the syntax of a typical regular expression query.
[0018] FIG. 7 is a flowchart of one embodiment of the Query Request
Handler module.
[0019] FIG. 8, comprised of FIGS. 8A through 8C, is a flowchart of
the processing of the probe data importer.
[0020] FIG. 9, comprised of FIGS. 9A and 9B, is a diagram of the
modules in the system and a flowchart of the processing of the NRDB
Access manager module.
[0021] FIG. 10 is a block diagram of one embodiment of the overall
system including the major functional modules in the central server
called Megha, where the query request processing for analysis of
performance metric data occurs and where the NRDB stores the
performance metric data and configuration data.
[0022] FIG. 11 is a flowchart of the processing by one embodiment
of the Query Request Processor.
[0023] FIG. 12 is an example of time-series data.
[0024] FIG. 13 is a flowchart of the processing of evaluation.
[0025] FIG. 14 is an example of syntax of searching query.
[0026] FIG. 15 is a flowchart of the processing.
[0027] FIG. 16 is a flowchart of the processing.
[0028] FIG. 17 is a flowchart of the processing.
[0029] FIG. 18 is a flowchart of the processing.
[0030] FIG. 19 is a flowchart of the processing.
[0031] FIG. 20 is a flowchart of the processing.
[0032] FIG. 21 is a flowchart of the processing.
[0033] FIG. 22 is an example of pattern matching.
[0034] FIG. 23 is a flowchart of the processing.
[0035] FIG. 24 is an example of pattern matching.
[0036] FIG. 25 is a flowchart of the processing.
[0037] FIG. 26 is example of time slicing.
[0038] FIG. 27 is a flowchart of the processing.
DETAILED DESCRIPTION OF THE VARIOUS EMBODIMENTS
[0039] There is disclosed herein apparatus and processes for
infrastructure performance data analysis (and analysis of other
large amounts of performance data) which uses search techniques
instead of relational databases to store and organize data. Data is
stored in a special folder and directory structure with one
directory for every day's worth of performance data. This allows
data to be collected, processed and stored at a faster rate. All
the performance data collected from one or more resources in an IT
environment or one or more sensors in some other environment on the
day corresponding to the directory is stored in files within the
directory. There is a subdirectory for each resource where the
directory name is the signature for that resource. There is one
file for a group of attributes. Each attribute file has N sections,
one for each attribute defined to be in the group. Each section has
M values, where M values comprise the entire times series of values
for that attribute for the entire day corresponding to the
resource.
[0040] The result is that all the collected performance data is
stored as patterns; the patterns being data from many sources which
are sorted and stored in a time series in the special directory
structure described above; so all data from all sources for a
particular day is stored in one directory structure. This data
structure allows the data set to be searched with time as one axis
and each data element as the other axis.
[0041] Attribute values are stored either as band values or delta
values. Each value for an attribute for a particular reading on a
particular day is stored as Java UTF-8 encoded string with each
value encoded as a single Unicode character. In other words, the
numbers of each performance metric value are converted to letters
of a Java UTF-8 encoded string. This allows searching using
standard regular expressions the syntax of which is known and
comprises a form of formal language. The various elements of syntax
can be used to construct search queries which search through the
performance data for patterns. Regular expressions can only search
text and not numbers and that is why the performance metric
readings or values have their numbers converted to text before
storage.
[0042] The syntax of regular expression is rich with tools that
allow complex searches and pattern analysis simply by writing an
expression of the proper syntax thereby eliminating the time
consuming need to write a custom program or "stored procedure" in
SQL to do the same thing in searching the data of a relational
database.
[0043] Unicode is a computing industry standard for the consistent
encoding, representation and handling of text expressed in most of
the world's writing systems. It is a set of approximately 1 million
characters that span from hex 0 to hex 10FFFF. There are enough
unicode characters to devote a single one to every symbol in the
Japanese and Chinese languages and all the alphabets in the world
and all the numbers in which performance metrics are expressed.
Each performance metric value received from an agent is converted
to one of these unicode characters.
[0044] Searching the performance data with regular expressions
defining particular patterns of data from certain resources which
satisfy certain conditions expressed in the regular expressions is
analogous to searching large amounts of text for keywords and
reporting only those portions of the text which fit a certain
semantic usage or search string. This means the data pattern can be
identified by use of regular expression to define the search
criteria or a nondeterministic automaton as an equivalent thereto
by encoding in advance the measured data to a describable code,
i.e., a code searchable by use of a regular expression.
[0045] Therefore, the system according to the claimed technology
encodes the performance metric data using an encoding method such
as the encoding method of Unicode which results in the performance
metric data being converted into characters that can be searched
using regular expressions. Specific examples of the code that can
be used to encode the performance data include Unicode. The Unicode
standard defines 110,000 codes, which is an amount enough to encode
the numerical values of the performance metric data. The following
description is made on the assumption that the encoding is
performed by Unicode, but in a system according to the technology
claimed herein, any encoding method other than Unicode can be
carried out as long as the encoded performance data can be searched
using regular expressions.
[0046] The use of regular expressions allows complex patterns of
performance data to be searched without having to write complex,
custom programs called "stored procedures" which would be necessary
if a relational database was used to store the data and SQL was
used to search the database.
[0047] The system claimed herein allows users to draft their search
queries as regular expressions. The user must know the syntax of
regular expressions in order to do this unless the user wishes to
only use predefined searches which some embodiments of the claimed
technology provide for selection and execution by a user. A regular
expression provides a concise and flexible means for matching
strings of text, such as particular characters, words, or patterns
of characters.
[0048] A regular expression is written in a formal language that
can be interpreted by a regular expression processor, a program
that either serves as a parser generator or examines text and
identifies parts that match the provided specification. In the
preferred embodiment, the MARS query language detailed in Appendix
A attached here is the formal language from which searches
implemented using regular expression are expressed.
[0049] Storing the Unicode characters encoding the performance
metric data in the special directory structure described herein
eliminates the need for use of an expensive database system such as
Oracle even where very large amounts of data are collected and
stored.
[0050] The performance data is collected by agent programs which
are coupled to the sensors or are programmed on the IT resources
being monitored. These agent programs collect, compress and send
the performance data over the data link to the remote monitoring
server which collects it, converts it to Unicode and stores it in
the directory structure defined above. The remote monitoring server
also provides an interface for a user to compose regular expression
search queries and also provided "canned" searches which can be run
by a user, each canned search being a predefined regular expression
which the user may modify slightly to suit his or her purposes.
[0051] The process and apparatus for collecting, storing and
processing performance metric data differs from SQL Database
technology in at least two ways. First, the partition algorithm
stores performance data based upon time slices. By recording data
based on the time slice, it is possible to reduce the cost for
creation of Index relating to the time axis when the data is added,
which can suppress influences on the performance of the database.
This is important in order to speed up the data search in a
time-axis direction and to maintain the performance of data
addition.
[0052] Further, by recording data by creating a slice for each data
element in addition to the time slice, it is possible to perform
the data search in the time axis and the axis of each data element.
FIG. 2 and FIG. 3 are diagrams of a case where this method is
realized for the directory structure and the file on the file
system. The slice is realized by and implemented as the directory
structure, and the data is recorded as a file on the file system.
However, in addition thereto, by creating a key based on the date
and the identifier of the data element and associating the key with
the data, the same can be realized. The following description is
made on the assumption that the above-mentioned method is realized
by the directory structure and the file on the file system. In
addition, combining the data by previously slicing the data by the
time and the data element is important from the viewpoint of the
reduction in the capacity at the time of compressing and storing
the data. It is known that the data included in the same data
element tends to assume a similar value and is therefore high in
compressibility, i.e., the file size can be reduced by compaction
programs.
[0053] Storing the data in time slices in the special directory
structure, examples of which are shown in FIGS. 2 and 3, allows the
data to be searched with time as one axis and each data element as
the other axis. This is analogous to searching a large amount of
text for keywords and then reporting only those portions of text
that fit a certain semantic usage, such as matching the
keywords.
[0054] Further, by using this method, it is also possible to speed
up access to the stored data without affecting the entire database.
In a case where the pattern that is often used for a search is
known, such as a case where the data pattern that is often used for
the search is registered in advance or a case where there is a
tendency discernible from past searches or search results, it is
possible to speed up the search using such a data pattern by
performing the pattern matching before the storing. The processing
performed at that time is described with reference to FIG. 22. FIG.
22 illustrates time-series data 2202 and a set 2203 of query data
patterns that are often used for search. A time window 2201
represents a range of data that can be referred to when consecutive
processing is performed for data that flows without interruptions
such as the time-series data. In such a situation, it is possible
to speed up the referring to the data stored by performing the
processing as illustrated in FIG. 23.
[0055] First, after data is newly added to the time window, the
pattern matching is performed for data included in the time window
2201 by using the known query data pattern 2203 that is often used
for the search. After that, the data element at the head of the
time window is extracted and stored, and if matched in the
above-mentioned pattern matching at this time, it is predicted that
the data is often referred to, and information relating to the
stored location is stored together at the time of the storing.
[0056] The time required for this processing depends on the amount
of data included in the time window and the number of data patterns
used to perform the pattern matching. Further, the amount of data
included in the time window normally has a size enough to be loaded
in the memory. Therefore, there is a limitation on the referring to
the data, which does not affect the entire database.
