Author |
Message |
Dave  Send message
Joined: 13 Feb 12 Posts: 3171 ID: 130544 Credit: 2,233,579,151 RAC: 561,339
                           
|
What does the following mean? Iam trying to compete in the WSS challenge and both clients for the 2 GPUs keep stopping. Thanks
[2014-04-07 13:50:39 GDT] : Could not find completion line in log file [wwww.log]. Assuming user stopped with ^C
|
|
|
|
What does the following mean? Iam trying to compete in the WSS challenge and both clients for the 2 GPUs keep stopping. Thanks
[2014-04-07 13:50:39 GDT] : Could not find completion line in log file [wwww.log]. Assuming user stopped with ^C
Its not wise to stop wwwwcl with CTRL+C, something went wrong to write to wwww.log. Try to delete the log file and restart the prpclient. |
|
|
Dave  Send message
Joined: 13 Feb 12 Posts: 3171 ID: 130544 Credit: 2,233,579,151 RAC: 561,339
                           
|
Hi
I did not manually stop the units with ctrl-c in this instance (although I have done so in the past i.e today & yesterday). Deleted (renamed) wwww.log & starting afresh.
|
|
|
rogueVolunteer developer
 Send message
Joined: 8 Sep 07 Posts: 1255 ID: 12001 Credit: 18,565,548 RAC: 0
 
|
What does the following mean? Iam trying to compete in the WSS challenge and both clients for the 2 GPUs keep stopping. Thanks
[2014-04-07 13:50:39 GDT] : Could not find completion line in log file [wwww.log]. Assuming user stopped with ^C
Are there any other messages? This implies that the GPU might be throwing an error that isn't caught. Can you run one instance without problems? |
|
|
|
I posted this in the wall-sun-sun thread. http://www.primegrid.com/forum_thread.php?id=4196&nowrap=true#75191. Still not resolved.
./prpclient
[2014-04-05 18:41:58 CDT] PRPNet Client application v5.2.8 started
[2014-04-05 18:41:58 CDT] User name brinktastee at email address is brink@brainlaser.com
[2014-04-05 18:41:58 CDT] WALLSUNSUN: Getting work from server prpnet.primegrid.com at port 13001
[2014-04-05 18:41:59 CDT] WALLSUNSUN: PRPNet server is version 5.2.8
Hi! Welcome to PrimeGrid's Wall-Sun-Sun Prime Search.
wwwwcl v2.2.5, a GPU program to search for Wieferich and WallSunSun primes
Sieve started: (cmdline) 28713310000000000 <= p < 28713320000000000
Fatal Error: Not prime: p = 28713310000000049 c10 = 580 c11 = 15853. The code must have a bug.
[2014-04-05 18:42:08 CDT] : Could not find completion line in log file [wwww.log]. Assuming user stopped with ^C
[2014-04-05 18:42:08 CDT] Total Time: 0:00:11 Total Work Units: 0 Special Results Found: 0
[2014-04-05 18:42:08 CDT] Client shutdown complete
____________
|
|
|
|
Are these problems just the latest incarnation of issues that are reported in the "Call for wwwwcl beta testers" thread? It seems like there were unresolved issues for windows clients with wwwwcl, at least in the last several posts. Warning: that is an extremely long thread to read, covering many moons, and some of the older posts may not now be relevant.
As a workaround, can you just run prpclient in a loop? On a Linux or Mac "csh" shell, it would look something like:
while (1)
./prpclient
rm wwww.log
sleep 10
end
I used to have random infrequent failures with the GPU prpnet apps (GFN, wwww) that were recoverable if just retried. Note that this is by no means a "fix"... it just automates the restart.... if you have some sort of systematic error that causes immediate wwwwcl failure all the time, the above code won't help you at all. Be sure to include the "sleep" so that if "immediate crash and burn" is the case, the script won't go crazy monopolizing your machine. I'm sure there is a windows equivalent but I'm not the guy to ask for that.
--Gary |
|
|
Dave  Send message
Joined: 13 Feb 12 Posts: 3171 ID: 130544 Credit: 2,233,579,151 RAC: 561,339
                           
