Message boards :
Number crunching :
Huge pending credit
Message board moderation
| Author | Message |
|---|---|
|
Send message Joined: 29 Sep 04 Posts: 196 Credit: 207,040 RAC: 0 |
Anyone else have a large pending credit? 1,379 here at time of writing. I looked at a number of the workunits and they were sent to me and 4 of 5 results were yet to be sent out.. goes back as far as 14 Sept. Just a curiosity. :) 
|
 Paul D. BuckSend message Joined: 2 Sep 04 Posts: 545 Credit: 148,912 RAC: 0 |
Mine is 755 ... But, I only have 8 computers, and I run 5 projects on all of them (well, the PowerMac does not do LHC, so 7 computers). |
|
Send message Joined: 29 Sep 04 Posts: 196 Credit: 207,040 RAC: 0 |
Not bad, Paul.. mine's now up to 1,598.. Seven active hosts here.
|
|
Send message Joined: 13 Jul 05 Posts: 16 Credit: 45,621 RAC: 0 |
660 waiting, with 1 PC running 66%, 1 PC running 33% LHC. |
Jim Baize Send message Joined: 17 Sep 04 Posts: 103 Credit: 38,543 RAC: 0 
|
I have 4 active crunchers on LHC. I have a whooping 0.19 Pending credit. That's right... One result that is still pending and it was obviously a short WU. |
|
Send message Joined: 15 Jul 05 Posts: 17 Credit: 16,521 RAC: 0 |
I currently have about 270, with two CPU's running LHC (one @ 2/3 shared with 1/3 Einstein, the other @ 1/3 shared with 1/6 Einstein and 1/2 CPDN sulphur cycle). It depends upon how quickly a quorum is returned for each WU, and using multiple systems will obviously generate more pending credit. 
|
|
Send message Joined: 29 Sep 04 Posts: 196 Credit: 207,040 RAC: 0 |
I'm finally down to 1,281 .. looks like I'll have a decent credit posting for the previous day, considering. :) My pending probably has something to do with the fact all my PCs are connected via cable and my 'connect to network every x days' is set to 0.1 .. so I'm always getting the "newest" workunits .. and several days later as other people's pending queue is worked down I finally post a huge credit-day. Tis my best guess.
|
|
Send message Joined: 13 Jul 05 Posts: 16 Credit: 45,621 RAC: 0 |
Hi Travis DJ, I don't think your network settings does affect this so much - I did play with the settings and had to crunch a bulk of Einstein and predictor WU's - so I was not in time@lhc like you. although I had up to 700 credits waiting (down to 440 right now) with summarized 1 cpu running lhc 100%. It seems they had some problems validating - take a look at http://www.boincstats.com/stats/project_graph3.php?pr=lhc&table=credits on 12/13th they have a contributed much less credit than normal. |
|
Send message Joined: 1 Sep 04 Posts: 506 Credit: 118,619 RAC: 0 |
This is easy: CERN have released a substantial number of work units. Rather than issue one WUs results to five hosts before moving on to the next, they issue a WU's first result to one host, then the next WU has its first result issued to the next host and so on. When all WUs have had one result issued the cycle begins again with the next result. Since each WU requires 3 valid results for a quorum, this cycle issues three results for each WU, with the prospect that a quorum will be reached before new results are needed. This helps to optimise host workload as the fourth and fifth results may never be issued. A side effect is that the pending credit rises rapidly early on as each WU has only one or two results returned. Once the third and subsequent results are returned towards the end of the batch pending credit drops just as rapidly. No problem validating, and, hopefully, every valid result counts. Gaspode the UnDressed http://www.littlevale.co.uk |
©2025 CERN