Test-infra: milestone bot spam detracts from monitoring issues

Created on 1 Jun 2018  Â·  12Comments  Â·  Source: kubernetes/test-infra

It's very difficult to actually subscribe to Kubernetes issues / PRs when code slush / freeze comes around due to the enormous amount of these messages:

image

We should seriously consider if posting this once-a-day on each thread actually helps anyone.

/sig contribex
/cc @jberkus @cblecker @cjwagner @marun

sicontributor-experience sirelease

Most helpful comment

@cblecker yep, this can be closed.

Milestone munger was disabled:

/close

All 12 comments

/sig release

So the milestone bot nags folks daily or every 3 days (depending on the issue) because we expect progress in that window of time, and it's critical that issue/PR owners keep their issues updated.

While I agree that the milestone bot hasn't been spectacularly effective in doing this, do you have suggestions on something else we could do that isn't bot nagging?

Can we nag folks that actually have permissions to fix these things to look
through the issues that have some label instead? Most of the people
receiving these messages can't really do anything about them.

On Fri, Jun 1, 2018, 11:53 AM Josh Berkus notifications@github.com wrote:

So the milestone bot nags folks daily or every 3 days (depending on the
issue) because we expect progress in that window of time, and it's critical
that issue/PR owners keep their issues updated.

While I agree that the milestone bot hasn't been spectacularly effective
in doing this, do you have suggestions on something else we could do that
isn't bot nagging?

—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
https://github.com/kubernetes/test-infra/issues/8205#issuecomment-393977137,
or mute the thread
https://github.com/notifications/unsubscribe-auth/AA4BqwcYZ4XRLfmSGmoGPxf4tCZJfMZAks5t4Y27gaJpZM4UXHcr
.

That would be great. And ... a job for test-infra?

Step one is completing replacing the milestone munger with Prow. Until we do that, no other changes are happening.

I'd suggest some off-github method for doing this, or some way of aggregating the issues/PRs so that it's easier to review/digest, then a bunch of spam every 3 days.

I would actually say it's up to sig-release (not just the current release team, but those involved with the sig) to define what the desired process/procedure is. Then we can figure out engineering resources to make the code happen.

The key being, and feedback I'd agree with: The current process is ineffective, spammy, and presents a poor contributor experience. We don't really have time to fix it in the middle of the 1.11 cycle, but we need a long view to this process.

@kubernetes/sig-release-members @kubernetes/sig-contributor-experience-misc-use-only-as-a-last-resort

That would be great. And ... a job for test-infra?

The issue is filed in test-infra, I'm happy to work on something once we have consensus on what to do or even if we agree that we don't want the current situation, but sig-release and sig-contributor-experience need to be involved in discussing what we should do before any test-infra folk change this.

@cblecker's comment sums up my thoughts nicely:

The current process is ineffective, spammy, and presents a poor contributor experience. We don't really have time to fix it in the middle of the 1.11 cycle, but we need a long view to this process.

Looking forward beyond the current release I'd love to identify a better solution to getting things through the milestone properly and implement.

Edit: which does not include "let's take the same bad process and run it as part of Prow instead of mungegithub". I don't think that's particularly helpful.

Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle stale

/remove-lifecycle stale

Very much still active :)

Actually, did we just move away from the milestone maintainer completely?

@tpepper

@cblecker yep, this can be closed.

Milestone munger was disabled:

/close

Yes!! :tada:

Was this page helpful?
0 / 5 - 0 ratings

Related issues

fen4o picture fen4o  Â·  4Comments

cjwagner picture cjwagner  Â·  3Comments

sjenning picture sjenning  Â·  4Comments

cjwagner picture cjwagner  Â·  3Comments

chaosaffe picture chaosaffe  Â·  3Comments