I noticed for a few days that my Windows machine seemed to hang up on WU's, after the 19th I have to kill a few, and I can't get a steady stream of results from it.

And now I've noticed my mac has been working on a new WU for 48+ minutes. Looks like I've had a few WU's finish in 3900 seconds as opposed to 400 seconds, which I'm fine with if they return valid results. The only thing that's a little irksome is the fact that the credit dishout is the same. 4.07 for a 400 second WU vs 4.07 for a 3900 second WU...

Just i killed one on Ubuntu Hardy: it was running a couple of hours and no progress. All others on my other machines hasn´t running yet, so I killed them, before they waste CPU

And at the moment: "no new work" accepted from Milky, sorry !!

---

Sysadm@Nbg
Member of Team

I noticed for a few days that my Windows machine seemed to hang up on WU's, after the 19th I have to kill a few, and I can't get a steady stream of results from it.

And now I've noticed my mac has been working on a new WU for 48+ minutes. Looks like I've had a few WU's finish in 3900 seconds as opposed to 400 seconds, which I'm fine with if they return valid results. The only thing that's a little irksome is the fact that the credit dishout is the same. 4.07 for a 400 second WU vs 4.07 for a 3900 second WU...

This look like a different problem as the new WUs have only been available for the past day and when they do run they seem to be fractionallly faster - approx 13 min cpu time

My windows machine errors out in a matter of seconds with the new WUs

I noticed for a few days that my Windows machine seemed to hang up on WU's, after the 19th I have to kill a few, and I can't get a steady stream of results from it.

And now I've noticed my mac has been working on a new WU for 48+ minutes. Looks like I've had a few WU's finish in 3900 seconds as opposed to 400 seconds, which I'm fine with if they return valid results. The only thing that's a little irksome is the fact that the credit dishout is the same. 4.07 for a 400 second WU vs 4.07 for a 3900 second WU...

This look like a different problem as the new WUs have only been available for the past day and when they do run they seem to be fractionallly faster - approx 13 min cpu time

My windows machine errors out in a matter of seconds with the new WUs

This is understandable due to the NaNs. We are pulling all those erroring WUs from the database as we speak. Sorry about the inconvenience.

---

Dave Przybylo
MilkyWay@home Developer
Department of Computer Science
Rensselaer Polytechnic Institute

I've got two 2h-ones yesterday, and one 4h-one just now. I've killed them, have the project set to NNW and am just waiting to report them back before resetting the project. Or is there no need to report them?

---

Grüße vom Sänger

I've got two 2h-ones yesterday, and one 4h-one just now. I've killed them, have the project set to NNW and am just waiting to report them back before resetting the project. Or is there no need to report them?

we might have found the problem. we had the wrong script set up to dump the database every few days -- so while i took the server down this morning to get rid of the ones from the new search, the database was too slow to do anything.

i'm running a full purge (which is gonna take quite awhile), but after that i'll get the remaining gs_3737... WUs out of the system. i've already stopped the search so there wont be any new ones generated. hopefully they'll get out of the systems quickly.

---

I've got two 2h-ones yesterday, and one 4h-one just now. I've killed them, have the project set to NNW and am just waiting to report them back before resetting the project. Or is there no need to report them?

we might have found the problem. we had the wrong script set up to dump the database every few days -- so while i took the server down this morning to get rid of the ones from the new search, the database was too slow to do anything.

i'm running a full purge (which is gonna take quite awhile), but after that i'll get the remaining gs_3737... WUs out of the system. i've already stopped the search so there wont be any new ones generated. hopefully they'll get out of the systems quickly.

Thanks for the update! Looking forward to getting on with the new WU's once it sorts out :)

I've got two 2h-ones yesterday, and one 4h-one just now. I've killed them, have the project set to NNW and am just waiting to report them back before resetting the project. Or is there no need to report them?

we might have found the problem. we had the wrong script set up to dump the database every few days -- so while i took the server down this morning to get rid of the ones from the new search, the database was too slow to do anything.

i'm running a full purge (which is gonna take quite awhile), but after that i'll get the remaining gs_3737... WUs out of the system. i've already stopped the search so there wont be any new ones generated. hopefully they'll get out of the systems quickly.

Thanks for the update! Looking forward to getting on with the new WU's once it sorts out :)

I just got three of the 3737 wu's, I thought that they were all flushed from the system ????

EDIT: All my PC's are getting them now. I have set to "No new work" again till it is sorted out.

---

Proud Founder and member of



Have a look at my WebCam

I'm getting gs_59x ones; perhaps the problem has been resolved.

I'm getting gs_59x ones; perhaps the problem has been resolved.

NO, they are mixed together ...

Windows systems seem to error on these bad Tasks immediately ...

My OS-X computers seem to run them for an hour and they complete and validate ... oddly enough ...

---

I just got some 3737's mixed in...not many....2 of 20.

---

They are slowly getting rid of the bad W/U's with less and less coming through, I seem to be having a minor problem with validation (as in it ain't happening) getting quite a list of pending, but this may be due to them trying to weed out the bad w/u's???

