|
|
Chromium Code Reviews
DescriptionSet linux_chromium_headless_rel CQ experiment at 100%
BUG=632008
Committed: https://crrev.com/70747375bb3bd1f015faace785dd7a2fa93f0097
Cr-Commit-Position: refs/heads/master@{#440426}
Patch Set 1 #
Total comments: 2
Messages
Total messages: 15 (7 generated)
perezju@chromium.org changed reviewers: + phajdan.jr@chromium.org, skyostil@chromium.org
Non-owner lgtm with a question. https://codereview.chromium.org/2595523002/diff/1/infra/config/cq.cfg File infra/config/cq.cfg (right): https://codereview.chromium.org/2595523002/diff/1/infra/config/cq.cfg#newcode57 infra/config/cq.cfg:57: experiment_percentage: 100 Is this equivalent to just removing experiment_percentage?
I'm seeing pending queues on this new builder with 10% load already. Should more capacity be added first? It should be possible to estimate based on 10% experiment load. Also, update the CL description - it's _rel now, not _dbg. https://codereview.chromium.org/2595523002/diff/1/infra/config/cq.cfg File infra/config/cq.cfg (right): https://codereview.chromium.org/2595523002/diff/1/infra/config/cq.cfg#newcode57 infra/config/cq.cfg:57: experiment_percentage: 100 On 2016/12/20 14:52:09, Sami wrote: > Is this equivalent to just removing experiment_percentage? Nope. A 100% experiment is still an experiment. It doesn't block CQ.
Description was changed from ========== Set linux_chromium_headless_dbg CQ experiment at 100% BUG=632008 ========== to ========== Set linux_chromium_headless_dbg CQ experiment at 100% BUG=632008 ==========
Description was changed from ========== Set linux_chromium_headless_dbg CQ experiment at 100% BUG=632008 ========== to ========== Set linux_chromium_headless_rel CQ experiment at 100% BUG=632008 ==========
On 2016/12/20 16:19:52, Paweł Hajdan Jr. wrote: > I'm seeing pending queues on this new builder with 10% load already. Should more > capacity be added first? It should be possible to estimate based on 10% > experiment load. I'm trying to figure out the amount of hardware to add. From stip's comment: "We usually get about 40 builds per hour peak[1] for CQ bots, with occasional spikes. So let's plan for 50 builds per hour. The mean of this bot bounces around a bit[2], but let's say 30 minute mean time for the bot. Targeting 99% of builds to queue less than 10 minutes, we plug these numbers into Erlang-C [3] to get 33 machines. Let's round up to 35." https://bugs.chromium.org/p/chromium/issues/detail?id=633985#c9 All of those constraints still seem to hold up, namely: [1] 40 builds per hour peak http://shortn/_3QSeeGvdy0 [2] 30 minutes min time https://viceroy.corp.google.com/chrome_infra/Buildbot/overview_v2?duration=30... Why do you think we sometimes have those unusual spikes of pending builds? How much additional hardware should we add?
Pending builds might have been an effect of https://bugs.chromium.org/p/chromium/issues/detail?id=669297#c85 . Let's try this. LGTM.
The CQ bit was checked by perezju@chromium.org
CQ is trying da patch. Follow status at https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...
CQ is committing da patch.
Bot data: {"patchset_id": 1, "attempt_start_ts": 1482420508282990, "parent_rev":
"9982464e533a4670dd74b8e5a792f74096127b58", "commit_rev":
"1ba42b4ba7ea4163c0a65b274026481491be53ea"}
Message was sent while issue was closed.
Description was changed from ========== Set linux_chromium_headless_rel CQ experiment at 100% BUG=632008 ========== to ========== Set linux_chromium_headless_rel CQ experiment at 100% BUG=632008 Review-Url: https://codereview.chromium.org/2595523002 ==========
Message was sent while issue was closed.
Committed patchset #1 (id:1)
Message was sent while issue was closed.
Description was changed from ========== Set linux_chromium_headless_rel CQ experiment at 100% BUG=632008 Review-Url: https://codereview.chromium.org/2595523002 ========== to ========== Set linux_chromium_headless_rel CQ experiment at 100% BUG=632008 Committed: https://crrev.com/70747375bb3bd1f015faace785dd7a2fa93f0097 Cr-Commit-Position: refs/heads/master@{#440426} ==========
Message was sent while issue was closed.
Patchset 1 (id:??) landed as https://crrev.com/70747375bb3bd1f015faace785dd7a2fa93f0097 Cr-Commit-Position: refs/heads/master@{#440426} |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
