New issue
Advanced search Search tips
Note: Color blocks (like or ) mean that a user may not be available. Tooltip shows the reason.

Issue 593404 link

Starred by 2 users

Issue metadata

Status: WontFix
Owner:
Last visit > 30 days ago
Closed: Apr 2016
Cc:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 1
Type: Bug



Sign in to add a comment

jetstream_GuestFirewall failed: firewall rules do not match expectations

Project Member Reported by shuqianz@chromium.org, Mar 9 2016

Issue description

Is this a test flake?
https://uberchromegw.corp.google.com/i/chromeos/builders/whirlwind-paladin/builds/3942/steps/HWTest%20%5Bjetstream_cq%5D/logs/stdio

3/09 09:49:30.171 DEBUG|        base_utils:0177| Running '/usr/bin/ssh -a -x   -N -o ControlMaster=yes -o ControlPath=/tmp/_autotmp_P52mVtssh-master/socket -o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null -o BatchMode=yes -o ConnectTimeout=30 -o ServerAliveInterval=900 -o ServerAliveCountMax=3 -o ConnectionAttempts=4 -o Protocol=2 -l root -p 22 chromeos4-row10-rack12-host7'
03/09 09:49:31.397 INFO | site_crashcollect:0034| symbol_dir: /usr/local/autotest/server/../../../lib/debug
03/09 09:49:31.398 DEBUG|        base_utils:0177| Running 'minidump_stackwalk "/usr/local/autotest/results/56229440-chromeos-test/chromeos4-row10-rack12-host7/jetstream_GuestFirewall/sysinfo/iteration.1/var/spool/crash/ap_update_manager.20160303.233820.1172.dmp" "/usr/local/autotest/server/../../../lib/debug" > "/usr/local/autotest/results/56229440-chromeos-test/chromeos4-row10-rack12-host7/jetstream_GuestFirewall/sysinfo/iteration.1/var/spool/crash/ap_update_manager.20160303.233820.1172.dmp.txt"'
03/09 09:49:31.446 WARNI| site_crashcollect:0103| Failed to generate stack trace locally for dump /usr/local/autotest/results/56229440-chromeos-test/chromeos4-row10-rack12-host7/jetstream_GuestFirewall/sysinfo/iteration.1/var/spool/crash/ap_update_manager.20160303.233820.1172.dmp (rc=127):
CmdError('minidump_stackwalk "/usr/local/autotest/results/56229440-chromeos-test/chromeos4-row10-rack12-host7/jetstream_GuestFirewall/sysinfo/iteration.1/var/spool/crash/ap_update_manager.20160303.233820.1172.dmp" "/usr/local/autotest/server/../../../lib/debug" > "/usr/local/autotest/results/56229440-chromeos-test/chromeos4-row10-rack12-host7/jetstream_GuestFirewall/sysinfo/iteration.1/var/spool/crash/ap_update_manager.20160303.233820.1172.dmp.txt"', * Command: 
    minidump_stackwalk "/usr/local/autotest/results/56229440-chromeos-test/ch
    romeos4-row10-rack12-host7/jetstream_GuestFirewall/sysinfo/iteration.1/var
    /spool/crash/ap_update_manager.20160303.233820.1172.dmp"
    "/usr/local/autotest/server/../../../lib/debug" >
    "/usr/local/autotest/results/56229440-chromeos-test/chromeos4-row10-rack12
    -host7/jetstream_GuestFirewall/sysinfo/iteration.1/var/spool/crash/ap_upda
    te_manager.20160303.233820.1172.dmp.txt"
Exit status: 127
Duration: 0.00887393951416

stderr:
/bin/bash: minidump_stackwalk: command not found, 'Command returned non-zero exit status')
03/09 09:49:31.451 INFO | site_crashcollect:0107| Generating stack trace for /usr/local/autotest/results/56229440-chromeos-test/chromeos4-row10-rack12-host7/jetstream_GuestFirewall/sysinfo/iteration.1/var/spool/crash/ap_update_manager.20160303.233820.1172.dmp
03/09 09:49:31.655 DEBUG|        dev_server:0312| Apache client count is not collected from devserver.
03/09 09:49:31.656 DEBUG|        dev_server:0312| Apache client count is not collected from devserver.
03/09 09:49:31.682 INFO |    connectionpool:0190| Starting new HTTP connection (1): 100.107.160.6
 

Comment 1 by cbook@google.com, Mar 9 2016

This was a legit test catch.  I updated a path location but missed updating the same path in the test case.

Comment 2 by cbook@google.com, Mar 9 2016

Status: WontFix (was: Untriaged)
closing (is that right?) as this is not a bug in the test.
thanks!
Status: Available (was: WontFix)
Wait, so you will have a fix to update the path in the test, right? 

Comment 5 by cbook@google.com, Mar 9 2016

Yes, my patch has been updated to include the autotest path:

https://chrome-internal-review.googlesource.com/#/c/250605
So, we can expect this test will pass next time in the cq, right?

Comment 7 by cbook@google.com, Mar 9 2016

Yes, I ran the test and it passes locally now.

Comment 8 by cbook@google.com, Mar 13 2016

Status: WontFix (was: Available)
this was resolved.
Cc: shchen@chromium.org adlr@chromium.org

Comment 12 by cbook@google.com, Apr 6 2016

Not sure why this bug was even raised in the first place.  A change was added to the CQ and rejected (properly) because it failed the test.  The change was fixed and then later passed the test and then landed.  Why would a bug be created for a legit CQ failure catch?

Not sure about these new failures?  Is another new change in the CQ breaking?

Comment 13 by cbook@google.com, Apr 6 2016

Looks like this change may be the culprit of the CQ failure:
https://chrome-internal-review.googlesource.com/#/c/252996/
Cc: col...@chromium.org pauldev@chromium.org
consecutive CQ failure due to 'minidump_stackwalk', not sure what's going on,

whirlwind-paladin: The HWTest [jetstream_cq] stage failed:
https://uberchromegw.corp.google.com/i/chromeos/builders/master-paladin/builds/10717

log: 
https://pantheon.corp.google.com/storage/browser/chromeos-autotest-results/59143184-chromeos-test/chromeos4-row10-rack12-host2/debug/

a debug log is attached. 

Agree that maybe a CL causes this failure.

ļ¼ pauldev, @colish, Could you take a look at it?
59143184-chromeos-test-chromeos4-row10-rack12-host2-debug-autoserv.txt
166 KB View Download
I think this one is mine. trying to add a seccomp filter for miniupnpd.

My change was readded to the commit queue this morning as we believed it was a flaky test, but I've since found where the problem lies. We've removed it from the commit queue at this stage, pending a much higher confidence that this wouldn't happen again.
Thanks! 

and thanks for your help @cbook.
CQ is broken again because of the same reason, seems not this CL?
I haven't added back to the recommit queue. Unless there is something I'm missing to remove it out (Took the +1 on commit queue away), it shouldn't be this CL.

Comment 19 by cbook@google.com, Apr 6 2016

Do you have a link?  What is the list of changes for this new CQ failure?
A bad cdd got past cq, my fix is https://chrome-internal-review.googlesource.com/#/c/253670/.
Cc: cjgrant@chromium.org
The other bad cdd CL fix is https://chrome-internal-review.googlesource.com/#/c/253681/
I just chumped my fix to unblock.

Comment 23 by cbook@google.com, Apr 18 2016

Status: WontFix (was: Untriaged)
closing

Sign in to add a comment