New issue
Advanced search Search tips

Issue 758256 link

Starred by 3 users

Issue metadata

Status: Fixed
Owner:
Closed: Jul 19
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Fuchsia
Pri: 2
Type: Bug

Blocked on:
issue 864164



Sign in to add a comment

Optional/Experimental Fuchsia Debug bot for CQ

Project Member Reported by scottmg@chromium.org, Aug 23 2017

Issue description

Now that we've got "Fuchsia Compile" on the waterfall (thanks John! https://bugs.chromium.org/p/chromium/issues/detail?id=745086), we of course managed to have someone break Fuchsia Debug :)

Is it easier to add the Debug bot (in particular with is_component_build=true) now that we have the Release one? I guess we still have to go through the staged load-testing, etc?

This isn't as high priority as Release (as we don't ship it of course) but it would be nice to keep it working.
 
Owner: scottmg@chromium.org
Status: Started (was: Available)
Summary: Fuchsia bots for waterfall/cq (was: Fuchsia Compile (dbg) for waterfall/cq)
Retitling/consolidating:
- ARM64 compile on main cq/waterfall: https://bugs.chromium.org/p/chromium/issues/detail?id=771841
- ARM64 fyi https://bugs.chromium.org/p/chromium/issues/detail?id=774313

I'm going to rename "Fuchsia Compile" to "Fuchsia x64" as we have ARM now, and we'll be adding test running to the bots waterfall bots.

We should also add debug compile bots per the original bug here, but for now focus on release-with-dchecks.
Rename SGTM.

Comment 3 by w...@chromium.org, Oct 16 2017

Labels: M-64
Project Member

Comment 4 by bugdroid1@chromium.org, Oct 18 2017

Project Member

Comment 5 by bugdroid1@chromium.org, Oct 18 2017

Project Member

Comment 6 by bugdroid1@chromium.org, Oct 18 2017

The following revision refers to this bug:
  https://chrome-internal.googlesource.com/infradata/master-manager/+/82018a8b03b1fc3607864b780a06c3fd1d5a3590

commit 82018a8b03b1fc3607864b780a06c3fd1d5a3590
Author: Dirk Pranke <dpranke@google.com>
Date: Wed Oct 18 03:58:39 2017

Project Member

Comment 7 by bugdroid1@chromium.org, Oct 19 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/fc2b2f04ed20d88113c10951036ca403891c435f

commit fc2b2f04ed20d88113c10951036ca403891c435f
Author: Scott Graham <scottmg@chromium.org>
Date: Thu Oct 19 17:14:10 2017

fuchsia: Remove fuchsia_compile from CQ temporarily

Step 1 of ye ol' 7 step plan:
https://chromium-review.googlesource.com/c/chromium/src/+/726314#message-4899c168cc8c4d3ec18d1442a7697fe9645ee185

Bug:  758256 
Change-Id: I1002527dae80f0af3ac670914bf0c8b4d1dab6d9
Reviewed-on: https://chromium-review.googlesource.com/727140
Reviewed-by: Dirk Pranke <dpranke@chromium.org>
Commit-Queue: Scott Graham <scottmg@chromium.org>
Cr-Commit-Position: refs/heads/master@{#510115}
[modify] https://crrev.com/fc2b2f04ed20d88113c10951036ca403891c435f/infra/config/cq.cfg

Project Member

Comment 9 by bugdroid1@chromium.org, Oct 20 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/0bdc4ab790ec470e5bd54022caa15e0d39a55b37

commit 0bdc4ab790ec470e5bd54022caa15e0d39a55b37
Author: Scott Graham <scottmg@chromium.org>
Date: Fri Oct 20 04:08:06 2017

Add note on how to upload buildbucket.config changes

Bug:  758256 
Change-Id: I5e7874eefa3265acc6bfc1e14f18a48a87ef6170
Reviewed-on: https://chromium-review.googlesource.com/729203
Reviewed-by: Dirk Pranke <dpranke@chromium.org>
Commit-Queue: Scott Graham <scottmg@chromium.org>
Cr-Commit-Position: refs/heads/master@{#510330}
[modify] https://crrev.com/0bdc4ab790ec470e5bd54022caa15e0d39a55b37/tools/mb/docs/user_guide.md

Project Member

Comment 10 by bugdroid1@chromium.org, Oct 20 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/tools/build/+/14fe557860738d310becbf8dc06a1da98861d0c2

commit 14fe557860738d310becbf8dc06a1da98861d0c2
Author: Scott Graham <scottmg@chromium.org>
Date: Fri Oct 20 15:09:48 2017

fuchsia: Move fuchsia_x64 into cq pool

Step 4 of revised 7 step program. https://chromium-review.googlesource.com/c/chromium/src/+/726314#message-4899c168cc8c4d3ec18d1442a7697fe9645ee185

