Anything relating to AstroPulse tasks

Message boards : Number crunching : Anything relating to AstroPulse tasks
Message board moderation

To post messages, you must log in.

Previous · 1 . . . 12 · 13 · 14 · 15 · 16 · 17 · 18 . . . 120 · Next

AuthorMessage
Profile Brent Norman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester

Send message
Joined: 1 Dec 99
Posts: 2786
Credit: 685,657,289
RAC: 835
Canada
Message 1706643 - Posted: 30 Jul 2015, 10:18:33 UTC

Question for the tester's,

Do you have any use for about 95 AP files that got abandoned by my computer during a crash? I can throw them on my server if you can use them.

BOINC doesn't seem to be cleaning them up, so just been waiting for a good date spread of new/old so I don't delete the wrong ones.
ID: 1706643 · Report as offensive
Profile betreger Project Donor
Avatar

Send message
Joined: 29 Jun 99
Posts: 11360
Credit: 29,581,041
RAC: 66
United States
Message 1706727 - Posted: 30 Jul 2015, 19:58:22 UTC

Does anybody know if the APs that are out that claim the application is
Not in DB
will validate?
ID: 1706727 · Report as offensive
Profile Zalster Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 27 May 99
Posts: 5517
Credit: 528,817,460
RAC: 242
United States
Message 1706730 - Posted: 30 Jul 2015, 20:02:19 UTC - in response to Message 1706727.  
Last modified: 30 Jul 2015, 20:04:04 UTC

Someone claimed they had a "not in DB" AP that validated

Edit..

I have MBs that are "not in DB" that are validating
ID: 1706730 · Report as offensive
Profile HAL9000
Volunteer tester
Avatar

Send message
Joined: 11 Sep 99
Posts: 6534
Credit: 196,805,888
RAC: 57
United States
Message 1706745 - Posted: 30 Jul 2015, 20:21:47 UTC - in response to Message 1706727.  

Does anybody know if the APs that are out that claim the application is
Not in DB
will validate?

The message "not in db" is in the application field. From what I have seen it looks like there was an issue populating the name field for hosts using Anonymous platform. I would expect this to only be a cosmetic issue that seems to have already been resolved.
SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[
ID: 1706745 · Report as offensive
Profile Brent Norman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester

Send message
Joined: 1 Dec 99
Posts: 2786
Credit: 685,657,289
RAC: 835
Canada
Message 1706757 - Posted: 30 Jul 2015, 20:33:58 UTC - in response to Message 1706745.  

The only ones I can find on my computers are from

29 Jul 2015, 23:11:49 UTC to
30 Jul 2015, 0:29:33 UTC

And there are only 3 of them (that are in progress)

But it's not like I have a big dataset to compare to ... I'm almost dry :((
ID: 1706757 · Report as offensive
Profile Donald L. Johnson
Avatar

Send message
Joined: 5 Aug 02
Posts: 8240
Credit: 14,654,533
RAC: 20
United States
Message 1706879 - Posted: 31 Jul 2015, 1:57:38 UTC - in response to Message 1706745.  
Last modified: 31 Jul 2015, 2:00:26 UTC

Does anybody know if the APs that are out that claim the application is
Not in DB
will validate?

The message "not in db" is in the application field. From what I have seen it looks like there was an issue populating the name field for hosts using Anonymous platform. I would expect this to only be a cosmetic issue that seems to have already been resolved.

Here's a message from Jord in the Panic_Mode_On thread that explains the "Not in DB" comment.

OBTW, my i5/Win7 box got 6! APs at 31Jul2015 0123 UTC. Most I've had on that box at any one time since.......
Donald
Infernal Optimist / Submariner, retired
ID: 1706879 · Report as offensive
Profile Jimbocous Project Donor
Volunteer tester
Avatar

Send message
Joined: 1 Apr 13
Posts: 1849
Credit: 268,616,081
RAC: 1,349
United States
Message 1706916 - Posted: 31 Jul 2015, 6:34:06 UTC - in response to Message 1706909.  

Does anybody know if the APs that are out that claim the application is
Not in DB
will validate?

So far all mine have . . .
ID: 1706916 · Report as offensive
Profile Todderbert
Avatar

Send message
Joined: 17 Jun 99
Posts: 221
Credit: 53,153,779
RAC: 0
United States
Message 1707028 - Posted: 31 Jul 2015, 15:57:57 UTC
Last modified: 31 Jul 2015, 15:58:35 UTC

My rig with two 750Tis locked up again this morning. It only does this when running two APs at a time.
My command line is as follows:
-use_sleep -unroll 10 -oclFFT_plan 256 16 512 -ffa_block 12288 -ffa_block_fetch 6144

Are there more conservative parameters that can be changed for this line?
Should I just change my AP tasks to one per gpu?

