I am having 2 problems with WUs recently.

Over the last several weeks I have had WUs run for days (24+ hours) and show still 0% completion. I then quit BOINC and restart it. This generally has gotten things straightened out.

The latest problem is a WU that is stuck. It has been at 77.027% for a really long time. When I select the WU in BOINC manager and click on show graphics I see it says it is doing Baseline smoothing. Watching the display it is apparent it is doing the SAME THINGS over and over again.

I tried the restart BOINC manager thing but the WU is still stuck in the same place.

There is no indication of any problems in the BOINC messages.

What can I do here? Abort the WU? Something else?

Particulars: BOINC 6.10.21, MAC G5 (dual PPCs), OSX 10.5.8

---

try restarting BOINC first

---

In a rich man's house there is no place to spit but his face.
Diogenes Of Sinope

Make sure when you stop BOINC manager that you also shut down connected client. The client, in this case SETI, can still run even though BOINC Manager is shut down.

---

PROUD MEMBER OF Team Starfire World BOINC

try restarting BOINC first

It appears that he has already tried that....

"Over the last several weeks I have had WUs run for days (24+ hours) and show still 0% completion. I then quit BOINC and restart it. This generally has gotten things straightened out."

---

"Freedom is just Chaos, with better lighting." Alan Dean Foster

You might need to check your computing prefs.
Go to your prefs in boinc manager and check your pref settings., one or more maybe causing your wu's to stop

Bob

While I do not remember all the checkin information the Mac side of the house. there were some issues.

Off the top of my head, Boinc is having the issue. I do not recall what the last Stable release before 6.10.21 was. I would consider a rollback (if you still have a copy). If you recall the version number you can find it here Boinc Downloads

Regards

---

Please consider a Donation to the Seti Project.

I've already rolled back from 6.10.31 (Hence running 6.10.21).

Are you suggesting to roll back further?

I've restarted BOINC several times now with no effect on this WU.

Since the SETI processes run under the user boinc_project when you quit BOINC manager the SETI processes go away (as opposed to continue running as someone suggested).

Since the machine is a dual processor, I have 2 WUs going at a time. The other WU is merrily going on its way but this one is really stuck in the mud and muck.

It seems a waste of time and resources to let this WU continue running but I'm not sure how to stop it other than aborting it and if there is some issue with the WU someone will end up with it...

---

Since the SETI processes run under the user boinc_project when you quit BOINC manager the SETI processes go away (as opposed to continue running as someone suggested).

That's only on Windows. On the Mac and OS X you'll need to use this tool, which makes BOINC a daemon, to achieve this feat.

As for the stuck task, I'd say abort it. Another person's computer may be able to do it.

---

Bob, if it is in the work unit itself, maybe aborting it is the best thing. Once enough people send it back in it gets marked as no good and is discarded. Have you checked the work unit to see if it is a reissue already? If the WU name ends in 0 or 1 it is a fresh one. If it ends in 2 or more it's been around the block some.

---

PROUD MEMBER OF Team Starfire World BOINC

I changed preferences so that BOINC would only use one processor and this seemed to get the 'stuck' WU finally going, BUT the WU that was going OK popped off these errors when BOINC reset to using the single processor:

Wed Mar 17 13:18:37 2010 SETI@home Task 26fe07ad.6150.25021.14.10.70_2: no shared memory segment
Wed Mar 17 13:18:37 2010 SETI@home Task 26fe07ad.6150.25021.14.10.70_2 exited with zero status but no 'finished' file
Wed Mar 17 13:18:37 2010 SETI@home If this happens repeatedly you may need to reset the project.

The original WU that was stuck was task 25fe07ab.24367.2526.14.10.27_2

I have never had such issues in the past...

---