Bug:  758256 
Change-Id: I6f96d842f895651f3c9ab4c0715f4303b48609cc
Reviewed-on: https://chromium-review.googlesource.com/728865
Reviewed-by: Dirk Pranke <dpranke@chromium.org>
Commit-Queue: Scott Graham <scottmg@chromium.org>

[modify] https://crrev.com/14fe557860738d310becbf8dc06a1da98861d0c2/masters/master.tryserver.chromium.linux/slaves.cfg

Project Member

Comment 11 by bugdroid1@chromium.org, Oct 21 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/573b6c39d3de239a70e8fa672647b63bf0bd1f89

commit 573b6c39d3de239a70e8fa672647b63bf0bd1f89
Author: Scott Graham <scottmg@chromium.org>
Date: Sat Oct 21 03:33:01 2017

fuchsia: Renaming src side of fuchsia bots, part 5

Step 5 of revised 7 step program.
https://chromium-review.googlesource.com/c/chromium/src/+/726314#message-4899c168cc8c4d3ec18d1442a7697fe9645ee185

Bug:  758256 
Change-Id: Ib9ab90ee2038a2af72fe29a62042e55feb8fcecc
Reviewed-on: https://chromium-review.googlesource.com/728415
Reviewed-by: Dirk Pranke <dpranke@chromium.org>
Commit-Queue: Scott Graham <scottmg@chromium.org>
Cr-Commit-Position: refs/heads/master@{#510655}
[modify] https://crrev.com/573b6c39d3de239a70e8fa672647b63bf0bd1f89/infra/config/cq.cfg
[modify] https://crrev.com/573b6c39d3de239a70e8fa672647b63bf0bd1f89/tools/mb/mb_config.pyl

Project Member

Comment 12 by bugdroid1@chromium.org, Oct 21 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/4cdadc77f6092cb29561a3d38182e4b927041c1d

commit 4cdadc77f6092cb29561a3d38182e4b927041c1d
Author: Scott Graham <scottmg@chromium.org>
Date: Sat Oct 21 04:39:37 2017

fuchsia: put fyi name back to plain Fuchsia

NOTRY=true

Bug:  758256 
Change-Id: If8154e8773baf861760dea0728b5627a27374369
TBR: dpranke@chromium.org
Reviewed-on: https://chromium-review.googlesource.com/731865
Commit-Queue: Scott Graham <scottmg@chromium.org>
Reviewed-by: Scott Graham <scottmg@chromium.org>
Cr-Commit-Position: refs/heads/master@{#510661}
[modify] https://crrev.com/4cdadc77f6092cb29561a3d38182e4b927041c1d/tools/mb/mb_config.pyl

Project Member

Comment 13 by bugdroid1@chromium.org, Oct 23 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/f776318ab9b8863704ed1758d774961ab3d2c7bb

commit f776318ab9b8863704ed1758d774961ab3d2c7bb
Author: Scott Graham <scottmg@chromium.org>
Date: Mon Oct 23 09:01:28 2017

fuchsia: Add a few small tests to main CQ/waterfall

Starting with just a few that haven't flaked and run in a couple
seconds to make sure we don't overload anything.

Bug:  707029 ,  758256 
Change-Id: If287d719ed966cd3f251d47e72bf386067dac220
Reviewed-on: https://chromium-review.googlesource.com/726314
Reviewed-by: Dirk Pranke <dpranke@chromium.org>
Commit-Queue: Scott Graham <scottmg@chromium.org>
Cr-Commit-Position: refs/heads/master@{#510741}
[modify] https://crrev.com/f776318ab9b8863704ed1758d774961ab3d2c7bb/testing/buildbot/chromium.linux.json

Comment 14 by w...@chromium.org, Dec 2 2017

jbudorick: Now that we have Debug building, can we get it added to the waterfall/CQ?
We are somewhat capacity constrained, and even without that it really only makes sense to have bots in the CQ if they are going to frequently catch failures that won't be caught by other configs. We currently have fuchsia_{arm64,x64} in the CQ; would it make more sense for one of those to be debug and the other release? Or do you plan to run tests on both? Do we think that we'll see a large number of fuchsia-specific debug compile failures?

