412545 - more (possibly spurious) random orange on bm-xserve11

Status: RESOLVED WORKSFORME

(Core :: General, defect, P1)

Description

[Dan Mosedale (:dmosedale, :dmose)]

This is apparently unrelated to bug 412392.  Link to the log is in the URL field of this bug.  

The error message of interest:

Error: Leak Test Failed: Number of leaks 64 is greater than LeakFailureThreshold 

Rlk on previous and subsequent cycles was 0.

Potentially interesting chunk of the log:

--NEW-LEAKS-----------------------------------leaks------leaks%-----------------------
nsBaseAppShell                                   52          -
nsRunnable                                       12          -
TOTAL                                            64
--NEW-BLOAT-----------------------------------bloat------bloat%-----------------------
nsThread                                        504      40.00%
xpcPerThreadData                                392      40.00%
nsProxyEventObject                              512      14.29%
nsProxyObject                                   896      14.29%
nsStandardURL                                594000       1.32%
nsFrameConstructorState                       21560       1.03%
nsDeque                                      341016       0.74%
nsRunnable                                    17916       0.61%
nsDisplayXULImage                              5772       0.42%
nsDisplayClip                                 21708       0.33%
nsStringBuffer                               260136       0.11%
StringAdopt                                    2186       0.05%
nsHashKey                                    119136       0.04%
nsCStringKey                                  59320       0.03%
nsVoidArray                                   87772       0.02%
TOTAL                                       1532826

Comment 1

[Mike Schroepfer]

Moving TM to b4 - Dmose can you suggest an owner for this?

Comment 2

[Mike Schroepfer]

Dmose?

Comment 3

[Dan Mosedale (:dmosedale, :dmose)]

Unfortunately, I can't.  The problem, as I recall, is that it seemed very very unlikely that any of the checkins (if there were any) around the time of this orangeness caused this (tinderbox no longer indexes these, but they could be found with a sufficiently precise bonsai query).  A bit of archaeology shows no occurrences of the orangeness in the 4 days or so previous to this checkin, and no occurrences in the 4 days previous to now.  So it's perhaps reasonable to chalk this up to phase-of-the-moon and close as WORKSFORME.

Comment 4

[Damon Sicore (:damons)]

Are we still seeing this?

Comment 5

[Damon Sicore (:damons)]

If I don't hear more on this today, I'm going to take this off the blocker list.

Comment 6

[Damon Sicore (:damons)]

-'ing this as no one seems to care.

Comment 7

[Chris Cooper [:coop] (he/him)]

bm-xserve11 was re-imaged last year.