[0057] In addition, in the above-mentioned example, the stored
location of the data that matches the data pattern that is often
used for the search is stored together, but in the case where the
data is compressed and stored or other such case, it is conceivable
that target data cannot be extracted without decompressing the
whole time series of data containing the string which matched even
when the stored place is known. In that case, this problem can be
avoided by, as illustrated in FIG. 24, compressing and storing only
the data matching the query data pattern 2203 that is often used
for the search in the manner of being ready to be extracted. In
FIG. 24, the partial syntaxes 2401 and 2402 are data stored by a
normal compression method, and the partial syntax 2403 is obtained
by separately compressing the data matching the query data pattern
2203 that is often used for the search. Further, FIG. 25
illustrates the flow of the processing performed at this time. The
matching processing is performed for the data included in the time
window in the same manner as in FIG. 23, but the different point is
that, when the data matches the pattern (Step 2501:YES), the data
is encoded by another encoding method and stored (Step 2502).
[0058] Note that, the size of the time slice used in this method is
assumed to be specified in a range that allows the data search to
be performed at a satisfactorily high speed, but there is a fear
that the amount of data within the time slice may become too large
due to a change in the sampling rate or the like. In such a case,
aside from changing the size of the time slice, by calculating
statistical values of the data included in a given time range and
then storing the statistical values together, it is possible to
speed up the referring to the data. FIG. 26 illustrates an example
thereof. In this case, it is assumed that such data is stored in
two time slices 2601 and 2602 which include data of a, b, c, d, and
e and data of d, e, f, and g, respectively, as the statistical
values. At this time, assuming that a search is performed for the
data string of fg, by referring to the statistical values in
advance, it turns out that there is no need to search the time
slice 2601, which can reduce the time required for the search.
[0059] FIG. 27 illustrates this flow. The statistical values are
calculated with respect to the data included in the time window,
and the results are stored along with the data. In this case, the
range of data used to store the statistical values can be freely
set in accordance with the degree of detail of the statistical
values necessary for the data search, the response time obtained at
the time of the data search, or the like, and the ranges of data
may overlap with each other or may include data that does not
include in any of the ranges.
[0060] The second difference between the claimed technology and the
prior art is that the method of analysis and search of the
performance data is based upon regular expressions which are used
to search Unicode encoded text where the performance metric numbers
have been converted to Unicode text characters. Regular expressions
have a fixed, predefined syntax and semantics (together considered
a grammar) and a variety of expressions can be formed using this
syntax and semantics to search the performance data for patterns
that meet criteria expressed in the regular expressions composed
for the custom search. Regular expressions can be derived for all
different kinds of search to limit the search to particular
resources, particular attributes of those resources, particular
days or particular time intervals during particular days, etc.
Great flexibility is provided without the complexity and labor of
having to write custom programs in the form of stored procedures to
find the right data and analyze it.
[0061] The processes described here to search and analyze
performance metric data are inspired by and somewhat similar to
XPATH technology. XPATH is a technique used to traverse XML
document data. XPATH-like techniques are used here to analyze
infrastructure performance metric data and changes to that data
over time. The processes described herein extend the XPATH notions
to the search and analysis of data organized and stored by time
slice which makes the search and analysis techniques taught herein
efficient and fast. Search and analysis of the performance data is
done using path-based techniques. A graph is created that
represents the data. The graph G is a representation of vertex and
edges (V,E). An edge connects two vertices and vertex has the
ability to evaluate an expression and then, based on the
expression, allow for a traversal through an appropriate edge.
[0062] FIG. 1 is a block diagram of a typical server on which the
processes described herein for organizing, storing and searching
performance data can run. Computer system 100 includes a bus 102 or
other communication mechanism for communicating information, and a
processor 104 coupled with bus 102 for processing information.
Computer system 100 also includes a main memory 106, such as a
random access memory (RAM) or other dynamic storage device, coupled
to bus 102 for storing information and instructions to be executed
by processor 104. Main memory 106 also may be used for storing
temporary variables or other intermediate information during
execution of instructions to be executed by processor 104. Computer
system 100 further usually includes a read only memory (ROM) 108 or
other static storage device coupled to bus 102 for storing static
information and instructions for processor 104. A storage device
110, such as a magnetic disk or optical disk, is provided and
coupled to bus 102 for storing information and instructions.
Usually the performance data is stored in special directory
structures on storage device 110.
[0063] Computer system 100 may be coupled via bus 102 to a display
112, such as a cathode ray tube (CRT) of flat screen, for
displaying information to a computer user who is analyzing the
performance data. An input device 114, including alphanumeric and
other keys, is coupled to bus 102 for communicating information and
command selections to processor 104. Another type of user input
device is cursor control 116, such as a mouse, a trackball, a
touchpad or cursor direction keys for communicating direction
information and command selections to processor 104 and for
controlling cursor movement on display 112. This input device
typically has two degrees of freedom in two axes, a first axis
(e.g., x) and a second axis (e.g., y), that allows the device to
specify positions in a plane.
[0064] The processes described herein to organize, store and search
performance data uses computer system 100 as its hardware platform,
but other computer configurations may also be used such as
distributed processing. According to one embodiment, the process to
receive, organize, store and search performance data is provided by
computer system 100 in response to processor 104 executing one or
more sequences of one or more instructions contained in main memory
106. Such instructions may be read into main memory 106 from
another computer-readable medium, such as storage device 110.
Execution of the sequences of instructions contained in main memory
106 causes processor 104 to perform the process steps described
herein. One or more processors in a multi-processing arrangement
may also be employed to execute the sequences of instructions
contained in main memory 106. In alternative embodiments,
hard-wired circuitry may be used in place of or in combination with
software instructions to implement the invention. Thus, embodiments
of the claimed technology are not limited to any specific
combination of hardware circuitry and software.
[0065] The term "computer-readable medium" as used herein refers to
any medium that participates in providing instructions to processor
104 for execution. Such a medium may take many forms, including but
not limited to, non-volatile media, volatile media, and
transmission media. Non-volatile media include, for example,
optical or magnetic disks, such as storage device 110.
[0066] Volatile media include dynamic memory, such as main memory
106. Transmission media include coaxial cables, copper wire and
fiber optics, including the wires that comprise bus 102.
Transmission media can also take the form of acoustic or light
waves, such as those generated during radio frequency (RF) and
infrared (IR) data communications. Common forms of
computer-readable media include, for example, a floppy disk, a
flexible disk, hard disk, magnetic tape, any other magnetic medium,
a CD-ROM, DVD, any other optical medium, punch cards, paper tape,
any other physical medium with patterns of holes, a RAM, a PROM,
and EPROM, a FLASH-EPROM, any other memory chip or cartridge, a
carrier wave as described hereinafter, or any other medium from
which a computer can read.
[0067] Various forms of computer readable media may be involved in
supplying one or more sequences of one or more instructions to
processor 104 for execution. For example, the instructions may
initially be borne on a magnetic disk of a remote computer. The
remote computer can load the instructions into its dynamic memory
and send the instructions over a telephone line using a modem. A
modem local to computer system 100 can receive the data on a
telephone line or broadband link and use an infrared transmitter to
convert the data to an infrared signal. An infrared detector
coupled to bus 102 can receive the data carried in the infrared
signal and place the data on bus 102. Bus 102 carries the data to
main memory 106, from which processor 104 retrieves and executes
the instructions. The instructions received by main memory 106 may
optionally be stored on storage device 110 either before or after
execution by processor 104.
[0068] Computer system 100 also includes a communication interface
118 coupled to bus 102. Communication interface 118 provides a
two-way data communication coupling to a network link 120 that is
connected to a local network 122. For example, communication
interface 118 may be an integrated services digital network (ISDN)
card or a modem to provide a data communication connection to a
corresponding type of broadbank link to the internet. As another
example, communication interface 118 may be a local area network
(LAN) card to provide a data communication connection to a
compatible LAN. Wireless links may also be implemented. In any such
implementation, communication interface 118 sends and receives
electrical, electromagnetic or optical signals that carry digital
data streams representing various types of information.
[0069] Network link 120 typically provides data communication
through one or more networks to other data devices. For example,
network link 120 may provide a connection through local network 122
to a host computer 124 or to data equipment operated by an Internet
Service Provider (ISP) 126. ISP 126 in turn provides data
communication services through the worldwide packet data
communication network, now commonly referred to as the "Internet"
128. Local network 122 and Internet 128 both use electrical,
electromagnetic or optical signals that carry digital data streams.
The signals through the various networks and the signals on network
link 120 and through communication interface 118, which carry the
digital data to and from computer system 100, are exemplary forms
of carrier waves transporting the information.
[0070] Computer system 100 can send messages and receive data,
including program code, through the network(s), network link 120,
and communication interface 118. In the Internet example, a server
130 which is having its performance data monitored might transmit
performance data via an agent program that collects it through
Internet 128, ISP 126, local network 122 and communication
interface 118 to computer system 100. The received performance data
is stored and can be searched by the processes described later
herein.
[0071] The system according to the claimed technology has on the
software and data side the following components which are executed
and stored on the hardware platform described above or similar.
Data Store Manager;
Query Request Handler;
Data Access Manager;
Probe Interface; and
[0072] Proprietary non-relational database referred to as the NRDB
and detailed in the Directory Structure heading below and
illustrated in FIGS. 2 and 3
Data Store Manager
[0073] This component receives data from probes in a well defined
format, and stores the received data in the NRDB. A probe is an
external software program which collects data on a periodic basis
from an external data source and writes data into a format which
can be processed by Data Store Manager. The Data Store Manager can
have any program structure so long as it can receive data in the
probe data format described elsewhere herein, decompress it and
store it in the NRDB in the directory structure and data format
described herein for the NRDB. In the preferred embodiment, it will
have a program structure which can perform the processing of the
flowchart of FIG. 5. It can run on any off the shelf computer
having sufficient speed, memory capacity and disk capacity to store
the performance data being collected.
