Skip to content

Commit

Permalink
Address Documentation Issue #57: Include clarification on selecting m…
Browse files Browse the repository at this point in the history
…ultiple node selection parameters in USAGE.md (#59)

USAGE.md mentions how tests can be configured using a combination of parameters. The host, job, and nodelabel parameters all involve node selection. To clarify behavior when multiple node selection parameters are used together, I mentioned that Autopilot runs the tests on nodes that match any of the specified parameters (set union rather than set intersection).

I specifically added the following sentences under the parameters list in USAGE.md: Note that if multiple node selection parameters (host, job, nodelabel) are provided together, Autopilot will run tests on nodes that match any of the specified parameters (set union). For example, the following command will run the pciebw test on all nodes that either have the label label1 OR are running the job jobKey=job2 because both nodelabel and job parameters are provided in the input:

curl "http://<route-name>/status?check=pciebw&nodelabel=label1&job=default:jobKey=job2"


Signed-off-by: Anish Sinha <[email protected]>
  • Loading branch information
Anish701 authored Nov 18, 2024
1 parent b1b87c3 commit 9618968
Showing 1 changed file with 6 additions and 0 deletions.
6 changes: 6 additions & 0 deletions USAGE.md
Original file line number Diff line number Diff line change
Expand Up @@ -40,6 +40,12 @@ All tests can be tailored by a combination of:

Some health checks provide further customization. More details on all the tests can be found [here](https://github.com/IBM/autopilot/autopilot-daemon/HEALTH_CHECKS.md)

Note that if multiple node selection parameters (`host`, `job`, `nodelabel`) are provided together, Autopilot will run tests on nodes that match _any_ of the specified parameters (set union). For example, the following command will run the `pciebw` test on all nodes that either have the label `label1` OR are running the job `jobKey=job2` because both `nodelabel` and `job` parameters are provided in the input:

```bash
curl "http://<route-name>/status?check=pciebw&nodelabel=label1&job=default:jobKey=job2"
```

## DCGM

This test runs `dcgmi diag`, and we support only `r` as [parameter](https://docs.nvidia.com/datacenter/dcgm/latest/user-guide/dcgm-diagnostics.html#command-line-options).
Expand Down

0 comments on commit 9618968

Please sign in to comment.