I've got a bunch of pendings too.

---

Calm Chaos Forum...Join Calm Chaos Now

I've got a bunch of pendings too.

I believe when we took them out of the database, the bad workunits when and if they finish can't validate because BOINC doesn't know what they are. It thinks you're giving it a workunit that doesn't exist.

---

Dave Przybylo
MilkyWay@home Developer
Department of Computer Science
Rensselaer Polytechnic Institute

Task ID 30924039
Workunit 30447102
33014.631285
stderr out

<core_client_version>5.10.45</core_client_version>
<![CDATA[
<message>
Maximum CPU time exceeded
</message>
<stderr_txt>

</stderr_txt>
]]>

Validate state Invalid
Claimed credit 302.729235230077
Granted credit 0
application version 1.23

Dang thought I'd found the one that ran for 10 hours + but this only comes up to 9 or so hours. Xubuntu 64b, v8.04

UPDATED: Here we go...
Task ID 30921915
Name gs_3737082_1211945903_968749_3
Workunit 30465650
CPU time 37535.465819

You know what I wanna whine about... but will resist.

Signed, the credit whore


EDIT2: Oh so so sad... two more 9 hour WUs, a 10hr 45min one and my entry for the grand booby prize... over 11hours, 10min, 48 seconds on a 2.55GHz Phenom:

Task ID 30896339
Name gs_3737082_1211948989_978283_1
Workunit 30475184
CPU time 40248.863396

Did I win, huh? huh?

Task ID 30924039
Workunit 30447102
33014.631285
stderr out

<core_client_version>5.10.45</core_client_version>
<![CDATA[
<message>
Maximum CPU time exceeded
</message>
<stderr_txt>

</stderr_txt>
]]>

Validate state Invalid
Claimed credit 302.729235230077
Granted credit 0
application version 1.23

Dang thought I'd found the one that ran for 10 hours + but this only comes up to 9 or so hours. Xubuntu 64b, v8.04

UPDATED: Here we go...
Task ID 30921915
Name gs_3737082_1211945903_968749_3
Workunit 30465650
CPU time 37535.465819

You know what I wanna whine about... but will resist.

Signed, the credit whore


EDIT2: Oh so so sad... two more 9 hour WUs, a 10hr 45min one and my entry for the grand booby prize... over 11hours, 10min, 48 seconds on a 2.55GHz Phenom:

Task ID 30896339
Name gs_3737082_1211948989_978283_1
Workunit 30475184
CPU time 40248.863396

Did I win, huh? huh?

As they say in the game of Pool, Foul Stroke - Four Away, I believe I have pipped you at the post there Skip Da Shu, Try 13.5 hours:

Task ID 30899627
Name gs_3737082_1211952813_988295_0
Workunit 30485196
Created 27 May 2008 23:08:45 UTC
Sent 27 May 2008 23:09:38 UTC
Received 28 May 2008 19:16:42 UTC
Server state Over
Outcome Client error
Client state Compute error
Exit status -177 (0xffffffffffffff4f)
Computer ID 4778
Report deadline 1 Jun 2008 23:09:38 UTC
CPU time 48620.108619
stderr out

<core_client_version>5.10.21</core_client_version>
<![CDATA[
<message>
Maximum CPU time exceeded
</message>
<stderr_txt>

</stderr_txt>
]]>

Validate state Invalid
Claimed credit 165.045104245639
Granted credit 0
application version 1.22

I have a few more that are up to 1.5 hours now so I will abort them, all on my Linux system (Fedora Core 6), AMD Opteron 285.

---

I've got a bunch of pendings too.

I believe when we took them out of the database, the bad workunits when and if they finish can't validate because BOINC doesn't know what they are. It thinks you're giving it a workunit that doesn't exist.

You haven't got them out yet, I've been getting them as fast as I can abort them, I Abort 4 & get 4 back again, set to NNW for now again, too much work having to go thru all that just to get a few good ones ... 0_o

had the same for the last 30-40 min, but the last few updates haven't had any. They seem to be coming in waves, about an hour before I will have to stop monitoring my rigs, not sure im going to let MW run unattended, if I don't get anymore in the next hour I may risk it :)
I'm sure I will regret this decision :)

Got a task from this WU.
name: gs_3737082_1211956623_998318_3
WU name: gs_3737082_1211956623_998318
app version num: 122
checkpoint CPU time: 0.000000
current CPU time: 1258.242634
fraction done: 0.000000

According to the posts above it's funny that
the reverse of the fraction-done-value gives ...
NaN :-)

EDIT: when I'd reported my aborted task,
the next one was created immediately afterwards.

This may be overcome, as seen on SAH beta, by setting max#oftotaltasks to zero in the concerned WU's.

EDIT2: Anonymous got it onto it's host.

Since Windows WUs error out quickly, rather than taking many hours to get nowhere as they do on Linux, I've enabled Work Fetch on 2 of my boxes to help clear them out. Apart from lost credit, I'm hoping there won't be any problems.

You see, I don't only crunch for credits...

Rod