Query Request Handler
[0074] This component accepts search queries from external
applications or users, and sends back the results. The query
language is a proprietary syntax for regular expressions which is
given below under the Query Definition Language Heading, and which
provides constructs for specifying search patterns to analyze data.
The Query Request Handler can have any program structure which can
receive query requests with regular expressions embedded therein
having the syntax described below, and parse those queries and
perform the processing of the flowchart of FIG. 7. It can run on
any off the shelf computer having sufficient speed, memory capacity
and disk capacity to store the performance data being
collected.
Data Access Manager
[0075] This component provides access to the data stored in the
claimed technology's proprietary non-relational database (NRDB).
This component internally employs standard caching techniques to
provide results faster. The Data Access Manager can have any
program structure which can access directory structures like those
of the NRDB of which FIGS. 3 and 4 are examples, and which supports
the Query Request Handler requests for data from the NRDB to
perform the processing of the flowchart of FIG. 7. It can run on
any off the shelf computer having sufficient speed, memory capacity
and disk capacity to store the performance data being
collected.
Probe Interface
NRDB
[0076] All the data in the claimed technology is stored in NRDB.
NRDB uses a normal file system consisting of files and folders. It
uses a special folder structure and special encoding of data files
to optimize the storage and access of data.
[0077] The entire software that implements the Data Store Manager,
the Search Handler, the Data Access Manager and the Probe
Interface, in the preferred embodiment is designed to run on
commodity hardware inside a Java virtual machine. Commodity
hardware is defined as regularly available Intel x86/64
architecture based computers. Standard Linux distribution such as
CentOS is used as the base operating system.
[0078] As an example of how the system works to collect performance
metric data and analyze it, suppose server 130 is a server which
has a couple of virtual machines running on it the performance of
which is to be monitored. The performance metric data for each
virtual machine is collected by an agent or probe process (not
shown) or, in some embodiments, a separate probe process for every
virtual machine. The performance data is gathered on a per day
basis to measure various performance metrics on server 130.
Performance data of the server 130 itself such as CPU cycle
utilization, hard disk access time, hard disk capacity, etc. may
also be gathered. There are usually several metrics that are
measured simultaneously, often on a per minute basis.
[0079] This performance metric data gathered by the agent process
is compressed and packetized and the packets are sent over the
internet 128 to ISP 126 to which a local area network 122 is
connected. The local area network is coupled via a network line 120
to the communications interface 118 of the monitoring server system
100.
Probe Data Format
[0080] The performance metric data for every element is collected
by a probe. A probe is a program running on the computer having the
element or attribute being monitored. The probe for each element
periodically or sporadically (usually a call is made every minute)
makes application programmatic interface calls to the operating
system of the computer or other machine to gather the performance
data on the element it is monitoring. The probes can be any agent
hardware and/or software combination that can collect the desired
performance metric data and put it into the data format described
below for probe data.
[0081] Probes don't have to be just for IT attributes. They can
also gather data for mechanical structures or automotive systems.
For example, engineers designing bridges may attach temperature and
strain sensors at various positions on the structures, each of
which is read by a probe program running on a computer which
periodically interrogates each sensor from time to time, takes its
reading and sends it elsewhere for storage and analysis. The probe
gathers all the sensor data, formats the data into the data
structure format described below, compresses the data structure and
packetizes the compressed data for transmission over any data path
to a system elsewhere for analysis. Likewise for cars, engines,
etc. The probe system is more or less like the modern day
equivalent of telemetry systems used on satellites and missiles
that feed performance data back to an earth station by a radio
telemetry link.
[0082] The performance metric data values gathered by the probes
are typically packetized for transmission over the internet. The
primary objective of the probe data format is to reduce the amount
of data which probe will produce so as to reduce bandwidth
requirements on the data link over which the probe data is sent.
This reduces the amount of storage required to store the data and
also makes the transmission to another location faster. The probe
programs do not do the conversion of the performance metric data to
unicode in the preferred embodiment, but in some alternative
embodiments, they could.
[0083] The probe collects all the attribute data for one day on all
the elements it is monitoring and creates a directory structure
such as the one shown in FIG. 4. The directory structure contains
files which store the time series of attribute values (performance
metric data) for every attribute for which the probe collected
data. The attribute values are numbers and are not converted by the
probe to unicode values. That happens at the monitoring server
end.
[0084] In FIG. 4, block 180 represents the top level directory,
block 182 represents a folder for all host type elements, block 184
represents a folder for all disk type elements being monitored.
Each of the folders 182 and 184 preferably contains a text file
which contains the attribute values obtained by the probe for every
element being monitored of the type symbolized by the subdirectory.
Each text file preferably contains all the performance metric
values for all the monitored elements in the same group with one
row containing the performance metric values measured for one of
the elements being monitored in that group. For example, the host
folder 182 may have a single text file A1.txt, but that file
preferably contains multiple rows, one for each host element being
monitored. For example, blocks 186 and 188 contain the performance
metric values for two particular hosts being monitored in the group
within A1.txt called H1 and H2. H1 and H2 in blocks 186 and 188
represent unique strings uniquely identifying the hosts for which
the performance metric data was collected. H1 has 1440 performance
metric measurements stored in the row symbolized by the V1, V2 . .
. V1440 values in a comma delimited list. For host H1, a
performance value was measured every minute. The same is true for
host H2. Blocks 190 and 192 preferably contain performance metric
values collected by the probe for two disks D1 and D2 in the group
of monitored elements "disk" represented by folder 184. These
performance metric values for disks D1 and D2 are preferably stored
in different sections or rows of a text file named A2.txt.
[0085] The whole collection of data files and subdirectories is
preferably zipped by the probe into one zip file which is a
compressed version of the data structure. By sending a compressed
version of the data, the bandwidth requirement on the data path
between the probe and the monitoring server(s) is greatly reduced.
When the zip file is unzipped, the data structure like that in FIG.
4 (or whatever the data structure is the number of elements and
attributes being monitored) results.
[0086] Any payload produced by the probe must conform to the
following structure:
The first file named ListOfFiles<YYYYMMDD_HHmmSS>_<base64
encoded text of encrypted value of (SiteName+"_"+ServerName+"_"
ArraySerialNumber)>_<ProbeType>.txt
[0087] Each line inside this file preferably has the name of a file
which is part of this payload [0088] If the file has configuration
or events data, the file must be named
Conf<YYYYMMDD_HHmmSS>_<base64 encoded text of encrypted
value of (SiteName+"_"+ServerName
+"_"+ArraySerialNumber)>_<ProbeType>.zip.enc [0089] If the
file has performance data, the file must be named
Perf<YYYYMMDD_HHmmSS>_<base64 encoded text of encrypted
value of
(SiteName+"_"+ServerName+"_"+ArraySerialNumber)>_<ProbeType>.zip-
.enc
Where:
[0089] [0090] SiteName--name of the site assigned for the probe
[0091] ServerName--name of the entity from which data is being
collected, it is the text filled in by the user during probe
configuration. [0092] ArraySerialNumber--Optional additional
information to further identify the entity. [0093] ProbeType--Type
of entity from which data is being collected--VMWare, SMIS, NetApp,
Amazon ECS, Bridge Sensors One or more .zip file are identified in
the list of files The configuration zip file preferably contains
one or more files which can be of two types:
[0094] Snapshot
[0095] Mini-snapshot
Snapshot
[0096] The snapshot type file preferably contains the entire
configuration about the data source to which the probe is
connected. The name of this file is: <Site
Name>_<DataSource>_snapshot_<YYYYMMDD>_<HHMMSS>_<-
Version>.txt, where: [0097] 2. <Site Name>: Identifier for
location (actual physical site) where the probe is situated [0098]
3. <Data Source>: Identifier for the data source (resource,
i.e., host, disk array, printer, etc.) from which the data is being
collected [0099] 4. <YYYYMMDD>_<HHMMSS>: The date and
time when the snapshot was made [0100] 5. <Version>: Version
of the file. The file format of snapshot is preferably as
follows:
TABLE-US-00001 [0100] %meta probe_id:<Identifier>
probe_type:<Probe Type> probe_site:<Site Name>
probe_server:<Server Name> probe_version:<Probe
Version> %meta { t:<YYYYMMDD_HHMMSS> {
R:<ResourceType>#<Resource Id>
O:{<ResourceType>#<Another_Resource_id>,}+? b:
<Begin Time YYYYMMDD_HHMMSS >? e:<End Time YYYYMMDD_HHMMSS
>? a:{<Attribute Id>=<Attribute Value>}+
r:{<Resource Type>#<Resource Id>,}+ $:{<Event Id>
<space><Event String>}+ }+ }+
Example
TABLE-US-00002 [0101] %meta probe_id:Cust_192.168.0.63
probe_type:VMWare probe_site:Cust1 probe_server:192.168.0.63
probe_version:10 %meta t:20110624_062248
R:dc#Cust_192.168.0.63_datacenter-2 a:name=MTNVIEW
R:ds#Cust_192.168.0.63_datastore-205 a:name=FAS960_home
a:capacity=51322806272 a:freeSpace=42685091840
a:uncommitted=17323200512 a:provisionedSpace=25960914944 a:type=NFS
a:URL=netfs://192.168.0.50//vol/vol0/home/ a:sioc=disabled
r:h#Cust1_192.168.0.63_host-171, R:ds#Cust1
_192.168.0.63_datastore-10 a:name=Storage1 $:AlarmSnmpCompleted
Alarm `Host error`- an SNMP trap for entity 192.168.0.48 was
sent
Updates
[0102] As configuration changes and configuration related events
occur, they preferably will be written to a mini snapshot file. The
name of this file will be:
<Site name>_<Data
Source>_minisnapshot_<YYYYMMDD>_<HHMMSS>_<version>.tx-
t
<YYYYMMDD>_<HHMMSS>:
[0103] The format of this file is preferably exactly same as the
snapshot file. The primary difference is that it will only have a
subset of the data of the snapshot type of file. The subset
captures the changes which have occurred in configuration data
since the last time a snapshot file was made.
