All my batch 0000035 units are now in PVal state. All were suspended once with LAIM off. The Result Logs show "Quit requested: Exiting" at that point, followed by a resume at the start of the Task that was running when the previous checkpoint had taken place (from the Event Log). Is that the correct behaviour?

PS Keith - you might have intended a quorum of one, but they look like quorum of 2 to me; or maybe you meant to write quorum of 2, hehe

All seemed to checkpoint once. checkpointing revised to 60 seconds for this.

all seemed to re-start after suspend (LAIM off) Done before checkpoint.

poor efficiency observed typically losing 2.5 minutes or more every 20 mins of realtime up to 5 mins in 25 mins of realtime. cpu% in low 80's as observed in boinctasks.

e5-2650 Linux mint 17

All other tasks finished, this one got stuck at 1,667% without checkpoints, cpu time counting up:
7.04 Beta Test BETA_OET1_0000300_xMBGP-FA_rig_0041_1 01:49:44 (01:42:13) [0] 01:42:13 93,16 1,667 05:18:21 07-12-2014 21:15 Running linux_64-gny 30.58 MB 33.73 MB

This machine had also reported 2 jobs with cpu time reset:
World Community Grid 7.04 Beta Test BETA_OET1_0000035_xMBGP_4754_1 00:42:14 (00:00:33) 27-11-2014 21:56 27-11-2014 21:57 1,30 Reported: OK * linux_64-gny 48.73 MB 45.41 MB
World Community Grid 7.04 Beta Test BETA_OET1_0000035_xMBGP_3264_1 00:33:32 (00:00:25) 27-11-2014 21:14 27-11-2014 21:18 1,24 Reported: OK linux_64-gny 42.49 MB 39.15 MB

The other tasks reported as normal.

Perhaps a high failure rate with lots of resends is expected.

5,500 units with 2,200 errors/too late etc.
2,200 resends with 1,100 still errors
1,100 resends with 1,100 errors
1,100 resends with 1,100 errors
1,100 resends with 1,100 errors
1,100 resends with 1,100 errors
11,000 results from 5,500 units

Found 3 in the queue and pushed them ahead. Ran fine, adhered to the 1000 second checkpoint setting, a little less then optimal efficiency at 94-96 percent, but this could be an effect of running along with 5 ugm. Peak memory/vm usage modest at under 50mb. Did not look at any disk i/o to verify there's no 'write to disk at most' ignoring outside of the allowed checkpoints.

7.04 beta20 BETA_OET1_0000035_xMBGP_2011_0 00:25:25 (00:23:57) 11/28/2014 12:34:22 AM 94,23 Ready to report 43.96 MB 42.36 MB
7.04 beta20 BETA_OET1_0000035_xMBGP_4776_0 00:24:36 (00:23:17) 11/28/2014 12:34:22 AM 94,65 Ready to report 44.22 MB 42.61 MB
7.04 beta20 BETA_OET1_0000035_xMBGP_1356_0 00:25:56 (00:24:55) 11/28/2014 12:33:28 AM 96,08 Reported: OK * 47.21 MB 45.62 MB

The version number 7.04 implies this one did not need extended time in alpha.