Comment 16 by w...@chromium.org, Dec 7 2017

Re #15: We regularly see Debug compilation regress on the FYI bot, mainly due to component-build issues creeping in. I think we'd prefer to keep the tests running on both ARM & x64, since there are various bits of the system (e.g. bits of ASM) that differ, so let's leave things as they are for now - I'll set up gatekeeper to ping us about Debug failures on the FYI in the meantime.

Comment 17 by w...@chromium.org, Dec 8 2017

Labels: -M-64 M-66
Owner: ----
Status: Available (was: Started)
Summary: Fuchsia Debug bot for waterfall/cq (was: Fuchsia bots for waterfall/cq)
Marking this Available for M66, since we don't have an immediate need for Debug on CQ/waterfall right this minute.
Labels: -M-66 M-69
Summary: Optional/Experimental Fuchsia Debug bot for CQ (was: Fuchsia Debug bot for waterfall/cq)
Bumping this to Want for M69; we'd like an optional/experimental Fuchsia/Debug/component bot on the CQ to help Chrome-Fuchsia team break ourselves less often. ;)
Status: Untriaged (was: Available)
Marking untriaged so we see it in our queue.
Owner: martiniss@chromium.org
Status: Assigned (was: Untriaged)
I'll work on this.
Owner: dpranke@chromium.org
Actually, I'm not sure if we have capacity.... 

well, Dirk's comment in #15 makes me a bit nervous? I glanced at the CQ bots that already exist, and it looks like they just use linux capacity, which I don't think we're that severely constrained on.

I'd imagine we'd need maybe ~30 bots to power the compile, and then maybe some extra bots in swarming to run tests? Not sure. Dirk, what was your reasoning behind #15?
Owner: martiniss@chromium.org
We aren't particularly capacity constrained at this point (unlike where we were in December). 

Apart from that, the rest of the comments should be pretty self-explanatory. 

Ok, I'll create the bot then. I'll name it fuchsia_debug_arm64. The naming convention seems to use underscores rather than dashes, any reason why? 
Or, should it be x64? I think based on comments at the beginning of the bug it should be? So it'd be "fuchsia_debug_x64".
Yes, fuchsia_debug_x64.  We should start with that bot in a compile-only
configuration, IMO.
I think the "_" in names are from the builders that were set up before we switched to trying to use the bit.ly/chromium-build-naming conventions.

I really need to beef that doc up to be more explicit about something, and we're still working out the conventions, but you should use:

fuchsia-arm64-dbg

which is roughly $OS[-$ARCH][-$FEATURES]-$BUILD_CONFIG

where $TARGET is one of { android, chromeos, fuchsia, ios, linux, mac, win } and $BUILD_CONFIG is one of { dbg, rel } (for debug and release) for now (and will probably eventually grow { dcheck, perf, official }.


Does chromium-build-naming have a stated preference on trybot names being the same as the trybots they match?
Owner: dpranke@chromium.org
Dirk, can you comment on #27? The trybot would mirror 'fuchsia-fyi-x64-dbg' on chromium.fyi, which has a slightly different naming convention.
Owner: martiniss@chromium.org
Ideally we'd rename the CI fyi bot to match the "preferred" convention, but since that's still something of a work in progress, I'd leave the name alone for now.

The trybot should mirror the CI bot's name, i.e., I think matching the name should probably override matching the convention.
I need to add those bits to the doc as well; I'll try to do so if I get a quiet few minutes during my current sheriffing stint, and if I don't, then I'll add them ASAP afterwards.
in that case, preferred would be fuchsia-x64-fyi-dbg?

I recently renamed these to undupe names w/ non-fyi bots, hence the fyi infix. Chose the ordering because lexicographically grouping fyi before grouping arch seemed desirable.

In theory, yes, it'd be "fuchsia-x64-fyi-dbg" according to what I wrote above.

Really I think what we need to do is rename a large number of builders according to one scheme or another and see which we like in aggregate. Doing things piecemeal makes it harder to assess the readability.

Separate from that, though, this particular builder also trips over the current open issue in the the bit.ly/chromium-build-naming doc, i.e., how do we distinguish between otherwise similar builders on different masters if we get rid of the mastername concept? I'm not sure if using "fyi" for this as part of the name is the best possible answer, but it might be at least acceptably good.
Do I name the trybot 'fuchsia-x64-fyi-dbg'? I thought I would leave out the fyi, and name it 'fuchsia-x64-dbg', since it's not on an fyi waterfall.