Performance Data
[0104] The performance data is a zip file which preferably has the
following directory structure: [0105] 2.
<YYYYMMDD_HHMMSS>--This directory name is the start time of
the time series specified in this data set [0106] 3. <Resource
Type>--One directory for each resource type [0107] 4.
<Attribute Id>.txt--One file for each performance metric Each
<Attribute Id>.txt has one or more lines where each line has
the following format:
<Resource Signature>`,` {Value} `,` {`,`<Value>}+
[0108] The value list is a time ordered series of values for that
performance metric for the resource specified at the beginning of
the time. If the metric value does not exist for a particular point
in time, then a blank or empty value is allowed.
NRDB File System Structure
[0109] The performance metric data is preferably stored in a file
system structure as defined below. One directory is preferably
created for each day in the format YYYYMMDD. All performance data
for all the resources in the data model for a particular day are
preferably stored in this directory. Under this directory, there is
preferably a directory for each resource where the directory name
is preferably the signature of that resource. Under this directory,
there is preferably one file for a group of attributes. The
directory will preferably look something like this:
TABLE-US-00003 <YYYYMMDD> - One Folder for each day
<Resource Type> <AttributeGroupId>.perf
[0110] 5. <YYYYMMDD_HHMMSS>--This directory name preferably
contains the start time of the time series specified in this data
set [0111] 6. <Resource Type>--preferably one directory for
each resource type [0112] <Attribute Id>.txt--preferably one
file for each performance metric <AttributeGroupld>.perf file
preferably stores processed values for each sample in a compressed
format. This format is now described in detail. The file is
preferably divided into "n" number of sections. Where "n` is the
attributes which are defined to be in the same group. Each section
preferably will hold "m" number of values--the entire time series
values of that day for that resource's attribute. So, for example,
if the probe sampling interval is 1 minute then there will be 1440
(1440 minutes in a day) values. Each <Attribute Id>.txt
preferably has one or more lines where each line has the following
format:
<Resource Signature>`,` {Value} `,` {`,`<Value>}+
[0113] The value list is preferably a time ordered series of values
for that performance metric for the resource specified at the
beginning of the time. If the metric value does not exist for a
particular point in time, then a blank or empty value is
allowed.
[0114] Currently, corresponding to each raw value of a performance
metric attribute received from the probe, preferably two types of
processed value are stored:
[0115] Band Value [0116] An attribute can define the "fidelity"
with which it will store the raw value. This is called in Band
Factor. Band factor is preferably an integer with a minimum value
preferably of 1 and maximum of preferably any positive integer
value. With a band factor of 1, there is no loss of fidelity. The
processed value is same as raw value. With a band factor 10, the
processed value will preferably be 1/10.sup.th of the raw value
rounded to the nearest integer.
[0117] Delta Value [0118] It is preferably the change in percentage
from band value at time t-1 and band value at time t.
[0119] Each set of 1440 values of a performance metric attribute
(assuming one value is measured every minute) are stored preferably
as a Java UTF-8 encoded String. Each performance metric attribute
value is preferably encoded as a single Unicode character in the
String.
[0120] FIG. 2 is an example of a directory structure storing one
day's performance data on a resource the performance of which is
being monitored remotely. The processor 104 in FIG. 1 is programmed
by instructions stored in main memory 106, according to one
embodiment of the invention, to create a special directory
structure with preferably one directory for each day's worth of
data, and preferably one subdirectory for each resource for which
performance metric data is being received. In FIG. 2, block 150
represents the directory created for storing the performance metric
data collected on Aug. 14, 2011. The subdirectory represented by
block 152 represents the subdirectory where performance data for
the resource E1 is to be stored. Suppose in this example, that
resource E1 is the server 130 in FIG. 1.
[0121] Each subdirectory preferably has the directory name in its
signature. In this case, subdirectory 152 has 20110814 in its
directory name which is preferably the name of the directory of
which it is a part.
[0122] Each subdirectory preferably contains one attribute file for
each group of attributes that are being measured by the performance
metric data that stores performance metric values. Each attribute
file preferably has N sections, one section for each attribute
defined to be in the group for which the file was created. Each
section preferably holds M performance metric values for the
particular attribute whose values are recorded in that section.
That section's data preferably comprises the entire time series of
values for the attribute to which the section is devoted.
[0123] In the example of FIG. 2, there are only two groups of
attributes in subdirectory 152 so there are only two files 154 and
156. Suppose each of these files represents one of the virtual
machines running on server 130. Each file is a time slice of
performance metric data values that records the entire day's worth
of a metric in the section of that file devoted to storing values
for that performance metric. Typically, if a metric has a measured
value every minute, the section of the file devoted to that metric
will have 140 comma delimited values for that metric encoded as a
Java UTF-8 encoded string. UTF-8 is a multibyte character encoding
for unicode. UTF-8 can represent every character in the unicode
character set. Each of the 1,112,064 code points in the unicode
character set is encoded in a UTF-8 string comprised of one to four
8-bit bytes termed octets. The earlier characters in the unicode
character set are encoded using fewer bytes leading to greater
efficiency. The first 128 unicode character set coincide with the
128 ASCII characters.
[0124] The system of the claimed technology has a mapping table
that maps performance metric values into unicode characters and
then encodes them with UTF-8. Since unicode only supports positive
values, the unicode range is split and a first range of unicode
values is mapped to positive performance metric values and a second
range of unicode values is mapped to negative performance metric
values.
[0125] However, this mapping method is merely an example. By
examining the occurrence frequency of the values and assigning the
code having a small size to the value having a high occurrence
frequency, it is possible to reduce the storage size necessary to
store the data. Further, if the distribution of the values differs
depending on a kind of performance metric value and a time slot
(time slice) for the measurement of data, by changing the mapping
depending on the kind of performance metric value and the time slot
for the measurement of data, it is possible to realize further
reduction in the storage size necessary to store the data. In
addition, a mapping method which assigns (quantizes) a plurality of
values to one code at the time of the mapping is conceivable, but
by adjusting a range of quantization in that case in accordance
with a degree of detail (resolution) obtained when the data is
used, it is possible to reduce the number of necessary codes. As a
result, it is possible to use codes having a short code length,
which can realize a reduction in the necessary storage size.
[0126] Each performance metric value from a measurement is
preferably encoded as a single unicode character in the hexadecimal
number system (hex).
[0127] Each new day's worth of data from all resources and all
probes is preferably stored in a new directory structure. The names
of the directories, subdirectories and files preferably include
information about the day during which the data was gathered, the
resources from which it was gathered and the particular group of
attributes whose performance metric data is stored in the various
sections of the file.
[0128] In the example of FIG. 2, the directory structure 150 has
files 154 and 156 for one day of metric data gathered every minute
for two different metrics from the same resource, represented by
subdirectory 152. In other words, there is only one resource being
monitored. Also, for the example of FIG. 2, there is only one
attribute in each group of attributes and only two attributes in
total have performance metric data gathered. The performance metric
data is gathered on Aug. 14, 2011 so the directory 150 created to
store that day's metric data is named 20110814. There is only one
resource being monitored called E1 so there is created a
subdirectory 152 called 20110814_E1. That subdirectory contains two
files. The first file 154 is named E1/G1, and it stores the metric
values for metric M1 in group 1 (which has only one section because
there is only one metric M1 in the group E1/G1). The values of
metric M1 are gathered every minute and are symbolized as values V1
through V1440 which are stored as a comma delimited list. The value
V1 is the value of metric M1 taken at time 00:01:01 on 8/14/2011,
i.e., the first minute of 8/14/2011. The value V2 is the value of
metric M1 taken at time 00:02:01 on 8/14/2011, the second minute of
8/14/2011. The value V1440 is the value of metric M1 taken at time
23:59:01 which is the last minute of 8/14/2011. Therefore, the
position of any particular value on the comma delimited list
denotes the time at which the value was captured on 8/14/2011.
[0129] The second file 156 in the resource E1 subdirectory is named
E1/G2 and it stores values for a metric M2 in group 2 (which also
only has one metric in the group so there is only one section in
the file). It has not been shown in detail since it has the same
structure as the file E1/G1.
[0130] The values stored in each position of the file are
preferably Unicode encoded meaning the numeric value of the
metric's value has been mapped to a text character or string of
characters in the encoding process. Other encoding processes which
convert numeric values to text values could be used in other
embodiments.
[0131] This allows these values to be searched using regular
expressions which are a form of formal language (used in the sense
computer scientists use the term "formal language") which has
predefined rules of syntax and semantics (together called its
grammar). The elements from which regular expressions can be formed
are known and each element has its own known syntax for how it is
structured and has its own unique and known semantics defining what
it means. Persons wishing to analyze the performance metric data in
any way, can compose a regular expression using the available
elements for composing a regular expression and their syntax and
semantics. Any regular expression syntax can be used to carry out
the teachings of the invention, but the claimed technology uses a
proprietary syntax which is optimized for the application and is
disclosed elsewhere herein.
