Skip to content

Tools for using FIO benchmarking, creating GNUPlots and serving them via Docker.

License

Notifications You must be signed in to change notification settings

wallnerryan/fio-tools

Repository files navigation

FIO-Tools

Alt text Alt text

How to use this repo

  1. Build the images or use the public images
  2. Create a Fio Jobfile
  3. Run the fio-tool image
docker run -v /tmp/fio-data:/tmp/fio-data \
-e JOBFILES=<your-fio-jobfile> \
portworx/fio-tool

If your file is a remote raw text file, you can use REMOTEFILES

docker run -v /tmp/fio-data:/tmp/fio-data \
-e REMOTEFILES="http://url.com/<your-job>.fio" \
-e JOBFILES=<your-fio-jobfile> wallnerryan/fio-tool
  1. Run the fio-genplots script
docker run -v /tmp/fio-data:/tmp/fio-data wallnerryan/fio-genplots \
<fio2gnuplot options>
  1. Serve your Graph Images and Log Files
docker run -p 8000:8000 -d -v /tmp/fio-data:/tmp/fio-data \
wallnerryan/fio-plotserve
  1. Easiest Way, run the "all in one" image. (Will auto produce IOPS and BW graphs and serve them)
docker run -p 8000:8000 -v /tmp/fio-data \
-e REMOTEFILES="http://url.com/<your-job>.fio" \
-e JOBFILES=<your-fio-jobfile> \
-e PLOTNAME=MyTest \
-d --name MyFioTest wallnerryan/fiotools-aio

Other Examples

To use with Kubernetes

Use a Deployment

kubectl apply -f kubernetes/fiotools-aio-portworx-read.yaml
storageclass.storage.k8s.io "fio-tester-class-read" created
persistentvolumeclaim "fio-data-read" created
deployment.extensions "fio-tester-read" created
service "fiotools-read" created

kubectl apply -f kubernetes/fiotools-aio-portworx-write.yaml
storageclass.storage.k8s.io "fio-tester-class-write" created
persistentvolumeclaim "fio-data-write" created
deployment.extensions "fio-tester-write" created
service "fiotools-write" created

kubectl get svc fiotools-read
kubectl get svc fiotools-write

Access the Output

kubectl port-forward service/fiotools-read 8001:8001 Visit http://localhost:8001

kubectl port-forward service/fiotools-write 8000:8000 Visit http://localhost:8000

OR

Visit http://:[8000|8001] as long as the firewall allows 8000 and 8001 to the workers.

Note, you can change the ENV variables to submit a new job. Just provide a new job url to REMOTEFILES and update the name of JOBFILES to the name of the .fio file and provide and optional new PLOTNAME

env:
  - name: REMOTEFILES
    value: "https://gist.githubusercontent.com/wallnerryan/06cb07d3d8bee67af025a60a88da053f/raw/a46d97f30b79c2a2a6b42333e7114d85e84c450f/editablejob.fio"
  - name: JOBFILES
    value: editablejob.fio
  - name: PLOTNAME
    value: editablejob
Important
  • *Your FIO JOBFILES should reference a `directory=/my/mounted/volume" to test against docker volumes
  • *If you want to run more than one all-in-one job, just use -v /tmp/fio-data instead of -v /tmp/fio-data:/tmp/fio-data This is only needed when you run the individual tool images seperately
To use with docker and docker volumes
docker run \
-e REMOTEFILES="https://gist.githubusercontent.com/wallnerryan/fd0146ee3122278d7b5f/raw/cdd8de476abbecb5fb5c56239ab9b6eb3cec3ed5/job.fio" \
-v /tmp/fio-data:/tmp/fio-data \
--volume-driver flocker \
-v myvol1:/myvol \
-e JOBFILES=job.fio wallnerryan/fio-tool

To produce graphs, run the genplots container, -p <pattern of your log files>

Produce Bandwidth Graphs

docker run -v /tmp/fio-data:/tmp/fio-data wallnerryan/fio-genplots \
-t My16kAWSRandomReadTest -b -g -p *_bw*

Produce IOPS graphs

docker run -v /tmp/fio-data:/tmp/fio-data wallnerryan/fio-genplots \
-t My16kAWSRandomReadTest -i -g -p *_iops*

Simply serve them on port 8000

docker run -p 8000:8000 -d \
-v /tmp/fio-data:/tmp/fio-data \
wallnerryan/fio-plotserve

To use the all-in-one image

docker run \
-p 8000:8000 \
-v /tmp/fio-data \
-e REMOTEFILES="https://gist.githubusercontent.com/wallnerryan/fd0146ee3122278d7b5f/raw/006ff707bc1a4aae570b33f4f4cd7729f7d88f43/job.fio" \
-e JOBFILES=job.fio \
-e PLOTNAME=MyTest \
—volume-driver flocker \
-v myvol1:/myvol \
-d \
—name MyTest wallnerryan/fiotools-aio
To use with docker-machine/boot2docker

You can use a remote fit configuration file using the REMOTEFILES env variable.

docker run \
-e REMOTEFILES="https://gist.githubusercontent.com/wallnerryan/fd0146ee3122278d7b5f/raw/d089b6321746fe2928ce3f89fe64b437d1f669df/job.fio" \
-e JOBFILES=job.fio \
-v /Users/wallnerryan/Desktop/fio:/tmp/fio-data \
wallnerryan/fio-tool

(or)

If you have a directory that already has them in it

docker run -v /Users/wallnerryan/Desktop/fio:/tmp/fio-data \
-e JOBFILES=job.fio wallnerryan/fio-tool

To produce graphs, run the genplots container, -p <pattern of your log files>

docker run \
-v /Users/wallnerryan/Desktop/fio:/tmp/fio-data wallnerryan/fio-genplots \
-t My16kAWSRandomReadTest -b -g -p *_bw*

Simply serve them on port 8000

docker run -v /Users/wallnerryan/Desktop/fio:/tmp/fio-data \
-d -p 8000:8000 wallnerryan/fio-plotserve
To use with docker that is not boot2docker or docker-machine , /tmp/fio-data is a still a VOLUME

You can use a remote configuration script

docker run \
-e REMOTEFILES="https://gist.githubusercontent.com/wallnerryan/fd0146ee3122278d7b5f/raw/2eb7d0ae9b77fa5a93662fe8088df2d83fff9ab2/job.fio" \
-v /tmp/fio-data:/tmp/fio-data \
-e JOBFILES=job.fio wallnerryan/fio-tool

(or)

You can create a directory and put it locally on the server where the container will run

mkdir /tmp/fio-data
cp <your FIO job file> /tmp/fio-data/
docker run -v /tmp/fio-data:/tmp/fio-data \
-e JOBFILES=<your FIO job> wallnerryan/fio-tool

To produce graphs, run the genplots container, -p <pattern of your log files>

docker run -v /tmp/fio-data:/tmp/fio-data wallnerryan/fio-genplots \
-t My16kAWSRandomReadTest -b -g -p *_bw*

Simply serve them on port 8000

docker run -p 8000:8000 -d \
wallnerryan/fio-plotserve
Notes
  • The fio-tools container will clean up the /tmp/fio-data volume by default when you re-run it. If you want to save any data, copy this data out or save the files locally.
  • When you serve on port 8000, you will have a list of all logs created and plots created, click on the .png files to see graph (see below for example screen)

Alt text

bw= and BW= explained

About

Tools for using FIO benchmarking, creating GNUPlots and serving them via Docker.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published