Message boards :
Number crunching :
inconsistant %complete for hours computed
Message board moderation
| Author | Message |
|---|---|
|
Send message Joined: 18 Aug 07 Posts: 15 Credit: 22,771,987 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]()
|
Greetings all, I consistently have wu\'s that move up to around 90% complete in the span of 3-4 hours, and then take around a total of 25ish to complete the last 10%. This is not to accurate. I know its a guess-timation, but its usually way off. Anyone else noticed this? |
Stefan LedwinaSend message Joined: 16 Jul 07 Posts: 464 Credit: 298,573,998 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]()
|
This is normal for PS3GRID and the behaviour since the project started... ;-) pixelicious.at - my little photoblog |
|
Send message Joined: 23 Nov 07 Posts: 13 Credit: 531,316 RAC: 0 Level ![]() Scientific publications
|
Well, while we are on the topic... Is there is reason for this? I knew the WUs take about 20-24 hours to complete, so it isnt a big deal. But what kinda work actually goes on to calculate the completion time? With my limited programming knowledge, i know \"% Complete\" monitors are sometimes tricky. Guessing there is just some \"loops\" that take longer to complete. So lets say there is 100 tasks in each workunit. It completes 1-90 relativly fast. Therefore the % would show that there is only 10% left. But, that 10% is just the work that takes the longest. that is just my speculation. Anyone actually have any real data on this? |
©2025 Universitat Pompeu Fabra