[0132] FIG. 3 is another example of a file system containing a
separate directory for storing performance metric data for three
different days for three different resources, each resource having
two groups of attributes. The file system storing metric data is
represented by block 158. Three days of performance data are stored
in directories 160, 162 and 164, respectively. Each of these
directories has three subdirectories named R1, R2 and R3, each of
which is a folder which contains actual files of text data encoding
performance metric values that have been measured and transmitted
by the agents. Blocks 166 and 168 represent comma delimited text
files named GRP1.TXT and GRP2.TXT storing the performance metric
data gathered on Jul. 27, 2011 for resource 1 for group 1 and group
2 attributes, respectively.
[0133] The reason for grouping different attributes performance
values in the same file is for speed of loading and analysis.
Typically, an analysis of a resource will involve looking at
patterns or values or value changes of several different attributes
over a particular interval. If the attributes involved in the
analysis are all grouped in the same group, they will preferably be
stored in the same file. In this way, all the data needed to do the
analysis can preferably be loaded into memory for analysis simply
by reading the appropriate file containing the attribute group for
the resource under analysis from the directory structure
corresponding to the day of interest. That file is loaded into
memory by a standard file access call to the operating system, and
the regular expression search or searches can be performed on the
data. This is faster than having to load several different files or
having to do SQL queries to a database which would require a larger
number of reads.
[0134] FIG. 5 is a high level flowchart of the process the
monitoring server preferably performs to receive the zip files of
performance metric data from a probe, recover the data and store
it. Block 200 represents the process of receiving the zip file of
performance metric data from the probe. Block 202 represents the
process of decompressing the zip file to recover the data structure
such as that shown in FIG. 4. Block 204 represents the process of
converting the numerical performance metric values stored in the
text files preferably to unicode characters using a mapping table
the server uses for such purposes. Block 206 represents the process
of storing the unicode data structure derived in step 204
preferably in the appropriate parts of the NRDB data structure.
Usually this just entails storing the entire directory and all its
files on disk since the data structure is already structured as one
directory for the particular day on which the data was collected
preferably with individual text files of metric data for each
element being monitored in subdirectories for the type of element
each text file represents.
Example of how a Regular Expression can be Used to Analyze the
Metric Performance Data
[0135] Suppose an analyst wanted to know if CPU utilization was
between 90% and 100% for at least 5 minutes or more. The regular
expression syntax to make a search and analysis of the performance
metric data for CPU utilization would be in generic syntax:
[U90-U100]{5,} -100 -200
[0136] To convert this regular syntax to take into account the
unicode encoding of the CPU utilization metric values, suppose a
CPU utilization metric value representing 90% utilization is mapped
to unicode hex character a, 92.5% CPU utilization is mapped to
unicode hex character b, 95% to hex character c, 97.5% to hex
character d, and 100% to hex character e. If CPU utilization metric
values are measured every minute, then a regular expression to
determine if the CPU utilization was between 90% and 100% for at
least 5 minutes would be:
[a-e]{5}[g] which means if five consecutive values in the file
storing CPU utilization values for the CPU in question on the day
in question were any combination of hex characters a through e,
then the expression evaluates to true. This means that every time
on that particular day the CPU utilization metric values had five
consecutive values which were any combination of hex a through hex
e, then for each of those intervals, the CPU utilization was
between 90% and 100%. This may mean the CPU is maxing out and
another CPU should be added.
[0137] In the preferred embodiment of the claimed technology, the
user must know the syntax of regular expressions in order to
compose his or her query. In alternative embodiments, a user
interface is provided which allows the user to think in the problem
space and compose his queries in plain English, and the system
converts that query into the proper syntax for a regular expression
which will perform that query and analysis. In some embodiments,
the software portion of the system of the claimed technology
presents a user interface which has a set of predefined searches
which the user can use to do various forms of analysis. Each
predefined search, when selected causes a regular expression to be
generated and used to search the performance metric data and return
the results. In some embodiments, these predefined searches are
templates which have variables that can be set by the user. For
example, there may be a predefined search to determine if CPU
utilization is between x % and y % for more than z minutes where x,
y and z are variables that the user can set before the search is
run.
[0138] To run a search/query, in the preferred embodiment, the
software of the system displays a query expression box and two time
range boxes, one for a start time and one for an end time. These
start and end time boxes are calendars in the preferred embodiment,
and the user simply picks the first day for which data is to be
examined and picks a second day in the end time calendar which is
the last day of data to be examined. He then types his query into
the query expression box in the syntax of the regular expression
and hits return. The software then automatically accesses the
appropriate directory structures for the day or days specified by
the user, accesses the appropriate files that contain the
performance metric attribute values as specified in the query
expression, reads those attribute values into memory and examines
the data using the logic specified in the query expression.
[0139] FIG. 6 is a template for a regular expression used to
explain the syntax of a typical regular expression query. The h at
the beginning of the regular expression indicates that this
particular query is designed to search host performance metric
data. If the query was about disks or something else, something
indicative of the type of resource in question would be in the
place of the h.
[0140] The large left bracket indicates the beginning of the actual
query expression. The @ symbol at the beginning of the query
expression is a keyword. The "CPU usage" term is the name of the
attribute data to be searched and it is this attribute name which
causes the software to look up the correct file name which contains
the performance metric data for CPU usage. The "rx" term indicates
that what follows is a regular expression, and the "b" term
indicates that the type of search is for band data as opposed to
delta data. The [U90-U100] {5} is a regular expression that
indicates the actual criteria to be used in performing the band
data search, i.e., it defines which performance metric data satisfy
the query and which do not. The regular expression could also be a
pointer to another regular expression stored in a file. The pointer
would contain a unique ID for the regular expression to be
used.
[0141] The band values are computed or mapped values for internal
representation of numbers which are greater than the highest number
which can be unicoded (around 1,000,000). For example, if a data
transfer rate is 20 million bits per second and the metric is
20,000,000, a band value will be computed for that metric using a
reduction factor of, for example 10 million so as to reduce the 20
million number to the number 2 before it is unicoded. Any reduction
factor that brings the range of a performance metric which is a
high number down into the unicode range may be used for internal
representation purposes. The searches are then done on the computed
band values and not the actual performance metric numbers.
[0142] Delta values are useful for analyzing performance metric
data that spikes. A delta value records how much a value has
changed since the previous time it was measured.
[0143] The system, in the preferred embodiment, calculates and
stores both a band value and a delta value for some or all
performance metrics.
Query Definition Language
Objectives
[0144] 7. Be able to traverse from a set of resources to another
set of related resources and so on [0145] 8. At each stage of
traversal apply certain filtering criteria: [0146] 9. Configuration
attributes: Matching certain value, change in value [0147] 10.
Relations: Addition or deletion of a relation [0148] 11.
Performance metrics: Matching certain patterns
[0149] Before a detailed description of the query definition
language, a description is made of a basic operation of this search
method by taking an example. FIG. 12 illustrates a data example
used for the description of the example. First, it is assumed that
there are two kinds of resource type, vm and host, as an
environment for measuring the time-series data. It is assumed that
there are vm1 and vm2 as resources having a resource type
classified as vm and that there is host 1 as a resource having a
resource type classified as host. It is assumed that the resource
having the resource type of vm has two attributes of readlatency
and readiop and that the resource having the resource type of host
has the attribute of readlatency.
[0150] At this time, the time-series data is stored in a state in
which an ID 1204 that can identify the time slice and ID 1205 and
1206 are used for identifying the resource which are assigned
thereto. In FIG. 12, for easy understanding, a hierarchical
structure formed of time slice-resource identifiers and a plurality
of pieces of time-series data are expressed in the form of one
table 1207, but those pieces of time-series data do not need to
form one table or do not need to have a time stamp 1208 assigned to
each row thereof. Further, in the example of FIG. 12, for easy
understanding, measurement data is shown in the form of numerical
values, but is actually stored in an encoded form as alphanumeric
characters in order to make it easier to perform the pattern
matching of the time-series data. Further, it is assumed that a
relationship is defined in advance between the data having resource
IDs of vm1 and vm2 and the data having a resource ID of h1.
[0151] It is now assumed that a syntax as illustrated in FIG. 14 is
given as a search query. The search query of FIG. 14 is one large
regular expression which is comprised of three smaller regular
expressions. Then, the system disclosed in the present
specification interprets the syntax of the single large regular
expression based on a predefined syntax rule which, using the rules
of grammar of the formal language which define the syntax that must
be used in all searches encoded in regular expression written using
that formal language, interprets the syntax of each of the three
smaller regular expressions in a predefined order, each search
restricted to searching through only the data found by the previous
search. In this case, the syntax of the large regular expression is
divided into three smaller regular expressions having the partial
syntaxes 1401, 1402, and 1403, which are evaluated in order from
the beginning. So in this case, the "predefined syntax rule" is:
evaluate the three smaller regular expressions in the order in
which they appear in the larger regular expression, each search
restricted to searching through the matching data found in the
previous search except that the first search encoded in regular
expression 1401 is performed on the time sequence data in the time
window defined by the user who initiated the searches." The
predefined syntax rule can also define a tree-structured
relationship between the searches carried out by the smaller
regular expressions which comprise a larger regular expression. In
such a case, the first search is called the root search and then
smaller regular expressions that form the branches or leafs of the
tree are simultaneously performed on the data found by the root
search. So the predefined syntax rule in such a case would be:
"perform the root search first, and then perform the searches of
the leafs of the tree on the data found by the root search and so
on in case other searches branch off from a branch of the tree.
Another predefined syntax rule is perform a search using as the
search criteria a notation denoting a condition that matches a
plurality of encoded values wherein a range search for a value is
performable, said notation being like a wild card which can match
to anything or anything within a range of values.
