Author |
Message |
|
These come to my PC stating 11 + hours. I have 8 still running at 2 d 19:39:41 and counting.
It counts seconds up in elapsed with the clock, but the remaining does not decrement for about 10-15 seconds per second of elapsed....These at taking forever.
I had to cancel lots of WU's because they would never finish and some let alone start in time.
If you do not know how long a WU will take, do not use a short time for the projected Elapsed, say it will take 2 days.
____________
I got in, cool, more work. |
|
|
rebirtherVolunteer moderator Project administrator Project developer Project tester Project scientist
Send message
Joined: 2 Jan 13 Posts: 7479 Credit: 43,682,241 RAC: 42,410 |
These come to my PC stating 11 + hours. I have 8 still running at 2 d 19:39:41 and counting.
It counts seconds up in elapsed with the clock, but the remaining does not decrement for about 10-15 seconds per second of elapsed....These at taking forever.
I had to cancel lots of WU's because they would never finish and some let alone start in time.
If you do not know how long a WU will take, do not use a short time for the projected Elapsed, say it will take 2 days.
Check the stderr.txt on windows and calculate the max runtime. All WUs will be finished in time and has no endless loop. |
|
|
|
@Viking69: This ongoing R1024 batch is VERY a demanding one on the CPUs.
It can easily run 30h+ on a bit slower (AVX based) CPU, wouldn't even try to run them on the slower ones (ok, I have HT on (Intel) as I still need to feed the GPUs).
But I use the preferences to set up the right application depending on the box, no use to run too big ones on slower boxes, they won't make it in time.
It's also good to check out Reb's predicted (tested I guess, pls correct me if I'm wrong) times when bases are loaded and the Server Status page when they are running (the times usually get longer towards the end of the batch (not so much with the short ones)).
And as we are doing very different ones, even within the same application, the Boinc client won't be able to "learn" predict the right time...
Mankka* |
|
|
|
For this program BOINC I have not had to do additional math to determine the length of a WU for other projects.
The last 10 minutes of these last WU's have taken over 14 hours to get down to 2:13 to go......It is now over 3 minutes of process time before the remaining time decrements a single second. If this is the case then it will never finish, and I am over the deadline by 13 + hours now.
If I don't get credit for these I will be pissed!
____________
I got in, cool, more work. |
|
|
rebirtherVolunteer moderator Project administrator Project developer Project tester Project scientist
Send message
Joined: 2 Jan 13 Posts: 7479 Credit: 43,682,241 RAC: 42,410 |
For this program BOINC I have not had to do additional math to determine the length of a WU for other projects.
The last 10 minutes of these last WU's have taken over 14 hours to get down to 2:13 to go......It is now over 3 minutes of process time before the remaining time decrements a single second. If this is the case then it will never finish, and I am over the deadline by 13 + hours now.
If I don't get credit for these I will be pissed!
The deadline for the long WUs is 8 days. |
|
|
|
I had 8 WU's that were at 3 days + crunch time and I only see 2 in my completed....
task 165249769
work unit 162812451
&
task 165249727
work unit 162812409
9/22/2016 2:08:31 PM | SRBase | Finished upload of R1024_780-850k_wu_364_0_0
9/22/2016 12:08:11 PM | SRBase | Finished upload of R1024_780-850k_wu_350_0_0
9/22/2016 11:47:16 AM | SRBase | Finished upload of R1024_780-850k_wu_310_0_0
9/22/2016 11:30:07 AM | SRBase | Finished upload of R1024_780-850k_wu_351_0_0
9/22/2016 11:11:46 AM | SRBase | Finished upload of R1024_780-850k_wu_275_0_0
9/22/2016 11:09:17 AM | SRBase | Finished upload of R1024_780-850k_wu_309_0_0
9/22/2016 10:55:38 AM | SRBase | Finished upload of R1024_780-850k_wu_343_0_0
9/22/2016 9:39:01 AM | SRBase | Finished upload of R1024_780-850k_wu_306_0_0
____________
I got in, cool, more work. |
|
|