Make the builder failure tolerance threshold configurable through a feature

Bug #681426 reported by Julian Edwards
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
Launchpad itself
Triaged
Low
Unassigned

Bug Description

This is the only way we can poke new values on the fly without cowboying code/config. Until the current timeout problems are resolved we want to configure this threshold on the fly. A value of zero should disable automatic builder failing completely.

Changed in soyuz:
status: New → Triaged
importance: Undecided → High
status: Triaged → In Progress
tags: added: buildd-manager
Changed in soyuz:
assignee: nobody → Julian Edwards (julian-edwards)
Revision history for this message
Robert Collins (lifeless) wrote : Re: [Bug 681426] [NEW] Make the builder failure tolerance threshold configurable through a feature

0-or-missing for simplicity I think.That way you don't need any hard
coded default.

-Rob

Revision history for this message
Julian Edwards (julian-edwards) wrote :

On Thursday 25 November 2010 17:54:03 you wrote:
> 0-or-missing for simplicity I think.That way you don't need any hard
> coded default.
>
> -Rob

Well, I'd rather default to the hard-coded figure so that it DTRT with no
feature configuration needed. I thought the point of the feature config was
to tweak the defaults.

Revision history for this message
Robert Collins (lifeless) wrote :

On Fri, Nov 26, 2010 at 9:10 AM, Julian Edwards
<email address hidden> wrote:
> Well, I'd rather default to the hard-coded figure so that it DTRT with no
> feature configuration needed.  I thought the point of the feature config was
> to tweak the defaults.

Its to let us do things dynamically and in a relaxed fashion without
needing lockstep code + config changes.

The main reason I was saying no default was so the code could be a
little simpler, but its going to be pretty much a wash.

I doubt our getting the right value out of the gates though :)

Anyhow, all variations are fine.

-rob

Revision history for this message
Julian Edwards (julian-edwards) wrote :

On Thursday 25 November 2010 21:40:27 you wrote:
> I doubt our getting the right value out of the gates though :)

I'm quite confident of the default actually - once this other problem is
sorted (which I'll land a work around for tomorrow) we'll see a heck of lot
more stability anyway and given my observations of the logs files over the
last 2 or 3 years when a builder goes bong, it goes bong big stylee.

Changed in soyuz:
status: In Progress → Triaged
Revision history for this message
Launchpad QA Bot (lpqabot) wrote : Bug fixed by a commit
Changed in soyuz:
milestone: none → 10.12
tags: added: qa-needstesting
Changed in soyuz:
status: Triaged → Fix Committed
Revision history for this message
Julian Edwards (julian-edwards) wrote :

Hello? QA bot?

Changed in soyuz:
status: Fix Committed → Triaged
tags: removed: qa-needstesting
tags: added: qa-ok
Revision history for this message
Julian Edwards (julian-edwards) wrote :

Low priority now, the latest fixes to avoid timeouts are working well.

Changed in soyuz:
importance: High → Low
Revision history for this message
Ursula Junque (ursinha) wrote : Re: [Bug 681426] Re: Make the builder failure tolerance threshold configurable through a feature

On Mon, Nov 29, 2010 at 7:35 AM, Julian Edwards
<email address hidden> wrote:
> Hello?  QA bot?

Hi Julian, the bot should or shouldn't have touched this bug? Revno
11986 has this bug number on its branch nick:
timeouts-feature-bug-681426. Is that wrong somehow?

>
> ** Changed in: soyuz
>       Status: Fix Committed => Triaged
>
> --
> Make the builder failure tolerance threshold configurable through a feature
> https://bugs.launchpad.net/bugs/681426
> You received this bug notification because you are subscribed to
> Launchpad Suite.
>

--
Ursinha (Ursula Junque)
<email address hidden>
<email address hidden>
--
Ubuntu - I am because we are
--
Linux user #289453 - Ubuntu user #31144

Revision history for this message
Julian Edwards (julian-edwards) wrote :

On Monday 06 December 2010 19:12:56 you wrote:
> On Mon, Nov 29, 2010 at 7:35 AM, Julian Edwards
>
> <email address hidden> wrote:
> > Hello? QA bot?
>
> Hi Julian, the bot should or shouldn't have touched this bug? Revno
> 11986 has this bug number on its branch nick:
> timeouts-feature-bug-681426. Is that wrong somehow?

Yeah I was confused because I hadn't linked the branch - then I realised that
I'd stupidly put the wrong bug number on the branch nick!

I think that it might be better to let people explicitly link branches but
that's just me. Worth asking around?

Thanks!

Changed in soyuz:
milestone: 10.12 → none
William Grant (wgrant)
tags: removed: qa-ok
William Grant (wgrant)
Changed in launchpad:
assignee: Julian Edwards (julian-edwards) → nobody
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.