[0152] FIG. 13 illustrates a flow of a series of searches where
each search is performed on the data found by the preceding search.
First, a query for readlatency performance metrics in a
predetermined range for resources having resource IDs vm is done,
the query having the regular expression partial syntax shown at
1401 in FIG. 14 which is obtained by the dividing, said query being
represented by step 1301 in the flowchart of FIG. 13. At this time,
in this query, first, the data must have the resource type of vm,
which in this case, translates to the resource IDs of vm1 and vm2.
In this query, the attribute of readlatency is read, and the query
is looking for performance metric values between 20 and 1000 as
given by the syntax of the query shown at 1401, said performance
metric values recorded in the table of FIG. 12 each being encoded
by an encoding method at a time of data storage. In other words,
the query established by the regular expression segment 1401 in
FIG. 14 is looking for all performance metric values for
readlatency for vm type resources, which, in the case of the
timeslice data of FIG. 12 means resources having IDs of vm1 and
vm2, that have values between 20 and 1000. After that, a search
formula based on the regular expression is created from the encoded
values of 20 and 1000. At this time, if a search language supports
a range search, the search formula created may be written in
accordance therewith. If the search language does not support the
range search, after the values between 20 and 1000 are each
encoded, the matching may be performed for a character set obtained
by encoding the values included in the range from 20 to 1000. Then,
in accordance with the search formula based on the regular
expression, the pattern matching is performed for the thus-read
data to acquire matched data. In this case, the data having the
resource IDs of vm1 and vm2, the range of time, and the resource
IDs of vm1 and vm2 and having readlatency performance metrics are
acquired from among data included in the table of FIG. 12, the
acquired data being in an area surrounded by a broken line 1203 in
the table of FIG. 12. In short, the searches encoded in the regular
expressions of 1401, 1402 and 1403 are performed in that order,
each search restricted to searching through the matching data found
in the previous search.
[0153] Next, a query for host performance metrics for attribute
readlatency is done using a regular expression having the partial
syntax 1402 obtained by the dividing, this query being represented
by Step 1302 in FIG. 13. At this time, the search range of the data
is limited to the data having the range of time obtained in Step
1301 among the data having the resource type of h, in this case,
the resource ID of h1, and having the attribute of readlatency.
Processing details of the search are obtained by evaluating the
partial syntax 1402 with respect to this data, and these process
details are applied in step 1302. As a result, the data having the
resource ID of host 1, the range of time, and the resource ID of h1
are acquired from among data included in an area surrounded by a
broken line 1202.
[0154] After that, a third, narrowing search is done using a
regular expression having the partial syntax 1403, obtained by the
dividing, as represented by Step 1303. At this time, the search
range of the data is limited to the data having the range of time
obtained in Step 1302 and the search is done on the data having the
resource type of vm, which in this case, translates to the resource
IDs of vm1 and vm2, the search being done on data having the
attribute of readiop. Processing details for the search are
obtained by evaluating the partial syntax 1403 with respect to this
data are applied in step 1303. As a result, the data having the
resource ID of vm2, the range of time, and the resource ID of vm2
are acquired from among data included in an area surrounded by a
broken line 1203.
[0155] Note that, in this case, the input syntax is evaluated at a
time of carrying out search processing, but such an embodiment can
be applied that the syntax is converted in advance in a case of
repeatedly using the same syntax, in other words, automatic
reporting. In some embodiments, if the search just described is
carried out often, it can be included as a search template which
can be customized by the user or a "canned search" that can be
invoked from a user interface that allows the user to work in the
problem space isolating the user from the need to know the syntax
of the regular expression.
[0156] In the present specification, such a step of narrowing down
data is expressed as "traverse".
Basic Syntax Building Blocks that May be Used to Build a Query
XPath style data processing/filtering and this processing will be
applied to various search queries.
TABLE-US-00004 <Resource Type>/<*Related resource
type>[= <conf attrId> rx <regex> ORIAND
...][~<conf attr id>,,,][@<perf attr id> <rx
bld>IrxId <regex or regex pattern id>][$<event
id,,,][+l-<related resource type]/{Related resource
type/...}{Related resource type/...}
Relation Traversal:
[0157] <resource type>/<related resource type>/ . . .
Ex: v/h/d The above expression will result the following path:
v->h->d
Multiple Traversal Paths:
[0158] <resource type>/{related resource type>/ . . .
}{another related type>/ . . . }
[0159] The flow of the processing using this multiple traversal
paths syntax rule is described with reference to FIG. 15. In FIG.
15, v, h1, and h2 each represent a query represented by a partial
syntax. First, the query represented by the partial syntax of v is
evaluated (Step 1501). After that, the queries represented by the
partial syntaxes of h1 and h2 are each applied to the evaluation
result of (found data) from the query represented by partial syntax
v and evaluated with respect thereto (Steps 1502 and 1503). After
that, the obtained results are merged (Step 1504).
Example 2
[0160] Ex: v/{h/n}{r/d}
[0161] A description is made of the flow of the processing
performed in a case of the above-mentioned sample where v, h, n, r
and d all represent queries expressed as regular expressions which
are part of the syntax of the overall query having syntax
v/{h/n}{r/d}. First, v is evaluated, and then h and r are each
applied to the evaluation result of v. After that, n and d are
respectively applied to the results of applying h and r. Finally,
as the flow of the processing, the following two paths are executed
to obtain a processing result by merging the two results. [0162]
v/h/n (v->h->n) [0163] v/r/d (v->r->d)
Another Example
[0164] Note: There is no limit on number of queries or number of
sub paths and any number of levels of nested paths are supported as
shown in the following sample:
Ex: v/{h/{r1/d1}{n}}{r2/d2}
[0165] A description is made of the flow of the processing
performed in a case of the above-mentioned sample. First, v is
evaluated, and then h and r2 are each applied to the evaluation
result of v. To the result of applying r2, d2 is applied. On the
other hand, r1 and n are each applied to the result of applying h,
and d1 is further applied to the result of applying r1. Finally,
the following three paths are executed to obtain a processing
result by merging the three results. [0166] v/h/r1/d1 [0167] v/h/n
[0168] v/r2/d2
Look for Changes in Configuration:
[0169] <resource type>[.about.<attr id>, <attr
id> , , ,] Ex: v/h[.about.attr1,attr2]/n It takes all resources
of type `v`, finds the related resources of type `h` which have
configuration attributes attr1 and attr2 which have changes in the
given time window. Then it finds resources of type `n` the
resulting resources of type `h`.
Find Patterns in Performance Data:
TABLE-US-00005 [0170]<resource type>[@*<attr id> <rx
bld> IrxId <expression or id>][@....] <resource
type>[@*#tw1 #<attr Id> rx bld <expr....>]/<r
type>[@{circumflex over ( )}tw1 {circumflex over ( )} <attr
id> <rx bl d> ....] <resource type>[@*#tw1 #<attr
Id> rx bld <expr....>]/<r type>[@#tw2#{circumflex
over ( )}tw1 {circumflex over ( )} <attr id> <rx bld>
....] Where *: ignores the resulted data 1) can be used to derive
time windows for subsequent use 2) can be used to build logical
pattern b: for banded datad: for delta values
Special note: Any numeric value in actual regex
(exclusion=>quantifiers) should be prefixed with "U" e.g
[40-90]{5} will become [U40-U90]{5}. Here numbers within the
character class have been modified but not the quantifier i.e
{5}.
Examples of Regular Expression Queries of Various Types
EXAMPLES
[0171] v[@attr1 rx b U90+]/h
[0172] A query having this syntax finds all the virtual machines
which have performance data of metric attr1 value that equals or
exceeds 90 in the given time window. Then it finds the respective
hosts. It also returns the matched performance data
[0173] v[@attr1 rxId rxp1]/h
[0174] This query is similar to the example just preceding, but it
specifies the regex pattern id which will be defined in a separate
file.
[0175] Reuse of Processing Result:
[0176] rt1[@#tw# attr1 attrCond1]/rt2[@{circumflex over (
)}tw{circumflex over ( )} attr2 attrCond2]
[0177] FIG. 16 illustrates the flow of the processing performed at
this time. First, the evaluation result of the query represented by
the partial syntax of rt1[attr1 attrCond1] is bound to tw. After
that, the query represented by the partial syntax of rt2[attr2
attrCond2] is evaluated, and the range of data processed at that
time is a time range bound to tw.
Example
[0178] v[@#tw1# attr1 rx b U90+]/h[@{circumflex over (
)}tw1{circumflex over ( )} attr2 rx b U80+]
[0179] The first metric has defined a time span Id (tw1) which can
be referred by any other metric in the subsequent path. If metric
attr1 has generated any matched data, the respective time windows
will be assigned the id "tw1" and the same time windows will be
used on metric attr2 [attr12 or attr2?]. Note that if the connected
host has more narrow time windows than the resulted tw1, the common
slots will be used on metric attr2.
Event Filter:
[0180] Syntax: [$*t:<regex pattern>,d:<regex
pattern>]
Where
[0181] *: ignores the resulted data (won't produce any output but
can be used to build logical patterns)t: will search against the
type of the eventd: will search against the description of the
event The following are valid:
TABLE-US-00006 C) [$t:rmAdded] // type check D) [$d:error] //
description check E) [$t:rmAdded,d:error] // logical OR F)
[$*t:rmAdded] // type check and ignore the result G) [$*d:error] //
description check and ignore the result H) [$*t:rmAdded,d:error] //
logical OR and ignore the result
Resource Addition/Deletion:
[0182] <resource type>[+<related resource types added>
, , ,][-<related resource types removed> , , ,] Ex:
v[+h,d,n][-h,d] The above expression will return resources of type
`v` on which relation of type `h`, `d`, `n` has added or relation
of type `h`, `d` has been removed. How to exclude the data of a
matched relation: <resource type>/*<related
resource>/<sub resource> Ex: v/*h/d The above express will
return resources of type `v` and the related resources of type `d`
directly. But, it will skip the data of the matched resources of
type `h` in the output. Note: One can mix any of the above
combinations. One can specify configuration changes, performance
data filters, events list, multiple paths, etc. in the same
query.
