Re: [chromium-dev] PSA: "CQ dry run" button in Rietveld

12 views
Skip to first unread message

Ravi

unread,
Mar 31, 2015, 1:33:07 PM3/31/15
to chromi...@chromium.org, sad...@google.com, rmi...@chromium.org, infra-dev


On Tuesday, March 31, 2015 at 12:10:26 PM UTC-4, Gabriel Charette wrote:
Nice! I've been using this instead of git cl try now, but I find it annoying that it spews the same messages as if I was trying to commit (potentially making my reviewers panic that I'm landing unfinished code?), could we make it not output CQ messages (or at least word them differently and perhaps only mail them to the owner of the issue)?

Thanks!
Gab

On Tue, Mar 31, 2015 at 8:54 AM Sadrul Chowdhury <sad...@chromium.org> wrote:
On Mon, Mar 30, 2015 at 9:13 PM, Ravi <rmi...@chromium.org> wrote:
>
> There have been recent changes made to the COMMIT=false keyword in CQ:
> * The LGTM check is skipped if the owner of the issue is a committer.
> * The issue is no longer automatically closed after the CL goes through the
> CQ successfully (this only made sense for e2e test CLs created by
> commit-bot).
>
> The new "CQ dry run" button in Rietveld (added in chromium:467912) works
> similar to the COMMIT=false keyword, it runs all CQ verifiers including
> triggering the CQ trybots, but with the following advantages:
>
> * The username of who triggered the dry run is now saved in Rietveld. The CQ
> skips the LGTM check if the triggerer is a project committer. This makes it
> possible for a reviewer to CQ dry run the CL of an external contributor
> without having to give an LGTM first.
> * Easy way for developers to dry run their change with a few clicks instead
> of having to know about the magic COMMIT=false keyword.
> * Avoids accidental commits due to keyword misspells.
>
> This button/link is available in both Rietveld UIs. As with everything else
> this change looks and feels much better in the new UI.
>
> Next I will work on adding a new flag in "git cl upload" that sets the CQ
> dry run bit when specified.

Would it make sense to have 'git cl try' pick the same set of trybots,
or is that already the case?

+infra-dev

I do not think this is the case right now but infra-dev would know more.
After the CQ config moves to the individual repos it could be possible to add the ability to run something like 'git cl try -t cq'. It would be cool to be able to define different trybot sets in the config as well.

The change I am working on will allow you to trigger a CQ dry run by running 'git cl upload --dry_run' or 'git cl dry_run', this will kick off your CQ trybots in addition to the other CQ checks.

 

Sadrul

>
>
> Thanks,
> Ravi
>
> --
> --
> Chromium Developers mailing list: chromi...@chromium.org
> View archives, change email options, or unsubscribe:
> http://groups.google.com/a/chromium.org/group/chromium-dev
>
> To unsubscribe from this group and stop receiving emails from it, send an
> email to chromium-dev+unsubscribe@chromium.org.

--
--
Chromium Developers mailing list: chromi...@chromium.org
View archives, change email options, or unsubscribe:
    http://groups.google.com/a/chromium.org/group/chromium-dev

Ravi

unread,
Mar 31, 2015, 1:33:38 PM3/31/15
to chromi...@chromium.org, sad...@google.com, rmi...@chromium.org, infr...@chromium.org


On Tuesday, March 31, 2015 at 1:33:06 PM UTC-4, Ravi wrote:


On Tuesday, March 31, 2015 at 12:10:26 PM UTC-4, Gabriel Charette wrote:
Nice! I've been using this instead of git cl try now, but I find it annoying that it spews the same messages as if I was trying to commit (potentially making my reviewers panic that I'm landing unfinished code?), could we make it not output CQ messages (or at least word them differently and perhaps only mail them to the owner of the issue)?

I will work on a change to improve the CQ messages for dry runs.
Thanks for the feedback!

Paweł Hajdan, Jr.

unread,
Apr 1, 2015, 5:50:27 AM4/1/15
to Ravi, chromium-dev, Sadrul Chowdhury, infr...@chromium.org
Chromium "git cl try" is already synchronized with the CQ config since January (https://codereview.chromium.org/823823002).

Paweł

--
You received this message because you are subscribed to the Google Groups "infra-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to infra-dev+...@chromium.org.
To post to this group, send email to infr...@chromium.org.
To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/infra-dev/68d93017-cfac-4427-a617-a1c178ee2976%40chromium.org.