Although I'd imagine we don't like the idea of having non-fyi trybots mirror fyi bots. Sadly that seems to already be happening in places 
There's no real concept of an "fyi trybot", it's just an optional trybot that may mirror a bot on the main waterfall or on the fyi waterfall. I'd keep the name the same. 
Project Member

Comment 35 by bugdroid1@chromium.org, Jul 12

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/tools/build/+/73db86643d71a094e3575ee93644c14c21e9b0bf

commit 73db86643d71a094e3575ee93644c14c21e9b0bf
Author: Stephen Martinis <martiniss@chromium.org>
Date: Thu Jul 12 23:28:44 2018

Add new fuchsia debug trybot

Bug:  758256 
Change-Id: Id4f88e1b026f257c6b17b65a013ae530f9a9f599
Reviewed-on: https://chromium-review.googlesource.com/1130545
Reviewed-by: John Budorick <jbudorick@chromium.org>
Commit-Queue: Stephen Martinis <martiniss@chromium.org>

[modify] https://crrev.com/73db86643d71a094e3575ee93644c14c21e9b0bf/tests/masters_recipes_test.py
[modify] https://crrev.com/73db86643d71a094e3575ee93644c14c21e9b0bf/scripts/slave/recipe_modules/chromium_tests/trybots.py

Project Member

Comment 36 by bugdroid1@chromium.org, Jul 13

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/942ff8e3fd9b58f08158d6133e22b70dad6c59ce

commit 942ff8e3fd9b58f08158d6133e22b70dad6c59ce
Author: Stephen Martinis <martiniss@chromium.org>
Date: Fri Jul 13 01:17:11 2018

Add fuchsia debug trybot

Bug:  758256 
Change-Id: Icbd47e057881ec2dc2e2eb1cc733b5dd34326991
Reviewed-on: https://chromium-review.googlesource.com/1130499
Reviewed-by: John Budorick <jbudorick@chromium.org>
Reviewed-by: Ryan Tseng <hinoka@chromium.org>
Commit-Queue: Stephen Martinis <martiniss@chromium.org>
Cr-Commit-Position: refs/heads/master@{#574798}
[modify] https://crrev.com/942ff8e3fd9b58f08158d6133e22b70dad6c59ce/infra/config/global/cr-buildbucket.cfg
[modify] https://crrev.com/942ff8e3fd9b58f08158d6133e22b70dad6c59ce/infra/config/global/luci-milo.cfg
[modify] https://crrev.com/942ff8e3fd9b58f08158d6133e22b70dad6c59ce/tools/mb/mb_config.pyl

Just tried this out, and the bot failed with "internal failure".
I didn't add any actual machines to power this builder, so it doesn't work yet.
Blockedon: 864164
Project Member

Comment 40 by bugdroid1@chromium.org, Jul 17

The following revision refers to this bug:
  https://chrome-internal.googlesource.com/infradata/config/+/76d6be31c9d12ecbafa385822ec1826183ff081e

commit 76d6be31c9d12ecbafa385822ec1826183ff081e
Author: Stephen Martinis <martiniss@google.com>
Date: Tue Jul 17 22:12:25 2018

Project Member

Comment 41 by bugdroid1@chromium.org, Jul 17

The following revision refers to this bug:
  https://chrome-internal.googlesource.com/infradata/config/+/5dffcb3ebaa2b1231c907cb6b0729cdc3e29f73a

commit 5dffcb3ebaa2b1231c907cb6b0729cdc3e29f73a
Author: Stephen Martinis <martiniss@google.com>
Date: Tue Jul 17 23:26:35 2018

Project Member

Comment 42 by bugdroid1@chromium.org, Jul 18

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/bf550a6168f769ae3f6e293a28a8918d1bd60022

commit bf550a6168f769ae3f6e293a28a8918d1bd60022
Author: Stephen Martinis <martiniss@chromium.org>
Date: Wed Jul 18 22:00:45 2018

Fix typo in mb_config.pyl

I typo-ed the builder name, mispelled fuchsia

Bug:  758256 
Change-Id: I75d94e0cd2b6ed13e67dbfe71ec9f85c229a75d0
Reviewed-on: https://chromium-review.googlesource.com/1142443
Reviewed-by: John Budorick <jbudorick@chromium.org>
Commit-Queue: Stephen Martinis <martiniss@chromium.org>
Cr-Commit-Position: refs/heads/master@{#576230}
[modify] https://crrev.com/bf550a6168f769ae3f6e293a28a8918d1bd60022/tools/mb/mb_config.pyl

Status: Fixed (was: Assigned)
The bot seems to be up and running; https://ci.chromium.org/p/chromium/builders/luci.chromium.try/fuchsia-fyi-x64-dbg/5 seems to be running well.

I think this is done? Feel free to re-open if you want something else done.

Sign in to add a comment