Logical AND Operator
[0183] Logical AND operations are supported at path level and
filter level. Applicable conditions for processing can be narrowed
down by using logical AND operators.
At Path Level:
Syntax: P1/[&]P2/[&]P3/P4 . . . .
[0184] A description is made of the flow of the processing
performed in a case of using the logical AND operator at the path
level. First, FIG. 17 illustrates a basic flow of a process using
the AND operator. Here, in FIGS. 17, 18, and 19, p1, p2, and p3
each represent a query partial syntax. By describing p1/&p2,
the processing is executed as illustrated in FIG. 17, and p1 is
evaluated only when a condition specified by the query having
partial syntax of p2 is established.
Example 1
[0185] p1/&p2 p1 &&p2 Note: p1 qualifies only if p2
qualifies
[0186] Further, this operator can be used a plurality of times.
FIG. 18 illustrates the flow of the processing performed in that
case.
Example 2
[0187] p1/&p2/&p3 p1 &&p2&&p3 Note: p2 is
dependent on p3 and p1 is dependent on p2
[0188] Only the partial syntax immediately before the logical AND
operator is determined as to whether or not to be executed by the
operator, and the partial syntax before the above-mentioned partial
syntax is evaluated without conditions. FIG. 19 illustrates the
flow of the processing performed in that case.
Example 3
[0189] p1/p2/&p3 p1,p2&&p3 Note: p1 can qualify
irrespective of p2 status but p2 can qualify only if p3
qualifies
Example 4
[0190] p1/&p2/p3/&p4 p1 &&p2,p3&&p4 Note:
p2 can qualify irrespective of p3 status.
At Filter Level:
[0191] Syntax: P1 [filter1][&][filter 2][&][filter
3]/P2[filter 1][&][filter 2]
[0192] A description is made of the flow of the processing
performed in a case of using the logical AND operator at the filter
level. First, FIG. 20 illustrates a basic flow. Here, in FIGS. 20
and 21, p1 is the partial syntax representing the processing, and
f1, f2, and f3 represent filter processing for filtering data
supplied to the processing of p1. By describing p1 [f1]&[f2],
the processing is executed as illustrated in FIG. 20, and the data
that satisfies both conditions of f1 and f2 can be specified as the
data to which p1 is to be applied.
Example 1
[0193] p1[=1001 rx Demo3]&[@2001 rx b U10+] Note: P1 qualifies
only if both the filters find matches.
[0194] A case where the partial syntaxes that specify the filter
processing are simply arrayed is recognized as OR, which is applied
prior to the AND operation. FIG. 21 illustrates the flow of the
processing performed in that case.
Example 2
[0195] p[f1][f2]&[f3] (f1.parallel.f2)&&f3
Example 3
[0196] p[f1]&[f2][&f3] f1&&f2&&f3
Example 4
[0197] p[f1][f2][f3] f1.parallel.f2.parallel.f3
Example 5
[0198] p[f1]&[f2][f3] f1&&(f2.parallel.f3)
Note: if f1 fails, it exits (no processing of f2 or f3). Short
circuit execution on Logical 20 AND failure.
Example 6
[0199] p[f1]&&&&&[f2] f1 &&f2 Note:
multiple &s will be collapsed into one
Example 7
[0200] p[f1][f2]& f1 II f2 Note: trailing & will be
ignored
Others
[0201] Regular expression patterns can include brackets, but only
with matching pairs. When a resource is included in the higher
level path, it will not be repeated in lower level paths.
Example
[0202] v[=attr1 rx Demo3]/*h/v
[0203] In third level in the result, Demo3 will not be
repeated.
v[=attr1 rx Demo3]/h/v
Regex Patterns
[0204] Query supports both regular expression string or regular
expression pattern id which will be defined in a separate file in
the following format:
<PatternList><Pattern id=" " extraDataPoints="
"><![CDATA[<pattern>]]></Pattern></PatternList>-
;
Example
[0205] <PatternList><Pattern id="rxp1"
extraDataPoints="30">
[0206] <![CDATA[9+]]></Pattern></PatternList>
Pattern with id "rxp2" will directly apply the regular expression
pattern to the performance data. ExtraDataPoints will be used in
the result set to return additional data in addition to the matched
values. It adds 30 points before and after to the matched
values.
Query Processing Flow
[0207] The configuration data tells the system what types of
resources have performance metric data stored in the system and
what are the attributes of each type of resource, some of which may
have had performance data measured. The configuration data
basically tells what resources have existed for what periods of
time.
[0208] FIG. 7 is a flowchart of the processing of the query
processor. When the query processor starts, it first reads the
query to determine the start and end times of the interval of
performance data to be searched, and then reads a configuration
data file to determine for the time frame of the query (as set by
the user by setting the start date and end date for the query
expression) what resources exist or have existed. These processes
are represented by step 210. If a resource or resources existed for
only part of the relevant query interval, the query processor
determines from the configuration data the valid times these
resources existed during the relevant interval, and, if the
resources still exist, at what time they came into existence during
the relevant query interval. Resources can come and go such as when
a server is taken offline or a disk is swapped out. Reading the
query and the configuration data file and determining what
resources existed at any time during the relevant interval is
symbolized by step 210. The configuration file also contains data
which tells which resources are related to the resources named in
the query. For example, a disk which is contained in or connected
to a particular server is indicated as related to that server.
[0209] The server reads all this data in the configuration file
and, in step 212, creates a map of only the relevant resources,
i.e., the resources of the system that match the resource type
identified at 208 in the query of FIG. 6 and which existed at any
time during the query interval and any related resources. In the
preferred embodiment, the string at 208 identifies only a resource
type. In this example of FIG. 6, the resource type is a host. Step
214 represents the process of loading the entire day of performance
metric data for the relevant day, relevant resources (named
resource and related resources) and the relevant attribute (the
attribute named in the query). This results in all the performance
data for all resources of that type being loaded into memory as
described below for the entire day or days which include the
relevant interval starting at the start time and ending at the end
time identified in query. These start and end times are given by
the user in separate boxes (not shown) from the query expression
box when the user enters the query expression of FIG. 6 by
interacting with a display on a computer that shows the query box
and start and end time boxes.
[0210] This filtering out of performance data for resources not of
the named type allows the query processor to easily and quickly
find performance metric data which has been stored in the NRDB for
only the relevant resource types indicated at 208 in the query
syntax of FIG. 6.
[0211] The query processor then starts parsing the query expression
and determines from element 213 of the query of FIG. 6 what type of
attribute data for the resource type named at 208 which is stored
in the NRDB and which the query processor needs to perform the
query. In the example of the query of FIG. 6, parsing the query and
reading portion 213 thereof, the query processor determines it will
be performing a search on performance metric data for CPU usage on
all hosts as identified by the string at 208. This is symbolized by
step 214 of FIG. 7.
[0212] Also in step 214, the query processor examines the start
time (date and time) and end time (date and time) set by the user
on the query screen (not shown). The query processor then goes to
the NRDB and examines the directory structures and finds the
directory structures for the relevant day or days that contain the
start time and end time of the query. The query processor then
determines which subdirectory or subdirectories in these relevant
directories contain performance metric data for resources of the
type indicated at 208 in FIG. 6. The query processor then
determines the text files in the relevant subdirectories and
determines which text files contain the performance metric data for
the group of attributes which contain the attribute identified in
the query expression, i.e., the attribute identified at 213. The
query processor also determines from the configuration data file
what other resources are related to the resource types identified
at 208 and loads the performance metric data for these related
resources for the relevant interval into memory also, which is also
part of step 214 in some embodiments.
[0213] Next, in step 216, the query processor determines whether
the needed data is already stored in cache. If so, the needed data
is loaded from the cache memory to save the time of a disk read. If
the needed data is not stored in the cache, the query processor
sends a read request to the operating system API to read the
appropriate text file or files containing the data needed for the
query into memory in step 218. Step 218 loads the entire day's
worth of performance data for the resources of the type identified
in the string at 208 in FIG. 6 and for the group of attributes
including the attribute identified at 213 of the query
expression.
[0214] Now all the performance metric data for the file containing
the performance metric data for the entire group of attributes that
contain the relevant attribute, and for the entire day or days
spanning the start date and end date are stored in memory. The data
in memory contains both performance metric data for attributes not
named in the query as well as performance metric data for the
relevant attribute which is outside the start time and end time
given in the query. To eliminate this excess data, the query
process builds a new string containing only the data for the
relevant attribute and only starting at the starting time and
ending at the ending time named in the query. This process is
symbolized by step 220. To do this, the query processor finds the
row in the loaded file which contains the performance metric data
for the relevant attribute identified at 213 of the relevant
resource identified at 208 and counts entries until it reaches the
value recorded for the named start time. That performance metric
value and all subsequent values extending out to the end time are
copied to a new file in the same sequence they were stored in the
NRDB, all as symbolized by step 220.
