Phil Ringnalda (:philor) Assignee
	Description • 13 years ago

From bug 733663, the mercilessly spamming part:

[[[
As a result, I got 13 rapid-fire emails from Try in the span of ~6 minutes, all for the same platform.

Each email I received looked exactly like this:
=========
Your Try Server build (2a9660ea4911) had unknown problem (5) on builder try-win32-debug.

The full log for this build run is available at
]]]

There are two problems with that:

* this was a push without -e, and although trychooser carefully calls the default "email only on non-success" a RETRY shouldn't be considered either a success or a non-success - it's not a result, that's just an artifact of the way that the job is going to be split across two (or more) logs, and the first chunk has to say something about its "result" even though there wasn't one. The only time we should be emailing about RETRY is when someone has explicitly opted in to massive emailing with -e.

* "had unknown problem (5)" is a less-than-forthcoming way of describing it - "is being automatically retried (5)" would be a much better description

	Phil Ringnalda (:philor) Assignee
	Comment 1 • 13 years ago

I'm fairly sure it's safe to not mail on retry for the default, because the only people who should be able to force a retry with a push are people messing about with talos causing an "ERROR: We tried to download the talos.json file but something failed", and I hope we can afford to risk them looping until they look at tbpl.

	Phil Ringnalda (:philor) Assignee
	Comment 2 • 13 years ago

Probably go better if RETRY is defined, won't it?

	Lukas Blakk [:lsblakk] use ?needinfo
	Comment 4 • 12 years ago

Comment on attachment 653095 [details] [diff] [review]
fix v.2 unrotted

Moving the r? over to catlee so someone more au-courant can review.

	Chris AtLee [:catlee]
	Comment 5 • 12 years ago

Comment on attachment 653095 [details] [diff] [review]
fix v.2 unrotted

Review of attachment 653095 [details] [diff] [review]:
-----------------------------------------------------------------

lgtm...what could go wrong?

	Phil Ringnalda (:philor) Assignee
	Comment 6 • 12 years ago

An excellent question: I don't even know whether it deploys automatically, deploys on the next reconfig, or doesn't deploy until someone finds the instructions and does it.

http://hg.mozilla.org/build/buildbotcustom/rev/4b674b64a6db

	Chris AtLee [:catlee]
	Comment 7 • 12 years ago

It will get deployed on the next reconfig

	Phil Ringnalda (:philor) Assignee
	Comment 8 • 12 years ago

"Your Try Server test (fc2e8bbc64ef) is being automatically retried on builder try_tegra_android_test-remote-trobocheck3."