When I get home I'll be able to reboot the machine and if I can't find a stable solution for these AP tasks I may just remove astropluse and just crunch MBs on this machine.

Thanks for any help,
Todderbert
ID: 1707028 · Report as offensive
Profile Jeff Buck Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester

Send message
Joined: 11 Feb 00
Posts: 1441
Credit: 148,764,870
RAC: 0
United States
Message 1707031 - Posted: 31 Jul 2015, 16:05:16 UTC - in response to Message 1707028.  

Should I just change my AP tasks to one per gpu?

For my 750Tis I have my app_config.xml set up to run either 2 MB tasks or 1 MB plus 1 AP, never 2 APs at the same time.
Set <gpu_usage>.49</gpu_usage> for MB and <gpu_usage>.51</gpu_usage> for AP.
ID: 1707031 · Report as offensive
Profile Zalster Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 27 May 99
Posts: 5517
Credit: 528,817,460
RAC: 242
United States
Message 1707035 - Posted: 31 Jul 2015, 16:27:59 UTC - in response to Message 1707031.  

I run 2 APs at a time but I use lower Commandline values. I also limit CPU usage.
ID: 1707035 · Report as offensive
TBar
Volunteer tester

Send message
Joined: 22 May 99
Posts: 5204
Credit: 840,779,836
RAC: 2,768
United States
Message 1707038 - Posted: 31 Jul 2015, 16:39:20 UTC

If I run just One AP on the 750Ti in My Mac with -ffa_block 8192 -ffa_block_fetch 4096 the machine will immediately Reboot. It doesn't have that problem with lower numbers. I'd say the 750Ti doesn't like those higher numbers.
ID: 1707038 · Report as offensive
WezH
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 576
Credit: 67,033,957
RAC: 95
Finland
Message 1707052 - Posted: 31 Jul 2015, 17:12:38 UTC - in response to Message 1707028.  

My rig with two 750Tis locked up again this morning. It only does this when running two APs at a time.
My command line is as follows:
-use_sleep -unroll 10 -oclFFT_plan 256 16 512 -ffa_block 12288 -ffa_block_fetch 6144

Are there more conservative parameters that can be changed for this line?
Should I just change my AP tasks to one per gpu?

When I get home I'll be able to reboot the machine and if I can't find a stable solution for these AP tasks I may just remove astropluse and just crunch MBs on this machine.


My two 750ti host has this cmd line:

-use_sleep -unroll 15 -oclFFT_plan 256 16 256 -ffa_block 16384 -ffa_block_fetch 8192 -tune 1 64 4 1 -tune 2 64 4 1 -hp


But I run only one AP at time. I did test running two, but no go, same problems as You did have, host completely freezed...
ID: 1707052 · Report as offensive
Profile Zalster Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 27 May 99
Posts: 5517
Credit: 528,817,460
RAC: 242
United States
Message 1707063 - Posted: 31 Jul 2015, 17:49:31 UTC - in response to Message 1707052.  
Last modified: 31 Jul 2015, 18:23:17 UTC

Here is my commandline

-use_sleep -unroll 10 -oclfft_plan 256 16 256 -ffa_block 8192 -ffa_block_fetch 4096 -tune 1 64 4 1 -tune 2 64 4 1

I think the reason for the lock up are too many task at a time. IE Not enough resources.

I also removed the -hp

In my 4 core Intel, it has 1 -750 and it runs 3 CPU work and 2 GPU work units.

On my other 4 core AMD, it has 2-750s so it runs 4 GPU work units and I limit the CPU to 1 work unit for total of 5. If I didn't then it would try to run 2 core work units and 4 GPU work units, for a total of 6.

I think that was causing the lock up on that system.

I installed a <project_max_concurrent> to the app_config.xml in that one

Something to look at.

Edit..

Almost forgot. I also installed Process Lasso just in case.
ID: 1707063 · Report as offensive
OTS
Volunteer tester

Send message
Joined: 6 Jan 08
Posts: 369
Credit: 20,533,537
RAC: 0
United States
Message 1707080 - Posted: 31 Jul 2015, 19:47:51 UTC - in response to Message 1707028.  

My rig with two 750Tis locked up again this morning. It only does this when running two APs at a time.
My command line is as follows:
-use_sleep -unroll 10 -oclFFT_plan 256 16 512 -ffa_block 12288 -ffa_block_fetch 6144

Are there more conservative parameters that can be changed for this line?
Should I just change my AP tasks to one per gpu?

When I get home I'll be able to reboot the machine and if I can't find a stable solution for these AP tasks I may just remove astropluse and just crunch MBs on this machine.

Thanks for any help,
Todderbert



I have been using the exact same parameters you have been using. When I was using version 346.47 of Nvidia’s driver it would lock up almost every night when running one WU. Several weeks back in frustration I switched to driver version 346.59 and I thought I had the problem cured but it locked up again last night for the first time since the driver change. If it happens again I will probably try the parameters of WezH or Zalster and see if there is any difference.