|
Stable so far after deleting wwww.log. BOINC closed for now. Will reopen it later. |
|
|
|
Are these problems just the latest incarnation of issues that are reported in the "Call for wwwwcl beta testers" thread? It seems like there were unresolved issues for windows clients with wwwwcl, at least in the last several posts. Warning: that is an extremely long thread to read, covering many moons, and some of the older posts may not now be relevant.
As a workaround, can you just run prpclient in a loop? On a Linux or Mac "csh" shell, it would look something like:
while (1)
./prpclient
rm wwww.log
sleep 10
end
I used to have random infrequent failures with the GPU prpnet apps (GFN, wwww) that were recoverable if just retried. Note that this is by no means a "fix"... it just automates the restart.... if you have some sort of systematic error that causes immediate wwwwcl failure all the time, the above code won't help you at all. Be sure to include the "sleep" so that if "immediate crash and burn" is the case, the script won't go crazy monopolizing your machine. I'm sure there is a windows equivalent but I'm not the guy to ask for that.
--Gary
If my memory serves me correctly (which is always in question) I believe this is a new error for me. In particular the line below.
Fatal Error: Not prime: p = 28713310000000049 c10 = 580 c11 = 15853. The code must have a bug.
Running that script just does the same thing over and over. It never does finish even one task.
____________
|
|
|
rogueVolunteer developer
 Send message
Joined: 8 Sep 07 Posts: 1255 ID: 12001 Credit: 18,565,548 RAC: 0
 
|
If my memory serves me correctly (which is always in question) I believe this is a new error for me. In particular the line below.
Fatal Error: Not prime: p = 28713310000000049 c10 = 580 c11 = 15853. The code must have a bug.
Running that script just does the same thing over and over. It never does finish even one task.
28713310000000049 is a prime number, so this isn't an issue with primesieve.
Does anyone else have a problem with this range? If so, then it is a bug in the OpenCL code. If not, then this is a problem with the OpenCL SDK you are using to run the application. |
|
|
|
@brinktastee:
What GPU do you running? If an ATI card what driver do you have? |
|
|
|
@brinktastee:
What GPU do you running? If an ATI card what driver do you have?
It's a Nvidia gtx 470 running on a Mac Pro 1.1. Nothing else crunching. OSX 10.7.5
I have this nvidia driver loaded:
http://www.nvidia.com/object/macosx-cuda-5.5.47-driver.html |
|
|
Sysadm@Nbg Volunteer moderator Volunteer tester Project scientist
 Send message
Joined: 5 Feb 08 Posts: 1216 ID: 18646 Credit: 859,465,433 RAC: 204,020
                      
|
@rogue: just tested it, no problem
$ ./wwwwcl -TWallSunSun -s1000 -p28713310000000000 -P28713320000000000 -v
wwwwcl v2.2.5, a GPU program to search for Wieferich and WallSunSun primes
Compiled Jun 12 2013 with GCC 4.6.1
Platform 0 is a NVIDIA Corporation NVIDIA CUDA, version OpenCL 1.1 CUDA 4.2.1
Device 0 is a NVIDIA Corporation GeForce GTX 550 Ti
workGroupSize = 384000 = 3000 * 32 * 4 (blocks * workGroupSizeMultiple * deviceComputeUnits)
Running with 4 threads
Allocated memory (prior to sieving): 20 MB in CPU, 20 MB in GPU
Sieve started: (cmdline) 28713310000000000 <= p < 28713320000000000
p=28713315369690893, 2.335M p/sec, 1.95 CPU cores, 53.7% done. ETA 08 Apr 18:48
Sieve complete: 28713310000000001 <= p < 28713320000000000 263865524 primes tested
Clock time: 113.19 seconds at at 2331140 p/sec.
Processor time: 219.51 sec. (0.40 init + 219.11 sieve).
Seconds spent in CPU and GPU: 0.91 (cpu), 401.88 (gpu)
Percent of time spent in CPU vs. GPU: 0.23 (cpu), 99.77 (gpu)
CPU/GPU utilization: 1.94 (cores), 1.00 (devices)
Percent of GPU time waiting for GPU: 63.76
____________
Sysadm@Nbg
my current lucky number: 113856050^65536 + 1
PSA-PRPNet-Stats-URL: http://u-g-f.de/PRPNet/
|
|
|
rogueVolunteer developer
 Send message
Joined: 8 Sep 07 Posts: 1255 ID: 12001 Credit: 18,565,548 RAC: 0
 
|
@brinktastee:
What GPU do you running? If an ATI card what driver do you have?
It's a Nvidia gtx 470 running on a Mac Pro 1.1. Nothing else crunching. OSX 10.7.5
I have this nvidia driver loaded:
http://www.nvidia.com/object/macosx-cuda-5.5.47-driver.html
I have a MacBook Pro with an nvidia card, so I'll test the code there tonight. I have the latest OS X and Xcode installed so we'll see if this is a problem with the driver. |
|
|
|
Thank you rogue. :) |
|
|
rogueVolunteer developer
 Send message
Joined: 8 Sep 07 Posts: 1255 ID: 12001 Credit: 18,565,548 RAC: 0
 
