No new tasks: Message from server: Server error: can't attach shared memory

[Bobnova]

Mon 14 Dec 2009 07:04:28 PM PST rosetta@home Sending scheduler request: Requested by user. Requesting 60480 seconds of work, reporting 0 completed tasks
Mon 14 Dec 2009 07:04:33 PM PST rosetta@home Scheduler request succeeded: got 0 new tasks
Mon 14 Dec 2009 07:04:33 PM PST rosetta@home Message from server: Server error: can't attach shared memory

Only things i can find via the Great Evil (google) is to edit the config files to point towards srv4, i checked and they already are pointed there.

I got this first when re-attaching to rosetta, so i did a complete uninstall and reinstall, still did it, so i manually updated to the most recent client, and it still does it.

How do i fix this? I want to crunch rosetta!

 

[Jolly-Swagman]

Srv4 is not UP Running see Server Status Thread below

http://boinc.bakerlab.org/rosetta/rah_status.php

 

[baditude_df]

Only things i can find via the Great Evil (google) is to edit the config files to point towards srv4, i checked and they already are pointed there.

I got this first when re-attaching to rosetta, so i did a complete uninstall and reinstall, still did it, so i manually updated to the most recent client, and it still does it.

How do i fix this? I want to crunch rosetta!

And we want you to crunch Rosetta. Thx.

Srv4 is not UP Running see Server Status Thread below

http://boinc.bakerlab.org/rosetta/rah_status.php

Swag beat me to it. Thx bud!

 

[Bobnova]

Ahh, well now that is an issue.
Is there anything i can do aside from waiting for it to come back to life? It looks like it's the only Feeder server, i assume i have to connect via a feeder, right?

 

[baditude_df]

Try detatching from the project and try again, but I think you might end up waiting till that puppy is back on line. Not sure what the deal is there, and I'm quite sure it's never effected me. Even if it did, I wouldn't know for about a week as I'd have to burn through my work buffer to find out.

 

[Bobnova]

Ya, looks like i have to wait, tried detaching/reattaching, no dice.
Tried entering a different server name, no dice there either.

Ahh well, they'll probably get it going tomorrow sometime, or at least i hope so. If my comp gets half an hour of access i should have a 10 day buffer.

 

[baditude_df]

Yep, normally these problems are fixed within a day.
I wonder if this had anything to do with the new2.03 minirosetta.exe that was released today?
EDIT: Yep, looks like it. http://boinc.bakerlab.org/rosetta/forum_thread.php?id=5186

And yeah, once it's back up you should have a full buffer within a day or so depending on the server load. Typically I give myself a 5 day buffer which equals out to ~6.5pages of WU's in BOINC manager, and I've never run out.

 

[Bobnova]

Heh, that at least means it's not something on my end.
Looks like my client will try to connect every hour or so till it works, that's one of the reasons i like rosetta, it seems more resistant to such things.

 

[Bobnova]

They fixed it! I now have an 88 WU queue, that should hold me for a while
Looks like 5.5 days minimum, now i can weather another server crash.

That said, one failed at the 7000 second mark, which is strange as i'm running a known-stable OC right now.

 

[baditude_df]

I've had a few computation errors lately, but only on very certian WU's. The rest are fine.

 

[Bobnova]

Must just be a few defective WUs floating around.
I seem to remember my AthlonII 245 box getting a couple computation errors at stock speeds, now that i think about it.

 

[Bobnova]

Came home to find my computer locked solid, i think Rosetta is much more demanding on memory then pretty much anything else going. Dropped it a PL (to 6, from 5), much less frozen now
This explains the strange results and graphical artifacts rosetta was causing on my dads PhII 810 box(It'd run for ~20 minutes then throw computation errors out everthing and artifact like the GPU was at 200*c, despite all temps being good), the ram was overclocked (couldn't find the command rate, so it was running 1t instead of 2t) more then it liked.

This explains it all! Thanks rosetta!