SAS50 workunits 'break down' after a few seconds.

<stderr_txt>

Model crashed: INANCILA:integer header error tmp/xadae.pipe_dummy
Regional Worker:: CPDN process is not running, exiting, bRetVal = 1, checkPID=15784, selfPID=15784, iMonCtr=2
Controller:: CPDN process is not running, exiting, bRetVal = 1, checkPID=15784, selfPID=16112, iMonCtr=1
Model crash detected, will try to restart...
Leaving CPDN_ain::Monitor...
15:33:37 (16112): called boinc_finish(0)

---

Have you had a bunch of them do so? Are any tasks that you might have downloaded today still running?

They think there's a corruption in a restart file (or restart files) but are unsure how widespread it is.

No all workunits crashed so far.

---

I see that 5 on Bonsai's computer have crashed so far. https://www.cpdn.org/cpdnboinc/results.php?hostid=1377284

When a second one of mine crashed I had to wait an hour for the timeout before it would show as crashed on its page.

I see that 5 on Bonsai's computer have crashed so far. https://www.cpdn.org/cpdnboinc/results.php?hostid=1377284

When a second one of mine crashed I had to wait an hour for the timeout before it would show as crashed on its page.

You can update the project in boinc manager to show the crashed task immediately on the webpage. Of course if you are wanting it to request more tasks, that resets the communication time to 1 hour. I've been doing that since it appears that I am having no luck downloading a good task, and now I've "fulfilled my daily quota" for both PCs.

That's right.

I see that 5 on Bonsai's computer have crashed so far.

But I received 5 workunits more, and they all also crashed after 13 or 14 seconds.

---

You can update the project in boinc manager to show the crashed task immediately on the webpage.

Should have thought of that! Having also crashed my daily quota machine is now crunching other projects on one of its two cores. I am keeping my marginally faster machine running native Linux work in the hope that at some stage testing might lead to main site work for it again.

Nothing new: The next five workunits crashed.


15 wu Model crashed: INANCILA:integer header error

---

Nothing new: The next five workunits crashed.

Thanks,
If anyone has any of these tasks (batches 703,704 and 705) running past the first few seconds (13) on my box without crashing can you please let us know.

Unsent tasks may be withdrawn but if some are working OK the withdrawn ones will be re-issued as that would mean it was a science issue rather than a purely dodgy xml file one.

Edit: unsent tasks have been paused for now.

I've got two of these batch 705, they ran for 30 secs and crashed with INANCILA:integer header error. (running WINE, BOINC 7.8.3)

I've got two of these batch 705, they ran for 30 secs and crashed with INANCILA:integer header error. (running WINE, BOINC 7.8.3)

17 seconds longer than my 703s managed! Priority is anyone with them not crashing please post!

Just had a 704 crash after about 12 seconds.

No reports of these tasks running. Team at Oxford are investigating the problem.

Problem identified and when correct files are uploaded to the system the batches will go out again.

These have started pouring into the hopper and one from batch 706 is now about 8 minutes in, well past the time when they were crashing before due to some misonfigured files.