Just wondering if others are having new WU Fetch issues?

6.6.3 running for at least a month on all rigs....no changes....sudden work fetch issues...

I have 3 rigs running GpuGrid:

2x Gpu working 2 WUs and 2 in queue
1x Gpu working 1 WU and 1 in queue--should be 3 in queue
1x Gpu Zero Wus and Zero in queue--not requesting new work. I've detached/reattached, restarted Boinc, rebooted....nothing.

Just wondering if there is an issue on my end....or if it's a project issue....

WCG is working just fine...

Just sent a request for work on it's own and same reply: "not available for you type of computer". .....this rig has been running GpuGrid for months....

Another voluntary request. This time it says: " Project is shut down for maintenance".

Can't be the issue here though--it would have to be shut down for a long time for 1x 8800 to be out of work.....

I am getting the same message "not available for your type of computer" on 2 out of 4 quads - the other 2 have enoug in cache that they aren't requesting (nor could they finish in time if they had more) any new WUs. I don't think it is you...

I'm not able to get any either.

"3/18/2009 8:27:38 PM|GPUGRID|Message from server: Full-atom molecular dynamics on Cell processor is not available for your type of computer."

I'm having problems getting work on my Linux install, not tried windows yet, getting the 'not available for your type of computer' message, even when I was uploading a finished WU and asking for another.

I've tried all the old tricks, upping resource share, suspending all other work and reducing resource share, detach, reset, even re-installed Boinc and tried another version, all with no luck.

Luckily, I have a couple of WUs on my windows partition to see me through. Thanks in advance.

Mark

You should be getting work now.

ignasi

Thank You Everybody and ignasi for getting the problem fixed. The gpu has been idle all night, but I just did a manual update and received 2 WUs. It did a automatic update shortly after and the response was:

No work sent
(reached CPU limit of 1 tasks)
(Project has no jobs available)

Then immediately DLed 2 more WUs.....lol

So, I'm back in biz for now....hoping everyone else is too....

6.6.3 running for at least a month on all rigs....no changes....sudden work fetch issues...

A word of caution. Several of the issues with the 6.6.x versions do not show up right away. Some are ony visible when you are in a hole so deep that it is near impossible to dig out. I am not saying you should change, but, just because you have run 6.6.3 for a month does not mean that it is not going to bite you at some later date.

Just be cautious ...

I totally agree on that have been running for months without error with some versions and all of a sudden failure after failures and loss of units... settings too much to sum up here.
Untill now i try everytime the newest beta's but i keep coming back to good old 6.5.0 which has some minor issues but somehow is the most stable.

You should be getting work now.

ignasi

Thanks ignasi, all good now.

---

Proud member of BOINC@AUSTRALIA

6.6.3 running for at least a month on all rigs....no changes....sudden work fetch issues...

A word of caution. Several of the issues with the 6.6.x versions do not show up right away. Some are ony visible when you are in a hole so deep that it is near impossible to dig out. I am not saying you should change, but, just because you have run 6.6.3 for a month does not mean that it is not going to bite you at some later date.

Just be cautious ...

Didn't one or few of the early 6.6. clients have long term debt issues? This could explain why the cache on your 8800GT was so low, whereas normally it would have a cache of ~2 days.

MrS

---

Scanning for our furry friends since Jan 2002

Didn't one or few of the early 6.6. clients have long term debt issues? This could explain why the cache on your 8800GT was so low, whereas normally it would have a cache of ~2 days.

Yes, several did... as a matter of fact for some of the latest releases the scuttlebutt is to reset the LTD to zero on install so that the LTD from the old version does not snarl up the new version.

How do you reset the LTD?

How do you reset the LTD?

From Richard Haselgrove:

There's a new option for
cc_config.xml:

<zero_debts>1</zero_debts>

(documented at the end of
http://boinc.berkeley.edu/wiki/Client_configuration)

You have to add it to the 'options' section of cc_config, restart the BOINC
CC (reloading cc_config isn't enough), and then disable or remove it so
BOINC can manage the new debt system itself in future.

For versions of BOINC that do not support this, you have to edit the client state file with a txt editor.

I seem to have the same problem again that was the original subject of this thread - no work "for your type of computer".

This is on a machine that normally gets work no problems.

---

Proud member of BOINC@AUSTRALIA

I had 2 ignasi units error out then the same message as previous post
So no new work been received