• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

D²OL: Is erasing broken WU's with DSpy ok?

vss1980

Platinum Member
I can understand having it erase possible broken WU's helps stop you wasting processing time, but I was just wondering what the D²OL guys think about it.

I take it they re-issue work after it hasn't been recieved for a while?
 
I find that something in dspy is broken and doesn't delete the directories anymore (in Win2K anyway), so I just delete the entire directories. When d2ol starts up again, it redownloads the problem candidate. Usually, the corruption is due to part of the file not being downloaded, so redownloading it is fine 🙂 Remember to shut down the client before you do the alt-f or client that are in the process of being downloaded will be identified as corrupt.
 
I havent used it for that feature. I only use it to detect that they are stuck. Day before yesterday when SO MANY bad cands came in I figured if one is a bad candidate that more may be in that directory so I deleted all the stucts, the Inque file the outque file and the tasks file and restart the agent. That brings in a whole new fresh set of files and so far it has worked well.

But normally I will only delete the bad one from the directory. The past couple days were just due to the flood of traffic that fragged their systems and caused incomplete files to be downloaded via thier network. They are working to upgrade the whole system.

We can sure see a huge difference now that Sengent is no longer in charge.
 
Originally posted by: vss1980
Do you know how long it takes for work to get redistributed if you erase it?

I just deleted the tasks, inqueue, outqueue and tasks file on one of my machines, and my queue just filled up to 149 candidates in less than 5 minutes. If your system does not d/l any work after deleting those files, you may need to jump start it by copying a task.dat file to the Res directory. If you need a tasks.dat file, LMK and I can hook you up with a link.

 
No Daaavo, I dont mean how long it takes for your client to fill up, I mean how long it takes for work to be redistributed back from the servers which they haven't got a result for.....
 
Each candidate gets reprocessed several times, so there isn't any problem with fragmentation like in the RC5 project. This is done to ensure the integrity of the tests 🙂
 
I had the same problem with 2 machines on TSC, had to clean out the cands and they started crunching hapilly again. didnt run into that problem untill yesterday.

Shane
 
Back
Top