Super committers are a group of TSC-approved individuals within the ONAP community with the power to merge patches on behalf of projects during approved Release Activities.


Super committers are still affected by NACR Gerrit rules: https://gerrit-review.googlesource.com/Documentation/prolog-cookbook.html#NonAuthorCodeReview

This means that, even though super committers can +2/submit changes across repos, they will still not be able to bypass this rule for self authored changes. 

Super Committer Activities

Super committers are given super committer powers ONLY during TSC-approved activities and are not a power that is active on a regular basis. Once one of the TSC-approved activities are triggered, helpdesk will enable the permissions listed for the respective activities for the duration of that activity.

Code Freeze (M3)

This activity has been pre-approved by the TSC and does not require a TSC vote. Helpdesk should be notified to enable the permissions and again to disable the permissions once activities are complete.

Super committers are granted powers to merge blocking patches for the duration code of freeze until a release is approved. This permission is only granted for the specific branch that is to be released.

The following powers are granted:

  • +2 Code-Review
  • Submit button access

During this time Super Committers can ONLY merge patches that have been validated during E2E testing within the integration lab. The patch passes Jenkins Verify check. If neither of these conditions are met then DO NOT merge the patch. Generally, this should be used for High and Highest priority JIRA tickets that are on the integration blocker list.

Who can be super committers?

Anyone as long as it makes sense for the TSC.

It would seem useful to have super committers in more than one region so that they can help support testing/integration/merge activities across regions/time zones.

The Integration PTL will recommend super committer candidates, get PTL voice vote concurrence vote during a PTL status meeting and recommend the candidates to the TSC for approval


Approved Super Committersregion / time zone
Central European 

Jessica Wagantall (LF Support)

NAR (Pacific)

Bengt Thuree (LF Support)

APAC 

Matthew Watkins (LF Support)

GMT

Kevin Sandi (LF Support)

NAR ( Central )

Bin Yang   (Windriver)

China
Central European
Central European
Central European
India


Integration Team committers:

They are validating the ONAP release by running the E2E tests. They raise the bugs, and test their fixes. Giving them the rights to merge the fix would allow them to move faster towards a more sanitized code base.

Doc Team committers:

They are ensuring the documentation is as accurate as possible, by updating project's documentation. Giving them to rights to merge the doc update, they would have the ability to provide better documentation faster.

To request the permissions

  • Provide a specific timeline of how long should the permissions be enabled for (Can't be enabled all the time)
  • Get TSC approval
  • request the permissions to be turned on via support.linuxfoundation.org

Quality checks

Number of patches merged

Number of patches merged that later were found to cause errors

PTL feedback per release cycle



  • No labels

6 Comments

  1. Please find one comment:

    * How do we handle PTL -1 and Super committer +2? Who will have the last word? (smile)


    1. PTL has final say but PTL should do a -2 to block the merge not a -1

  2. No direct comment to the existing text, but I would like to read motivation for the proposal together with the proposal. Clearly there is an issue that has been recognised and the current proposal is to address that issue. It would be helpful to reiterate this so as to provide context for the reader. Also would be interesting to read what is the assumed change in the project teams' dynamics when "Super Committer" is in place.

    Maybe these have been brought up in TSC discussions leading to the proposal, maybe capturing those here would suffice.

  3. Brian Freeman Alexis de Talhouët

    I have added the super committers group in Gerrit and have added the permissions similar to how ODL has in:

    https://gerrit.onap.org/r/#/admin/projects/All-Projects,access


    Please let me know for any questions. 

  4. I tried a random sdnc gerrit review and only had +1 - (I did logout/login) - is there a time delay ?


  5. Brian Freeman  Can you please try once more? 

    I think ODL grants this permission whenever they decide is critical timing and remove after. That's why I didn't set it on my end 

    I have set it now but we might want to double check is this is something we want to keep "activated" or we only want to activate 

    whenever there is a deadline. Let me know.