| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Support
Forum: BOINC Agent Support
Thread: Is there any reason for results to be delayed by over 2 hours? |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 4
|
|
| Author |
|
|
Johnny Cool
Ace Cruncher USA Joined: Jul 28, 2005 Post Count: 8621 Status: Offline Project Badges: 
|
For the past two weeks (maybe it was one), I have noticed a somewhat strange pattern of seeing my results being delayed by two hours (or more). You know, when you finish one work unit and then have to wait. The work units in question are the HCMD Phase 2. I am running Boinc version 6.10.58.
----------------------------------------My PC is an Intel Core i7-860 Lynnfield 2.8GHz 8MB L3 Cache LGA 1156 95W Quad-Core Processor BX80605I7860 (not overclocked). I am running Windows 7 Ultimate with 2x 2GB Corsair XMS DDR3-1600 I have never seen gaps like this (results being delayed this long). Is this normal? Does not seem to be to me. That is quite a gap. Is it the work unit? Is it the WCG? Or is this just "one of those things".  Oh yes, I also want my 'wingman' to be tied to across a tree where I can throw tomatoes at this person like Randy Johnson can throw a fastball. Every minute, at will.  ---------------------------------------- [Edit 1 times, last edit by Johnny Cool at Oct 11, 2010 1:31:45 AM] |
||
|
|
anhhai
Veteran Cruncher Joined: Mar 22, 2005 Post Count: 839 Status: Offline Project Badges: 
|
what delay are you talking about?
----------------------------------------If it is the delay before your result get verified? Well that is all dependent on your wingmen (and a little bit of processing power from WCG). If it is delay before you result get reported to WCG, then it is all dependent on when you need a new WU, then it will get reported. If it is before you get a new WU to process then you should increase your cache, although 2 hr wait sounds off.  |
||
|
|
Johnny Cool
Ace Cruncher USA Joined: Jul 28, 2005 Post Count: 8621 Status: Offline Project Badges:
|
what delay are you talking about? If it is the delay before your result get verified? Well that is all dependent on your wingmen (and a little bit of processing power from WCG). If it is delay before you result get reported to WCG, then it is all dependent on when you need a new WU, then it will get reported. If it is before you get a new WU to process then you should increase your cache, although 2 hr wait sounds off. Did you read the ... You know, when you finish one work unit and then have to wait. No, from your Boinc client.  Usually, it is a steady flow after you have completed a work unit or 8 (or 16). One right after another ...  |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
hmmm, well I have to admit, not native English speaker, that your line is leaving much open to interpretation.
----------------------------------------Work is generally requested before finishing, so if crunching is uninterrupted all is fine. If your client is told to back off due it's very frequent reporting, as suggested, increase your cache. Some with these heavies had to increase cache to 2 days to keep the flow going (6.10 bug iIMHO). I wont run below 0.9 days Additional Buffer for 99% of the time. If you are permanently connected to the Intertube, set the "connect every xxx days" to zero... could be one reason. Clients are completely unaware of your science selections. If there are delays for one reason or the other to get a backfill then look in the message log and the project properties screen. There are 2 counters there: CPU Work Fetch deferred for: CPU Work Fetch deferral interval: I understand that if GPU is enabled and detected, the are 2 more lines, but than reading GPU Work... Let us know if any of this helped to get your normal expectations to be met.
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
|