Toby Sargeant

unread,
Apr 1, 2015, 9:26:29 AM4/1/15
to phajd...@chromium.org, Ravi, chromium-dev, Sadrul Chowdhury, infr...@chromium.org
Would it be possible to also skip the LGTM check if you have trybot access, but are not a committer? I guess that's a rare case (although I just stumbled over it) so it's probably too much work to justify...

Cheers,
Toby.

Ravi Mistry

unread,
Apr 1, 2015, 9:32:28 AM4/1/15
to infr...@chromium.org, phajd...@chromium.org, rmi...@chromium.org, chromi...@chromium.org, sad...@google.com, tobi...@chromium.org


On Wednesday, April 1, 2015 at 9:26:29 AM UTC-4, Toby Sargeant wrote:
Would it be possible to also skip the LGTM check if you have trybot access, but are not a committer? I guess that's a rare case (although I just stumbled over it) so it's probably too much work to justify...

That would be ideal. However I am not sure if there currently exists an easy way to check if a user has trybot access from the CQ. Again, infra-dev would know more.
 

Cheers,
Toby.



--
--
Chromium Developers mailing list: chromi...@chromium.org
View archives, change email options, or unsubscribe:
    http://groups.google.com/a/chromium.org/group/chromium-dev

--
You received this message because you are subscribed to the Google Groups "infra-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to infra-dev+...@chromium.org.
To post to this group, send email to infr...@chromium.org.
To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/infra-dev/68d93017-cfac-4427-a617-a1c178ee2976%40chromium.org.

Ravi

unread,
Apr 1, 2015, 9:34:04 AM4/1/15
to chromi...@chromium.org, phajd...@chromium.org, rmi...@chromium.org, sad...@google.com, infr...@chromium.org


On Wednesday, April 1, 2015 at 9:27:00 AM UTC-4, Toby Sargeant wrote:
Would it be possible to also skip the LGTM check if you have trybot access, but are not a committer? I guess that's a rare case (although I just stumbled over it) so it's probably too much work to justify...

That would be ideal. However I am not sure if there currently exists an easy way to check if a user has trybot access from the CQ. Again, infra-dev would know more.
 

Cheers,
Toby.

To unsubscribe from this group and stop receiving emails from it, send an email to infra-dev+unsubscribe@chromium.org.

To post to this group, send email to infr...@chromium.org.

Paweł Hajdan, Jr.

unread,
Apr 8, 2015, 6:04:59 AM4/8/15
to Ravi, Sergiy Byelozyorov, chromium-dev, Sadrul Chowdhury, infr...@chromium.org
We're considering some changes to how the access lists are stored.

A possible bug to watch is https://code.google.com/p/chromium/issues/detail?id=465911 (Googlers only).

I think Sergiy may comment more on this.

Paweł

Ravi Mistry

unread,
Apr 10, 2015, 7:22:52 AM4/10/15
to Paweł Hajdan, Jr., Sergiy Byelozyorov, chromium-dev, Sadrul Chowdhury, infr...@chromium.org
Update:

You can now trigger CQ dry runs from the command line when uploading patches with "git cl upload --cq-dry-run".
To use the new flag please sync your depot_tools checkout past r294699.
This functionality was added in chromium:472690.

Thanks,
Ravi

Greg Thompson

unread,
Apr 10, 2015, 11:32:10 AM4/10/15
to Ravi Mistry, Paweł Hajdan, Jr., Sergiy Byelozyorov, chromium-dev, Sadrul Chowdhury, infr...@chromium.org
On Fri, Apr 10, 2015 at 7:22 AM, Ravi Mistry <rmi...@chromium.org> wrote:
Update:

You can now trigger CQ dry runs from the command line when uploading patches with "git cl upload --cq-dry-run".
To use the new flag please sync your depot_tools checkout past r294699.
This functionality was added in chromium:472690.

Since the default trybots are in sync with the CQ, what is the benefit to using the CQ dry run over "git cl try"?

To unsubscribe from this group and stop receiving emails from it, send an email to infra-dev+...@chromium.org.

To post to this group, send email to infr...@chromium.org.

Ravi

unread,
Apr 10, 2015, 11:37:06 AM4/10/15
to chromi...@chromium.org, rmi...@chromium.org, phajd...@chromium.org, ser...@chromium.org, sad...@google.com, infr...@chromium.org


On Friday, April 10, 2015 at 11:32:54 AM UTC-4, Greg Thompson wrote:
On Fri, Apr 10, 2015 at 7:22 AM, Ravi Mistry <rmi...@chromium.org> wrote:
Update:

You can now trigger CQ dry runs from the command line when uploading patches with "git cl upload --cq-dry-run".
To use the new flag please sync your depot_tools checkout past r294699.
This functionality was added in chromium:472690.

