You're spot on, it's right there at top of stderr.txt log, number of tasks = 1. Not like you'd want to run these on parttime devices . Fortunately these 1 job tasks are held in memory even when the agent is not set with laim on, i.e. to only compute when system is idle.

Rather funky, while the tasks are running the cpu time properly accumulates, but on conclusion both the boinctasks history and result status page show just a few seconds. One way to never make progress on badge hunting

BETA_ OET1_ 0000040_ xEBGP_ 0009_ 1-- Valid 12/1/14 07:35:55 12/1/14 11:19:36 0.00 / 1.91 56.0 / 46.7
BETA_ OET1_ 0000039_ xEBGP_ 0429_ 0-- Valid 12/1/14 07:33:45 12/1/14 11:24:43 0.00 / 2.00 58.8 / 65.7
BETA_ OET1_ 0000039_ xEBGP_ 0467_ 1-- Pending Validation 12/1/14 07:31:35 12/1/14 11:31:45 0.00 / 2.21 64.8 / 0.0
BETA_ OET1_ 0000039_ xEBGP_ 0748_ 0-- Pending Validation 12/1/14 07:24:24 12/1/14 09:58:24 0.00 / 1.98 58.2 / 0.0

Not deemed to be a science app issue, more how the boinc wrapper picks up and passes on the the information.

I've experienced this problem with BETA_ OET1_ 0000304_ xZAGP-FA_ rig_ 0042_ 1-- and two others xZAGP-FA units.
I've got about 35% progress (it have take about 10 minutes) at progress bar but thet it drops to 10,000% and start slowly rising by 10,000% steps. When i've got 100% unit, BOINC switched to unit of another project. But xZAGP-FA unit was not finished. It switched to Waiting to run. I've start this unit manualy (I paused every other units) and then xZAGP-FA unit was finished and BOINC send it back.
Maybe that's why people getting 0,00 CPU time.

I've got 5 of the 7.05 ones running on a win 8.1 machine that are doing the whole 'no checkpoint and run a long time' thing.

They all seemed to start off running normally but at some point (so far ranging from 5-60%) the progress percent and remaining time got stuck. However when checking the properties tab, the cpu time is still going up at a constant rate. They also don't seem to be entirely stuck, but only incrementing 5% at a time (they all got 'stuck' at some multiple of 5). the remaining time is occasionally updating, but seems to be going up each time.
Four of them are xZAGP-FA tasks and all _1
the fifth is an xSDGP-L task and is a _0
I seem to recall something similar to this happening on this machine during the MCM beta.


EDIT: one of the tasks has the progress 'stuck' (jumping 5% at a time) but the remaining time is counting down normally. No checkpoints after 50 minutes though.

EDIT 2: after completion on the one where the cpu AND remaining time was counting normally, the CPU time was reset to 2 seconds like others have reported

Also,

We have batches in the 300's, they are a lot tougher jobs than the others. They do not have checkpointing as they are a single job. We are working on a way to add checkpointing within these complex jobs.

Thanks,
-Uplinger