|
My MacBook Prod has a GT 650M. It isn't particularly fast, but it doesn't have any problems with this work unit.
I suggest that you upgrade to the latest OS X if you can. |
|
|
|
I will need to do some hacking to get the latest os on it. Sounds like a weekend project. |
|
|
Dave  Send message
Joined: 13 Feb 12 Posts: 3171 ID: 130544 Credit: 2,233,579,151 RAC: 561,339
                           
|
Snap. I still need to keep stirring it.
Btw GDB your email is showing in that post at this time. |
|
|
Michael Goetz Volunteer moderator Project administrator
 Send message
Joined: 21 Jan 10 Posts: 13955 ID: 53948 Credit: 392,700,090 RAC: 171,419
                               
|
GDB,
I hid your post, and Rogue's reply to it, because your email address was included in the log you included.
Rogue's reply to you was:
Copy wwwwcl to another directory then run the range manually.
____________
My lucky number is 75898524288+1 |
|
|
Dave  Send message
Joined: 13 Feb 12 Posts: 3171 ID: 130544 Credit: 2,233,579,151 RAC: 561,339
                           
|
Why o why does this keep stopping:
[2014-04-11 20:02:06 GDT] : Could not find completion line in log file [wwww.log]. Assuming user stopped with ^C
which I did not.
[2014-04-11 20:02:06 GDT] Total Time: 2:03:06 Total Work Units: 37 Special Results Found: 0
[2014-04-11 20:02:06 GDT] WALLSUNSUN: Returning work to server prpnet.primegrid.com at port 13001
[2014-04-11 20:02:06 GDT] WALLSUNSUN: INFO: Test for range 32133210000000000:32133220000000000 was accepted
[2014-04-11 20:02:07 GDT] WALLSUNSUN: INFO: Test for range 32133220000000000:32133230000000000 was abandoned
[2014-04-11 20:02:07 GDT] WALLSUNSUN: INFO: All 2 test results were accepted
[2014-04-11 20:02:07 GDT] Client shutdown complete
Nothing else going e.g BOINC closed. Makes no difference. Kept deleting the logs files on both folders. |
|
|
rogueVolunteer developer
 Send message
Joined: 8 Sep 07 Posts: 1255 ID: 12001 Credit: 18,565,548 RAC: 0
 
|
Why o why does this keep stopping:
[2014-04-11 20:02:06 GDT] : Could not find completion line in log file [wwww.log]. Assuming user stopped with ^C
which I did not.
[2014-04-11 20:02:06 GDT] Total Time: 2:03:06 Total Work Units: 37 Special Results Found: 0
[2014-04-11 20:02:06 GDT] WALLSUNSUN: Returning work to server prpnet.primegrid.com at port 13001
[2014-04-11 20:02:06 GDT] WALLSUNSUN: INFO: Test for range 32133210000000000:32133220000000000 was accepted
[2014-04-11 20:02:07 GDT] WALLSUNSUN: INFO: Test for range 32133220000000000:32133230000000000 was abandoned
[2014-04-11 20:02:07 GDT] WALLSUNSUN: INFO: All 2 test results were accepted
[2014-04-11 20:02:07 GDT] Client shutdown complete
Nothing else going e.g BOINC closed. Makes no difference. Kept deleting the logs files on both folders.
Does it delete the log file before it shuts down? I would like to know what is in the log file.
Do only certain ranges fail or does this appear to be random? |
|
|
Dave  Send message
Joined: 13 Feb 12 Posts: 3171 ID: 130544 Credit: 2,233,579,151 RAC: 561,339
                           
|
Does it delete the log file before it shuts down?
No.
I would like to know what is in the log file.
As pasted above.
Do only certain ranges fail or does this appear to be random?
No idea. Not able to monitor it that closely as have major things on my plate at this time. Appears random.
|
|
|
|
Dave, I didn't read all of your posts so this may have already been answered but have you over clocked your card? I ask because for some reason WSS is very sensitive to over clocked cards on my system. WFS on the other hand loves the overclocking. Anyway, when I had my cards over clocked alot (more than 50% above normal) I'd get errors much like you are reporting. I can overclock maybe 10% and get it to run stable.
Like I said however, I may be way behind the times on your issue. |
|
|
JimB Honorary cruncher Send message
Joined: 4 Aug 11 Posts: 918 ID: 107307 Credit: 977,945,376 RAC: 0
                     
|
Stupid question, but you're not trying to run two clients in the same directory, are you? |
|
|
Lumiukko Volunteer tester Send message
Joined: 7 Jul 08 Posts: 165 ID: 25183 Credit: 870,450,997 RAC: 38,672
                           
