Ticket #11708 (new defect)

Opened 2 years ago

Last modified 2 years ago

sugar shell memory leak

Reported by: pgf Owned by:
Priority: normal Milestone: 11.3.1
Component: not assigned Version: Development build as of this date
Keywords: Cc: martin.langhoff
Action Needed: diagnose Verified: no
Deployments affected: Blocked By:
Blocking:

Description

richard and i have both observed machines which, when allowed to sit for long periods, have huge python sugar processes.

the attached logs are from a machine where the sugar shell was taking 41% of memory (385M virtual, 347M resident). it was not suspending/resuming. it was associated to the "olpc" network at twine, and so there were both a lot of laptops in view, and a lot of networks in view.

build is os30.

when i picked up the machine (i hadn't touched it in at least several days), the touchpad was responsive, but the system would not respond to keystrokes or mouse clicks. after a long time (30 seconds or a minute -- long enough to have a conversation about the symptoms, the machine started responding, and is now behaving "normally".

Attachments

logs.Unknown.2012-03-21.19-34-59.tar.bz2 (1.8 MB) - added by pgf 2 years ago.

Change History

Changed 2 years ago by pgf

Changed 2 years ago by martin.langhoff

  • cc martin.langhoff added

Looks a lot like #10386 -- that bug contains instructions to dump Py memory state with Guppy so if you have some time you can confirm whether the seen AP/ESSID objects are still leaking.

Note: See TracTickets for help on using tickets.