"Deterministic Linux" bot is having a bad time |
||||
Issue descriptionI was sheriffing today and spent a fair amount of time to investigate what's suddenly wrong with the "Deterministic Linux" bot. It has suddenly become very unstable, starting with https://build.chromium.org/p/chromium.linux/builders/Deterministic%20Linux/builds/3042 There are several bugs for it: * bug 727442 - Clang segfault * bug 727201 - First/second build fails * bug 727607 - Google storage download problems I filed the last one, which phajdan@ offered to help out with (making the recipe reuse the checkouts instead of recreating them from scratch). That may require a significant amount of work, so I'd like to get this properly tracked and prioritized here first.
,
May 31 2017
Other deterministic builders seem not be affected for this issue. https://build.chromium.org/p/chromium.fyi/waterfall?category=deterministic Since failure steps are not deterministic, I started to guess this is caused by machine trouble. Can we ask to change the machine?
,
May 31 2017
Issue 727442 has been merged into this issue.
,
May 31 2017
Issue 727201 has been merged into this issue.
,
May 31 2017
,
May 31 2017
I believe all of issues are the same reason with different aspects. "Deterministic Linux" is the builder to confirm deterministic build. It is provided as a part of crbug.com/439737. It does the same chromium build twice, and compare. If something differs, compare_build_artifacts step should fail. I do not expect other steps fail. Especially, second build failure while first build success sounds like infrastructure problem.
,
May 31 2017
,
Jun 7 2017
This seems to be green now. Can we close this bug or is there some followup work to be done?
,
Jun 7 2017
All I know is somebody restarted buildbot after it get stuck, then build become green. I do not think there is follow up work to be done. |
||||
►
Sign in to add a comment |
||||
Comment 1 by mar...@chromium.org
, May 30 2017