Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Add numactl RPM requirement #2879

Merged
merged 1 commit into from
Jun 6, 2022

Conversation

portante
Copy link
Member

@portante portante commented Jun 3, 2022

Both the pbench-uperf and pbench-linpack commands offer interfaces that rely on numactl.

Both the `pbench-uperf` and `pbench-linpack` commands offer interfaces
that rely on `numactl`.
@portante portante added bug Agent uperf pbench-uperf benchmark related packaging Issues related to software packaging Linpack Related to `pbench-linpack` labels Jun 3, 2022
@portante portante added this to the v0.72 milestone Jun 3, 2022
@portante portante requested review from ndokos and webbnh June 3, 2022 20:15
@portante portante self-assigned this Jun 3, 2022
agent/Makefile Outdated Show resolved Hide resolved
@portante
Copy link
Member Author

portante commented Jun 4, 2022

@ndokos asked elsewhere: "Why not modify pbench-uperf and pbench-linpack to complain and abort the run as with tools?"

The answer is that those benchmark convenience scripts are not tools. They are part of the interface provided by the pbench-agent, and all the dependencies for those interfaces to work properly should be provided by direct pbench-agent RPM dependencies.

If we take the logic that we are applying to tools, that is that the dependencies for tools are not applied to the pbench-agent RPM, then the pbench-agent RPM would have NO dependencies at all.

Further, the numactl command does not fall in the category of a "tool", since it does not gather metrics or other data about a program or the system. Rather, it offers controls over the system. It's sibling, the numastat command, is a tool, though it is provided by the numactl RPM.

@portante
Copy link
Member Author

portante commented Jun 4, 2022

FYI, the numactl RPM was removed in PR #1887 (commit 73ccffc).

Copy link
Member Author

@portante portante left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Remove the Fedora 36 changes from this PR.

@portante portante requested a review from dbutenhof June 5, 2022 00:36
@ndokos
Copy link
Member

ndokos commented Jun 5, 2022

But one could argue that the pbench-uperf interface also depends on uperf, and we don't make pbench-agent depend on uperf.

@dbutenhof
Copy link
Member

BTW: While using the branch name to link to Jira clearly works, it also means the final main git history won't include the Jira number, while putting it in the message will. I like having it in the history for future reference. (Granted, we can always edit the log on merge; but if we put it there in the first place, we don't have to.)

Copy link
Member

@webbnh webbnh left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

👍

@ndokos ndokos merged commit 2f93c6f into distributed-system-analysis:main Jun 6, 2022
@portante
Copy link
Member Author

portante commented Jun 6, 2022

I am not convinced of the Jira integration yet. Being able to search in Jira using the GitHub ID is sufficient. No need to litter GitHub with Jira references that nobody outside Red Hat can access.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Agent bug Linpack Related to `pbench-linpack` packaging Issues related to software packaging uperf pbench-uperf benchmark related
Projects
None yet
Development

Successfully merging this pull request may close these issues.

5 participants