The funny thing is that in some cases some other processes stopped working before the whole machine became unresponsive, e.g., I have had sendmail stop working, or lose network connectivity first. The converse is also true, SETI can lock up while sendmail continues to work for a few hours. I thought at first it was a hardware problem but the lockups continued even after the mandatory dust bunny check and the reseating of all the cables and memory. They only went away after I changed the driver – until last night that is :( . Perhaps it was a one time occurrence -- I can always hope.

In any case, perhaps a driver change will help with your problem.
ID: 1707080 · Report as offensive
Profile Todderbert
Avatar

Send message
Joined: 17 Jun 99
Posts: 221
Credit: 53,153,779
RAC: 0
United States
Message 1707219 - Posted: 1 Aug 2015, 2:56:44 UTC

Thanks for the information. The settings I choose to use were a mix. Using Zalster's command line, and Jeff Buck's app_config, which I applied in the app_info. The system is running smooth right now. I have max_concurrent set to six. My cpu utilization with the gpus running one MB and one AP each is around 50%.

Now here's a good question, has anyone had a 960 lock up from running three APs at a time? I ask this because I have a machine with two 970's and a 960, with MBs and APs set to .33, it hasn't locked up yet. I really like these 960s, there like little tanks.

On a side note I picked up a Titan Black to mess with, I'm guessing the settings for a 980/970 would be ideal for it. Anyone run one of these cards? its a hot little monster, but it really idles down when crunching APs.
ID: 1707219 · Report as offensive
Profile Jimbocous Project Donor
Volunteer tester
Avatar

Send message
Joined: 1 Apr 13
Posts: 1849
Credit: 268,616,081
RAC: 1,349
United States
Message 1707271 - Posted: 1 Aug 2015, 7:16:25 UTC - in response to Message 1707080.  
Last modified: 1 Aug 2015, 7:17:20 UTC

In any case, perhaps a driver change will help with your problem.

Fwiw, I'm still running 334.89 on my 750ti's, due to issues I heard about on later versions. Learned a while back that latest and greatest was a bad plan on these. As far as params, no tuning here; just whatever the Lunatics installer gave me. I do run two tasks per GPU. Have not experienced a lock-up on these.
ID: 1707271 · Report as offensive
Profile Mike Special Project $75 donor
Volunteer tester
Avatar

Send message
Joined: 17 Feb 01
Posts: 34253
Credit: 79,922,639
RAC: 80
Germany
Message 1707273 - Posted: 1 Aug 2015, 8:39:22 UTC

On a side note I picked up a Titan Black to mess with, I'm guessing the settings for a 980/970 would be ideal for it. Anyone run one of these cards? its a hot little monster, but it really idles down when crunching APs.


Yes, probably you can increase -unroll even more on that card.


With each crime and every kindness we birth our future.
ID: 1707273 · Report as offensive
Profile Brent Norman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester

Send message
Joined: 1 Dec 99
Posts: 2786
Credit: 685,657,289
RAC: 835
Canada
Message 1707293 - Posted: 1 Aug 2015, 10:57:06 UTC

I finally made it to that hard to find number ... I'm full of AP tasks :)
ID: 1707293 · Report as offensive
Profile Cliff Harding
Volunteer tester
Avatar

Send message
Joined: 18 Aug 99
Posts: 1432
Credit: 110,967,840
RAC: 67
United States
Message 1707309 - Posted: 1 Aug 2015, 12:20:44 UTC - in response to Message 1707052.  

My two 750ti host has this cmd line:

-use_sleep -unroll 15 -oclFFT_plan 256 16 256 -ffa_block 16384 -ffa_block_fetch 8192 -tune 1 64 4 1 -tune 2 64 4 1 -hp


But I run only one AP at time. I did test running two, but no go, same problems as You did have, host completely freezed...


I'm running 2 tasks each on 2x 750TIs basically use the same, except the -unroll is 10 instead of 15, and have no problems. In my app_config.xml, I specify .5 for both CPU & GPU. Est. run time for each is 1.12 hrs. each. The only time the host freezes is for overheating if ambient temp reaches high 80s low 90s, then I stop crunching.


I don't buy computers, I build them!!
ID: 1707309 · Report as offensive
Profile betreger Project Donor
Avatar

Send message
Joined: 29 Jun 99
Posts: 11360
Credit: 29,581,041
RAC: 66
United States
Message 1707533 - Posted: 2 Aug 2015, 2:00:26 UTC

With only 152 channels to do on AP it seems doubtful that this run will last thru Sunday. It has been fun.
ID: 1707533 · Report as offensive
Previous · 1 . . . 12 · 13 · 14 · 15 · 16 · 17 · 18 . . . 120 · Next

Message boards : Number crunching : Anything relating to AstroPulse tasks


 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.