[0215] In step 222, the logic of the regular expression shown at
221 is applied to the performance data in the new file created in
step 220 to find values which meet the criteria expressed in the
regular expression at 221 of the search query for every resource of
the type identified at step 208. The values so found are returned
and decoded from unicode back to the original performance metric
values received from the probe. If multiple substrings from
multiple resources of the type indicated at 208 are found which
match the query, all such matching substrings are returned along
with identifying data as to which resource returned each string. In
some embodiments including the preferred embodiment, the metadata
about the resource identity (the specific host identity in the
example of FIG. 6), the attribute identity (CPU usage in the
example of FIG. 6), as well as the start time and end time of the
query and the times the returned values were recorded is also
returned for help in analyzing the results. In some embodiments,
only a true or false result is returned. In some embodiments, if a
true result is returned, and the sub string of performance metric
values which matched the regular expression is also returned after
being decoded from unicode back to the performance metric value
received from the probe.
Nested Queries
[0216] Sometimes complex situations arise where trouble shooting of
the performance metric data is needed to solve a problem. An
example would be where a host is running multiple virtual machines
and one of them has slowed down considerably or stopped responding
and the reason why needs to be determined. In such a case, a set of
nested queries such as those given below can be used to determine
the source of the problem.
[0217] vm[@readlatency rx b [U20-U1000] {5}/h[@readlatency rx b
[U20-U1000]{5}/vm[@readiop rx b [U1000-U2000]{5}]
[0218] The above query is actually three nested queries designed to
drill down into the performance data to find out what the problem
is with a slow virtual machine.
[0219] The first part of the query is: vm[@readlatency rx b
[U20-U1000] {5}/This query looks at the readlatency attribute (a
measure of speed) of all virtual machines which is between U20 and
U1000 for 5 consecutive readings. This range U20-U1000 finds all
the virtual machines which are running pretty slow.
[0220] The question then becomes why are these virtual machines
running slowly. To find that out, one question would be are the
hosts that are executing the code of the virtual machines
themselves running slowly for some reason. In parsing this query,
the query processor determines all host type resources which are
related to the virtual machine type identified by the string vm at
the beginning of the query. The performance metric data for all
these hosts is loaded into memory when the virtual machine
performance metric data is loaded into memory according to the
processing of FIG. 7. In order to find out if the host or hosts are
running slowly, the second part of the query is used. That part
is:
[0221] h[@readlatency rx b [U20-U1000]{5}/
[0222] This second part of the query looks at all the readlatency
performance metric values for host type resources that are related
to the virtual machine resource type identified in the first part
of the query and determines which ones of these hosts are running
slowly. The returned data indicates which hosts have slow read
latency. The question then becomes why is this host or hosts
running slowly. To answer that, the third part of the query is
used. That part determines which virtual machines which are related
to the hosts have high 10 operations going on which are bogging
down the hosts. The third part of the query is:
[0223] vm[@readiop rx b [V1000-V2000]{5}]
[0224] This query returns the identities of the virtual machine
which have high levels of input/output operations going on. This
high level of I/O operation will bog down the hardware of the host
and will be the explanation why other virtual machines have slowed
down or stopped. The results can then be used to shut down the
virtual machine that is bogging down the system or modify its
operations somehow so as to not slow down the other virtual
machines.
[0225] The results returned, for example, might indicate that
virtual machine 1 on host 1 is running slowly and host 1 is running
slowly because virtual machine 3 on that host is running a high
number of I/O operations. Another set of data that matches the
three queries may show also that virtual machine 2 running on host
2 is running slowly because host 2 is running slowly because
virtual machine 4 running on host 2 is carrying out a high number
of I/O operations.
Module Processing Flows
[0226] FIG. 8, comprised of FIGS. 8A through 8C, is a flowchart of
the processing of the probe data importer. The Probe Data Importer
runs a Data Import Scheduler routine which runs data import
operations at regular intervals, as symbolized by step 230. Step
232 checks the probe data folder for new data to be processed. Test
234 determines if new data has arrived, and, if not, processing
returns to step 230. If new data has arrived, step 236 is performed
to parse the list of files to get the list of configuration and
performance metric data files in the new data in sorted order. Test
238 determines if the new data has performance metric data in it.
If so, step 240 is performed to import the performance data. If the
new data does not have performance data files in it, processing
skips from step 238 to step 242 where a test is performed to
determine if configuration data has arrived. If not, processing
returns to step 230 to wait for the next data import. If new
configuration data has arrived, step 244 is performed to import the
new configuration data.
[0227] Step 246 starts the processing of performance metric data
files listed in the sorted list. Related performance counters of
each resource will be grouped together for storage and access
optimization. Step 248 creates file groups based on performance
counter group wherein one file group is formed for each performance
counter group. Step 250 creates a thread pool and processes the
file groups in multiple threads. Using Java API
(java.util.concurrent package), it creates a pool of threads and
each thread will pick one FileGroup at a time and processes it.
After completion of one FileGroup processing, the same thread will
pick the next FileGroup, if any, for processing and the process
repeats until all the FileGroups are processed. Total thread count
in the thread pool is configured through application properties
file. Step 252 is the processing for each thread. In each thread,
the files are read and the resources identified in the files are
found and resource counter groups are created. There is one
resource counter group per each resource. In step 254, another
thread pool is formed, and the resource counter groups are
processed as explained above. In step 256, for each thread, the
resource counter group data is processed, and data structures in
memory are updated to reflect the collected performance metric data
for each resource. The resource counters are used to determine
where in each text file each performance metric data value is to be
stored to properly reflect the time at which it was gathered.
Finally, in step 258, the data structures created in memory, i.e.,
the text files created when the performance metric values are
converted to unicode and stored in text files per the structure
described elsewhere herein, are written to non volatile storage of
the NRDB.
[0228] Step 260 on FIG. 8C represents the start of processing of
the configuration files listed on the sorted list. In step 262, the
configuration data file is parsed and the timestamp and resource
signature is found. Test 264 determines whether the resource
identified by the resource signature is found in the NRDB. If not,
step 266 creates a minisnapshot file in the NRDB using the
available configuration data. If test 264 determines that the
resource identified in the configuration file is already in the
NRDB, step 268 is jumped to where the configuration changes and
events are saved in an updates file in the NRDB. Finally, in step
270, the in-memory configuration data is refreshed by re-loading it
from the NRDB.
[0229] FIG. 9, comprised of FIGS. 9A and 9B, is a module diagram
and flowchart of the processing of the NRDB Access manager module.
The NRDB access manager module 300 controls access to the non
relational data base file system 302 where the configuration data
and performance metric data is stored. The NRDB access manager
module 300 retrieves data from the NRDB and uses a cache 304 in
memory of the server which is running module 300 and a cache 306 in
the file system to store data which is frequently accessed to speed
up data access. Performance data and configuration data are
imported from the probes by the Probe Data Importer module 308 by
the processing previously described and put into the NRDB via the
NRDB access manage module 300. Query requests to analyze the
performance metric data in the NRDB are handled by Query Request
Handler module 310 which accesses the data in the NRDB via the NRDB
Access Manager module 300.
[0230] In FIG. 9B, the NRDB Access Manager processing starts with
receiving a request for performance metric data from the Query
Process Handler, this request symbolized by line 312. Step 314
determines if the requested performance data is in the performance
data cache 304 in the system RAM and in the file system. If it is,
step 316 is jumped to, and the performance data is returned from
the cache to the Query Process Handler 310. If test 314 determines
the performance data requested is not in the cache, step 318 is
performed to load the requested data from the NRDB file system into
the cache 304, and then step 316 returns the requested data to the
Query Process Handler 310.
[0231] The Probe Data Importer 308 adds updated and new
configuration data and new performance data via data path 321 to
the NRDB through step 320, and updates the respective configuration
data cache 323 in RAM or the performance data cache 304 in RAM and
in the NRDB file system itself. NRDB Access Manager before
processing performance metric data gets the in-memory
representation (Java object) of the performance metric data through
Performance cache. Performance cache first verifies in memory
whether it is already loaded from the file. If not, it loads the
data from the file for the given date. If data is not available, it
creates a file with template data (default values) for all the
sampling intervals for that day. Based on the start time, it
updates the in-memory performance metric data at appropriate
locations. Once all the metrics data in the group is processed, it
commits the changes back to the file. The data will be compressed
(deflate format) before saved into the file.
[0232] FIG. 10 is a block diagram of one embodiment of the overall
system including the major functional modules in the central server
called Megha.TM., where the query request processing for analysis
of performance metric data occurs and where the NRDB stores the
performance metric data and configuration data. Persons who want to
query the performance metric data send an asynchronous request
using a web browser running on a client computer 330 to a Web
Request Controller 332 running on the Megha server using a REST
application programmatic interface (API). The Web Request
Controller 332 receives the request, validates it and then forwards
it to the Query Request Processor module 310 with an asynchronous
Java API call. Then the Web Request Controller returns the status
to the client computer 330 by hinting that the client needs to come
back for the result. The Query Request Processor 310 processes the
request and incrementally saves the results in a Results Cache 311.
The client computer 330 then sends back a request for the results
to the Web Request Controller 332 which checks the Results Cache
311. The results are then returned by the Web Request Controller
332 to the client 330 in an XML format if available. If the Query
Request Processor is still processing the request, the Web Request
Controller send the status hint to the client indicating it needs
to send another request for the results later. The Report Engine
313 is a Java class object which sends query requests to the Query
Request Processor 310 using Java API invocation asynchronously and
reads the results data from the Result Cache 311 through a Java
API.
[0233] FIG. 11 is a flowchart of the processing by one embodiment
of the Query Request Processor. Step 320 parses the search query.
If the search query has an invalid format, the result cache is
updated with an error and processing is terminated. Each query
starts with a high level