Since the default trybots are in sync with the CQ, what is the benefit to using the CQ dry run over "git cl try"?

Not all projects have "git cl try" configured that way, also you run through the other CQ verifiers.
But if your project has "git cl try" configured to run CQ trybots and you do not care about the other verifiers then AFAIK there is no other benefit.
 

Dana Jansens

unread,
Apr 10, 2015, 5:28:09 PM4/10/15
to Ravi Mistry, chromium-dev, Paweł Hajdan, Jr., ser...@chromium.org, Sadrul Chowdhury, infr...@chromium.org
It's awesome to have this feature in the UI so you can git cl try from the webpage. But I really don't want to get emails for every CL that I am cc'd/reviewer'd on every time someone tries try bots.

Ravi Mistry

unread,
Apr 10, 2015, 5:47:37 PM4/10/15
to Dana Jansens, chromium-dev, Paweł Hajdan, Jr., ser...@chromium.org, Sadrul Chowdhury, infra-dev, jrob...@chromium.org
On Fri, Apr 10, 2015 at 5:27 PM, Dana Jansens <dan...@chromium.org> wrote:
It's awesome to have this feature in the UI so you can git cl try from the webpage. But I really don't want to get emails for every CL that I am cc'd/reviewer'd on every time someone tries try bots.

I agree that this causes spam for the cc'ed not interested in dry run results.
This is tricky though, because the emails are sent by the CQ, it is difficult to filter who gets the email.
If a CL non-owner requested the dry run then presumably everybody cc'ed is interested in the dry run results. If the CL owner requested it then maybe only the owner is interested. Thoughts?


 
To unsubscribe from this group and stop receiving emails from it, send an email to infra-dev+...@chromium.org.

To post to this group, send email to infr...@chromium.org.

Ilya Sherman

unread,
Apr 10, 2015, 5:54:52 PM4/10/15
to Ravi Mistry, Dana Jansens, chromium-dev, Paweł Hajdan, Jr., ser...@chromium.org, Sadrul Chowdhury, infra-dev, jrob...@chromium.org
On Fri, Apr 10, 2015 at 2:47 PM, Ravi Mistry <rmi...@chromium.org> wrote:
On Fri, Apr 10, 2015 at 5:27 PM, Dana Jansens <dan...@chromium.org> wrote:
It's awesome to have this feature in the UI so you can git cl try from the webpage. But I really don't want to get emails for every CL that I am cc'd/reviewer'd on every time someone tries try bots.

I agree that this causes spam for the cc'ed not interested in dry run results.
This is tricky though, because the emails are sent by the CQ, it is difficult to filter who gets the email.
If a CL non-owner requested the dry run then presumably everybody cc'ed is interested in the dry run results. If the CL owner requested it then maybe only the owner is interested. Thoughts?

IMO at most the CL owner and the person who initiated the dry run should get an email.

Dana Jansens

unread,
Apr 10, 2015, 5:54:58 PM4/10/15
to Ravi Mistry, chromium-dev, Paweł Hajdan, Jr., sergiyb, Sadrul Chowdhury, infra-dev, Jason Robbins
On Fri, Apr 10, 2015 at 2:47 PM, Ravi Mistry <rmi...@chromium.org> wrote:


On Fri, Apr 10, 2015 at 5:27 PM, Dana Jansens <dan...@chromium.org> wrote:
It's awesome to have this feature in the UI so you can git cl try from the webpage. But I really don't want to get emails for every CL that I am cc'd/reviewer'd on every time someone tries try bots.

I agree that this causes spam for the cc'ed not interested in dry run results.
This is tricky though, because the emails are sent by the CQ, it is difficult to filter who gets the email.
If a CL non-owner requested the dry run then presumably everybody cc'ed is interested in the dry run results. If the CL owner requested it then maybe only the owner is interested. Thoughts?

IMO no emails ever would be great. Sending them just to the person who initiated it would be reasonable, though I'd like to opt out or get them in a separate thread so my inbox can filter it.

I'm not sure why who hit the button other than me would change if I want to get an email or not. But then I never want the email (especially in the same thread).

