The problem is that PG is supplying work for a card that does not have sufficient memory on the card for the task to start. It will just sit there forever.
jm7 |------------> | From: | |------------> >--------------------------------------------------------------------------------------------------------------------------------------------------| |pat leitch <[email protected]> | >--------------------------------------------------------------------------------------------------------------------------------------------------| |------------> | To: | |------------> >--------------------------------------------------------------------------------------------------------------------------------------------------| |<[email protected]> | >--------------------------------------------------------------------------------------------------------------------------------------------------| |------------> | Cc: | |------------> >--------------------------------------------------------------------------------------------------------------------------------------------------| |<[email protected]>, BOINC_DEV Mailing List <[email protected]> | >--------------------------------------------------------------------------------------------------------------------------------------------------| |------------> | Date: | |------------> >--------------------------------------------------------------------------------------------------------------------------------------------------| |12/10/2011 09:34 AM | >--------------------------------------------------------------------------------------------------------------------------------------------------| |------------> | Subject: | |------------> >--------------------------------------------------------------------------------------------------------------------------------------------------| |Re: [boinc_alpha] Why does the server supply work if the client has no chance of handling it. | >--------------------------------------------------------------------------------------------------------------------------------------------------| Primegrid has done this for as long as I can remember. It happens on all their sub projects. I've gotten 1000 WU's that take 5 hours to process and had a 5 day TAT. I always thought it was a quirk on my end On Fri, Dec 9, 2011 at 12:10 PM, <[email protected]> wrote: BOINC downloads work from Prime Grid that the host apparently has no chance of ever completing. Why is this state allowed? The server should refuse to send the work in the first place. 949998 PrimeGrid 12/9/2011 12:51:54 PM [coproc] Insufficient NVIDIA for pps_sr2sieve_38371998_0; need 1, available 0 949999 PrimeGrid 12/9/2011 12:51:54 PM [coproc] some instances lack available memory jm7 _______________________________________________ boinc_alpha mailing list [email protected] http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_alpha To unsubscribe, visit the above URL and (near bottom of page) enter your email address. _______________________________________________ boinc_dev mailing list [email protected] http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_dev To unsubscribe, visit the above URL and (near bottom of page) enter your email address.
