Ubuntu
boinc package

BOINC World Community Grid stuck "Waiting for memory"

Bug #156536 reported by shanen (Shannon Jacobs)
8
Affects Status Importance Assigned to Milestone
boinc (Ubuntu)
Invalid
Undecided
Unassigned

Bug Description

This actually appeared some months ago. So far it has only affected one of my machines, but the machine has two disk partitions configured for Ubuntu. There are obviously no hardware differences between them, and they even share the same swap partition.

Before the Gutsy 7.10 upgrade, the BOINC Manager in one of the Ubuntu boots would frequently report that the process was stuck "Waiting for memory" (Wfm). At first it was rare, though it seemed most likely to occur at boot time. For example, it might run to 30% one day, then continue to 75% the next day, but on the third day it would just stick in the Wfm state. Later it became more common, and finally it seemed to occur right after the initial computations. It would download a fresh unit (after a Reset command), run for a few minutes, and then go directly to Wfm. I tried everything I could think of to fix it, including complete removals and reinstalls.

At this time, there was no problem in the other Ubuntu.

After the Gutsy 7:10 upgrade, both of them seem to be broken and immediately transition to the Wfm state. (I'm going to try a few of the other BOINC projects, though just for testing. WorldCommunityGrid is the one that is officially sanctioned by my employer.)

Revision history for this message
Greg Grossmeier (greg.grossmeier) wrote :

Thank you for taking the time to report this bug and helping to make Ubuntu better. You reported this bug a while ago and there hasn't been any activity in it recently. We were wondering is this still an issue for you? Can you try with latest Ubuntu release? Thanks in advance.

Revision history for this message
shanen (Shannon Jacobs) (shanen) wrote :

I still see it sometimes, but more rarely. It's become one of those things that's fluttering at the edge of traceability. That machine actually has two versions of Ubuntu on it right now. My feeling is that the problem is less common with Hardy Heron, but more common with Gutsy. When it does happen, the condition seems to persist, though sometimes resetting the project will start it on a work unit that does not hang. (Other times multiple resets have no effectiveness.)

I basically feel like this is outside of Ubuntu's scope, but is a work unit allocation problem on the project end. You'd think that the size of the problems would be matched to the resources of the volunteer's machine, but if that's supposed to happen, then sometimes it fails.

Since no one else has commented on the bug, I suppose we can assume it's quite rare, though the machine that shows it is a very vanilla IBM NetVista. On the old side, but certainly not a rare machine in its day. This one actually has some extra memory above the original factory installed.

Daniel T Chen (crimsun)
Changed in boinc:
status: Incomplete → New
Revision history for this message
Timmy Shih Jun Yee (shijun) wrote :

Closing this bug because the reporter said it rarely happens and cannot be reproduced easily.

Changed in boinc (Ubuntu):
status: New → Invalid
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.
Subscribing...

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.