|
Do you have more than one GPU?
Do you have "platform" and/or "device" defined in wwww.ini?
I was getting that same error:
"Could not find completion line in log file [wwww.log]. Assuming user stopped with ^C"
until I added "platform=0" into wwww.ini.
--
Lumiukko |
|
|
Dave  Send message
Joined: 13 Feb 12 Posts: 3171 ID: 130544 Credit: 2,233,579,151 RAC: 561,339
                           
|
Thanks for answers.
No overclocking - factory oc is 783 & it was this but now revved down to 750 last night so testing this now.
2 GPUs, 2 clients, 2 folders, -1 & -2.
Platform=0 is in wwww. |
|
|
Dave  Send message
Joined: 13 Feb 12 Posts: 3171 ID: 130544 Credit: 2,233,579,151 RAC: 561,339
                           
|
This has still behaved the same with the lowered clocks. Same log contents as supplied. Never seen a subproject that needs manual kicking.
Any further advice duely appreciated. |
|
|
rogueVolunteer developer
 Send message
Joined: 8 Sep 07 Posts: 1255 ID: 12001 Credit: 18,565,548 RAC: 0
 
|
This has still behaved the same with the lowered clocks. Same log contents as supplied. Never seen a subproject that needs manual kicking.
Any further advice duely appreciated.
What is in wwww.log when this happens? |
|
|
Dave  Send message
Joined: 13 Feb 12 Posts: 3171 ID: 130544 Credit: 2,233,579,151 RAC: 561,339
                           
|
Nothing. It's always 0 bytes. |
|
|
GDBSend message
Joined: 15 Nov 11 Posts: 284 ID: 119185 Credit: 3,931,363,173 RAC: 2,029,972
                      
|
Two out of 8 clients are having the same problem of not finishing, and then resuming. |
|
|
|
My MacBook Prod has a GT 650M. It isn't particularly fast, but it doesn't have any problems with this work unit.
I suggest that you upgrade to the latest OS X if you can.
That did it!
I was able to get OSX 10.9 on my old mac 1.1 and its working great.
Thank you all for your help!
____________
|
|
|
|
I'm getting this same "Could not find completion line in log file [wwww.log]. Assuming user stopped with ^C" error on just one of my two GPU clients.
They are running in two separate directories.
I do have "platform=0" defined in both wwww.ini files.
GPU0 is running W-S-S just fine. But GPU1 keeps stopping immediately with the wwww.log file complaint.
The wwww.log file is created as soon as I start prpclient.exe, and it is empty. I have tried deleting it and restarting prpclient.exe, but the same thing happens again.
Any ideas on why one client would run and the other not run?
Do you have more than one GPU?
Do you have "platform" and/or "device" defined in wwww.ini?
I was getting that same error:
"Could not find completion line in log file [wwww.log]. Assuming user stopped with ^C"
until I added "platform=0" into wwww.ini.
--
Lumiukko
____________
Proud member of Team Aggie the Pew
"Wir müssen wissen. Wir werden wissen."
"We must know, we shall know."
- David Hilbert, 1930 |
|
|
rogueVolunteer developer
 Send message
Joined: 8 Sep 07 Posts: 1255 ID: 12001 Credit: 18,565,548 RAC: 0
 
|
Run wwwwcl with the -l option. What does it show?
Does only one client fail or do both fail? |
|
|
|
a really silly question, but with 2 instances, the 2nd folder is the same as the 1st folder except for the ini files as far as all the files and subfolders go and you do have the permissions setup the same? I know, really silly questions. |
|
|
|
I'm getting this same "Could not find completion line in log file [wwww.log]. Assuming user stopped with ^C" error on just one of my two GPU clients.
They are running in two separate directories.
I do have "platform=0" defined in both wwww.ini files.
GPU0 is running W-S-S just fine. But GPU1 keeps stopping immediately with the wwww.log file complaint.
The wwww.log file is created as soon as I start prpclient.exe, and it is empty. I have tried deleting it and restarting prpclient.exe, but the same thing happens again.
Any ideas on why one client would run and the other not run?
SOLVED: It was a PEBKAC error. GPU1 was not fully plugged in after the move! The fan was running, but one of the two power connectors was not fully inserted. GPU1 was not even showing up in Device Manager. Once I got the power connector to "click," everything is back to normal and I now have no errors in wwwwcl64. *Headslap*
____________
Proud member of Team Aggie the Pew
"Wir müssen wissen. Wir werden wissen."
"We must know, we shall know."
- David Hilbert, 1930 |
|
|