Incident of failure to continue

[archae86]

I had a curious happening on one host this morning.  It is a Gallatin hyperthreaded host running Windows XP SP2.  I have the host set to 80% resource share SETI Enhanced, 20% Einstein, so it usually runs two SETI threads but often one SETI and one Einstein.

A few minutes ago, I noticed that the "active" SETI thread had not used CPU since 12:30 a.m. this morning. While BOINCmgr showed it as "running", the CPU time did not increment, my CPU showed as 50% idle, the System Idle Process accumulated over 7 hours of time, etc. 

I can identify the time when idle time went way up from the temperature logs, as CPU temp dropped 4 degrees C over a three minute period.  Reviewing the message logs says this happened plus or minus a couple of minutes at a time when BOINCmgr resumed a SETI job having paused an Einstein job.

On noting the oddity, I exited BOINCmgr and restarted.  On resumption two SETI jobs (including the one which had been failing to use CPU) resumed succesfully.

I'm running "Windows P4 SSE2 32-bit V5.15 'Chicken Good!' (R-1.3|+freq|xN)"
My client is TruxSoft tx36, and I run the BOINCmgr 5.2.13

The WU which was involved has been run to completion by two others my problem WU

I've not seen this specific behavior before, and don't know whether it is most likely something odd in the science ap, the client, BOINCmgr, or my system.

I'm reporting in case others see such a thing, we might build up a pattern.

I'll post this comment on the SETI forum as well, as a different set of users might see it there and be moved to report if they have seen the same thing.

The enemy of my enemy is not quite as much of an enemy as my enemy if they ask, and in either case, I will play nice to the enemy of my enemy only so far as it hurts my enemy for real.

- 13th century Mongol warlord trying to describe the current semiconductor marketplace after dining on tainted cheese