Ravi Mistry

unread,
Apr 10, 2015, 6:12:41 PM4/10/15
to Ilya Sherman, Dana Jansens, chromium-dev, Paweł Hajdan, Jr., Sergiy Byelozyorov, Sadrul Chowdhury, infra-dev, jrob...@chromium.org
On Fri, Apr 10, 2015 at 5:54 PM, Ilya Sherman <ishe...@chromium.org> wrote:
On Fri, Apr 10, 2015 at 2:47 PM, Ravi Mistry <rmi...@chromium.org> wrote:
On Fri, Apr 10, 2015 at 5:27 PM, Dana Jansens <dan...@chromium.org> wrote:
It's awesome to have this feature in the UI so you can git cl try from the webpage. But I really don't want to get emails for every CL that I am cc'd/reviewer'd on every time someone tries try bots.

I agree that this causes spam for the cc'ed not interested in dry run results.
This is tricky though, because the emails are sent by the CQ, it is difficult to filter who gets the email.
If a CL non-owner requested the dry run then presumably everybody cc'ed is interested in the dry run results. If the CL owner requested it then maybe only the owner is interested. Thoughts?

IMO at most the CL owner and the person who initiated the dry run should get an email.

Sending it to the CL owner and the person who requested the dry run makes sense to me. I will investigate how to make this happen (if it is possible without significant changes to the emailing system in Rietveld).
 

Gabriel Charette

unread,
Apr 13, 2015, 12:51:54 PM4/13/15
to rmi...@chromium.org, Ilya Sherman, Dana Jansens, chromium-dev, Paweł Hajdan, Jr., Sergiy Byelozyorov, Sadrul Chowdhury, infra-dev, jrob...@chromium.org
If sending no emails at all is simpler to implement, I think we should go with that. i.e., what I really want is a git cl try that is a true "pre-CQ check" (i.e., if it's green, then real CQ has nothing left to do but commit when ticked), but I don't really want emails (even as the CL owner) from this early check.

Greg Thompson

unread,
Apr 13, 2015, 2:27:07 PM4/13/15
to Gabriel Charette, jrob...@chromium.org, ishe...@chromium.org, Sadrul Chowdhury, infra-dev, Sergiy Byelozyorov, Paweł Hajdan, Jr., rmi...@chromium.org, chromium-dev, Dana Jansens

I'm with Gab here. No mail would be good for me.

Ravi

unread,
Apr 14, 2015, 7:17:59 AM4/14/15
to chromi...@chromium.org, g...@chromium.org, jrob...@chromium.org, ishe...@chromium.org, sad...@google.com, infr...@chromium.org, ser...@chromium.org, phajd...@chromium.org, rmi...@chromium.org, dan...@chromium.org


On Monday, April 13, 2015 at 2:27:45 PM UTC-4, Greg Thompson wrote:

I'm with Gab here. No mail would be good for me.

On Apr 13, 2015 6:51 PM, "Gabriel Charette" <g...@chromium.org> wrote:
If sending no emails at all is simpler to implement, I think we should go with that. i.e., what I really want is a git cl try that is a true "pre-CQ check" (i.e., if it's green, then real CQ has nothing left to do but commit when ticked), but I don't really want emails (even as the CL owner) from this early check.

I came up with an approach that will allow emails to be sent to only the dry run initiator and the CL owner, I am currently working on the implementation.

I am a little surprised to hear that no emails are ok, I have been using the feature often and like the ability to trigger it from the command line and get email updates without me having to keep checking the issue in Rietveld.
I suspect no emails were preferred only compared to the current behavior of spamming everyone that is CC'ed and that the ideal approach is emailing the initiator and the owner, which is what I am working on. Let me know if you disagree here.

Thanks!

 

--
--
Chromium Developers mailing list: chromi...@chromium.org
View archives, change email options, or unsubscribe:
http://groups.google.com/a/chromium.org/group/chromium-dev

--
You received this message because you are subscribed to the Google Groups "infra-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to infra-dev+unsubscribe@chromium.org.
To post to this group, send email to infr...@chromium.org.

--
--
Chromium Developers mailing list: chromi...@chromium.org
View archives, change email options, or unsubscribe:
http://groups.google.com/a/chromium.org/group/chromium-dev

--
You received this message because you are subscribed to the Google Groups "infra-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to infra-dev+unsubscribe@chromium.org.
To post to this group, send email to infr...@chromium.org.
Reply all
Reply to author
Forward
0 new messages