gridmix
Gridmix is a benchmark tool for Hadoop cluster. See more details about Gridmix in the Gridmix Guide.
The usage without configuration parameters is as follows:
$ hadoop gridmix [-generate <size>] [-users <users-list>] <iopath> <trace>
-R |
Recursively through the directory structure |
The usage with the configuration parameters is as follows:
$ hadoop gridmix \
-Dgridmix.client.submit.threads=10 -Dgridmix.output.directory=foo \
[-generate <size>] [-users <users-list>] <iopath> <trace>