Folding Forum

Do we really need to list up all of the "bad" 9201s? I got a few and I can see that some members of my team got 4-7 since the 15th. Put a plug into that, please.

No, Stanford does not need a long list of WUs that did or did not receive bonuses. The data already given will allow them to figure out what the problem is/was and fix it. Then, if history is any guide to the future, they'll process re-credits for everyone.

Hey everyone, apologies for this recurring stats issue. Rest assured that we're working on it on our end. Unfortunately I'm out of the country right now and can't access either the WS or AS so what I can do is rather limited, but we'll do our best to fix this ASAP.

Don't forget to look at psummary settings for your project, it appears twice from two different servers :

"ASAP" seems to be quite slow at stanford or is there a bigger problem ? @jadeshi: will you post here, when correct points are calculated and put to the accounts ?. I wish all a merry christmas, hear you later.

I guess the problem is the holidays

Any update here?

This and the recent 9430 issue is the first cases where points were not rewarded soon after the mistake was found. Before you could rest assure that whatever problems the stat had points were there shortly. Now I doubt and I will delete any WU related to a weak CS as soon as I spot a hick up. Got that jadesi and cxh?

I do not know what happened in this particular case; I can only report the facts behind similar reports in the past. Nevertheless, I suspect that your statement about a "weak CS" is incorrect.

Apparently the CS was doing it's job. It successfully uploaded what it classified as a corrupt WU or a duplicate WU and in either case, the WU needed to be discarded. If it was a corrupt WU, the CS is not responsible for the WU becoming corrupt. If the WU had been accepted by the WS but that fact was not reported back to your client, the CS should not accept a duplicate upload but, in fact, should tell your client to stop trying to send it.

It's been a month since this problem was first identified, and 3 weeks since jadeshi acknowledged the issue and indicated "we're working on it". We can speculate all day long on what the CS is doing to the WS with duplicate/corrupted WUs, but at this point, I really don't care what the technical issue is.

Respecting donor contributions means, in part, providing a reasonably solid estimate of when the problem will be fixed such that work submitted is credited correctly, both historically and ongoing. Scientific progress is measured by the WUs returned, donor progress is measured by the points accumulated. One part of that equation is lacking, and it's more than reasonable at this point to expect a timeframe for the fix.

bigblock990 wrote:Any update here?

Still wondering.

soon

Nathan_P wrote:soon

I thought of this situation when someone asked if the missing NaCl WUs would be credited.

The first step is ALWAYS to fix the basic problem so that future WUs will be credited. Once that has been addressed, finding "lost" WUs generally takes much longer. In my experience, most "lost" WUs are eventually found and re-credited although it often takes many weeks to fill in those previous WUs. (Logs and backups have to be processed manually and compared to credit records to avoid double credits.)

Was there credit issued yesterday? I see a lot of teams with an unusually high update on EOC at 12pm 1/19. More WUs and points than normal. There was not a missed EOC update before that like the 3pm update today. The overall points were higher yesterday than normal. The extra points compared to extra WUs would suggest the difference were from high point GPU WUs. I'm see like an extra 20-30% more WUs but an increase in nearly double the points on that update.