cbworkloadgen
Generates random data and performs reads and writes
SYNOPSIS
cbworkloadgen [--node <host>] [--username <user>] [--password <password>]
[--bucket <bucket>] [--ssl] [--threads <num>]
[--ratio-sets <ratio>] [--max-items <num>] [--prefix <str>]
[--size <bytes>] [--json] [--low-compression] [--loop]
[--xattr] [--collection <id>] [--verbose] [--help]
DESCRIPTION
The cbworkloadgen tool is useful to test a Couchbase cluster as well as to
generate test data sets. The tool is installed in the following locations:
| Operating system | Location |
|---|---|
Linux |
/opt/couchbase/bin/ |
Windows |
C:\Program Files\Couchbase\Server\bin\ |
Mac OS X |
/Applications/Couchbase Server.app/Contents/Resources/couchbase-core/bin/ |
Options
- -n,--node <host>
-
Target node address in the form
ip:portorhostname:portfor example-n 127.0.0.1:8091. If non provided it will default to127.0.0.1:8091. - -u,--username <user>
-
Specifies the username of the user executing the command. If you do not have a user account with permission to execute the command then it will fail with an unauthorized error.
- -p,--password <password>
-
Specifies the password of the user executing the command. If you do not have a user account with permission to execute the command then it will fail with an unauthorized error.
- -b,--bucket <bucket>
-
Specifies the bucket to operate on. If non provided it will use default.
- --ssl
-
Use SSL for transferring and reading data.
- -t,--threads <num>
-
Number of concurrent threads use to insert or get data. By default it will use 1.
- -r,--ratio-sets <ratio>
-
The percentage of the workload that are writes; the remainder will be reads. By default it is set to 0.95.
- -i,--max-items<num>
-
Number of inserted items. The default is 10000.
- --prefix <prefix>
-
Prefixed to use for the document keys. The default is
pymc. - --s,--size <bytes>
-
Minimum value size in bytes, it defaults to to bytes.
- --json
-
When specified the program will insert json documents.
- --low-compression
-
When specified the generated documents will be difficult to compress.
- --loop
-
The process will loop forever until the user interrupts it.
- --xattr
-
When specified the inserted documents will have extended attributes.
- -c,--collection <id>
-
The collection ID to perform the operations on. A collection with the given ID must exist. The ID is expected to be a number in hex format for example '0x16'.
- -v,--verbose
-
Verbose logging; more -Vs provide more verbosity.
- -h,--help
-
Shows help message.
Examples
To generate workload on a given Couchbase Server use:
$ cbworkloadgen -n 10.5.2.45:8091 -b my-bucket -u Administrator -p password
This command will create 10000 binary documents with keys prefixed by pymc
in the target bucket. The output of the command would look as follows:
[####################] 100.0% (1111110/estimated 1111110 msgs)
bucket: default, msgs transferred...
: total | last | per sec
byte : 111111000 | 111111000 | 636622.5
done
To generate a data set of one million JSON documents on a cluster use:
$ cbworkloadgen -n 10.5.2.45:8091 -b my-bucket -u Administrator -p password \ -i 1000000 -j -r 1
By default the documents generated by the tool are very compressible which
may not be ideal for some case scenarios to change this use the
--low-compression flag as follows:
$ cbworkloadgen -n 10.5.2.45:8091 -b my-bucket -u Administrator -p password \ -i 1000000 -j -r 1 --low-compression -s 20480
The command above will create documents with bodies of at least 20480 bytes (20Kb) that are hard to compress.
To generate documents into the collection my_collection you first must find the
collection ID. This can be done by using the command below:
$ cbstats -u Administrator -p Password -b my-bucket 10.5.2.45:11210 collections | grep ":name:.*my_collection" 0x0:0x8:name: my_collection
We can see that the collection ID is 0x08, so we can now use the -c flag to
generate the documents as shown below.
$ cbworkloadgen -n 10.5.2.45:8091 -b my-bucket -u Administrator -p password \ -i 1000000 -j -r 1 -c 0x08