Welcome to MilkyWay@home

Posts by europa

1) Message boards : Number crunching : Linux Mint Issue (Message 64358)
Posted 3 Mar 2016 by europa
Post:
One other possibility might be to deselect the "CPU task" box since I don't believe the N-body WU's use the GPU.

Regards,
Steve
2) Message boards : Number crunching : Issues With Milkyway & Linux (Message 64327)
Posted 8 Feb 2016 by europa
Post:
I'm running all my machines under Linux Mint 17.3. Three boxes are Nvidia and one is ATI.

In the past, people would use an app_info.xml file to customize the running of multiple simultaneous GPU WU's, for example. They could get pretty convoluted and I'm no programmer.

A couple of years back there was some movement to the simpler app_config.xml format and I've gone that route with my machines with no problems.

Here's mine for MW:
<app_config>

<app>
<name>milkyway</name>
<max_concurrent>8</max_concurrent>
<gpu_versions>
<gpu_usage>.5</gpu_usage>
<cpu_usage>.05</cpu_usage>
</gpu_versions>
</app>

<app>
<name>milkyway_separation__modified_fit</name>
<max_concurrent>8</max_concurrent>
<gpu_versions>
<gpu_usage>.5</gpu_usage>
<cpu_usage>.05</cpu_usage>
</gpu_versions>
</app>

</app_config>

Another thing to check is the cc_config.xml since that's where you tell it to use all GPU's. Boinc defaults to no GPUs.

Here's mine:
<cc_config>
<options>
<use_all_gpus>1</use_all_gpus>
<save_stats_days>180</save_stats_days>
<max_file_xfers>48</max_file_xfers>
<max_file_xfers_per_project>36</max_file_xfers_per_project>
</options>
<log_flags>
<task>1</task>
<file_xfer>1</file_xfer>
<sched_ops>1</sched_ops>
</log_flags>
</cc_config>

Finally, sometimes I've found it necessary to do the following in the terminal:

xhost +

and then:

sudo service boinc-client restart

I don't know if any of this will help. Sometimes I feel like I'm just doing the digital version of uttering incantations and shaking rattles!! :)

Regards,
Steve
3) Message boards : News : New Nbody Version 1.50 (Message 63629)
Posted 24 May 2015 by europa
Post:
I'm running Linux Mint and after an initial run of WU's running but getting "validation inconclusive" for them, then I started having the old problem of asymtotic WU's (approaching infinitely close but never completing).

I ended up aborting all of the wu's before I noticed that some were 1.50 and some were 1.50 (mt). I don't know if there was any pattern to the one's that just keep running and the one's that run but give the "validation incomplete" message.

I've delisted the N-body Simulation from my machines for now.

Regards,
Steve
4) Message boards : News : New Nbody Version 1.50 (Message 63618)
Posted 18 May 2015 by europa
Post:
Just to provide some extra information on troubleshooitng the N-body wu's....

I'm running Linux Mint 17.1 (Rebecca) on a 5 core AMD with ATI 7850 card.

Unlike previously, the N-body wu's are completing, however, I just checked and they are all "validation inconclusive."

I also notice that it only processes 1 N-body wu at a time and says it is using 4 cores. This seems to apply ONLY to the MW wu's since the Einstein GPU wu's are processing multiple WU's simultaneously and concurrently with the MW and say they're using the equivalent of 2 cores!

Using 6 cores simultaneously on a 5 core cpu!

Ya gotta love the math!! :)

Regards,
Steve
5) Message boards : News : New Nbody version 1.48 (Message 63229)
Posted 14 Mar 2015 by europa
Post:
My N-body WU's start off really cranking but gradually slow down as more and more of the WU is completed.

Finally, after additional hours and hours, they're at 99.999% and become asymtotic WU's...........approaching infinitely close but never actually completing!

Right now, they're just wasting alot of computing power and electricity.

Regards,
Steve
6) Message boards : Number crunching : Updated GPU Requirements (Currently not supporting GPU tasks) (Message 63097)
Posted 2 Feb 2015 by europa
Post:
Arkayn,

Thanks so much for the help. I'll get this onto my ATI machine right away.

Thanks again,
Steve
7) Message boards : Number crunching : ATI 7950 and 7850 not getting GPU WU's (Message 63096)
Posted 2 Feb 2015 by europa
Post:
Arkayn,

Thanks so much for the help. I'll get this onto the ATI machine. Hopefully, I can get through all of the E@H WUs quickly so that MW@H will say that I need work and send me some.

Thanks again,
Steve
8) Message boards : Number crunching : Updated GPU Requirements (Currently not supporting GPU tasks) (Message 63092)
Posted 2 Feb 2015 by europa
Post:
Well, I think I finally figured out why MW@H isn't sending me any GPU WU's on my ATI machine. It see's my HD7950 and HD7850 as:

"Message from server: ATI GPU R600 (R38xx) does not support OpenCL"

I'm using the Catalyst 14.4 driver and E@H is going like gangbusters crunching multiple GPU WU's per card.

Basically, E@H has this machine all to itself right now and would like to thank MW@H for their generosity!

Regards,
Steve


See my reply in your thread.
http://milkyway.cs.rpi.edu/milkyway/forum_thread.php?id=3710&postid=63088#63088



Arkayn,

Thanks for the suggestion but I'm on Linux Mint (17.1), not Windows.

Regards,
Steve
9) Message boards : Number crunching : Updated GPU Requirements (Currently not supporting GPU tasks) (Message 63086)
Posted 1 Feb 2015 by europa
Post:
Well, I think I finally figured out why MW@H isn't sending me any GPU WU's on my ATI machine. It see's my HD7950 and HD7850 as:

"Message from server: ATI GPU R600 (R38xx) does not support OpenCL"

I'm using the Catalyst 14.4 driver and E@H is going like gangbusters crunching multiple GPU WU's per card.

Basically, E@H has this machine all to itself right now and would like to thank MW@H for their generosity!

Regards,
Steve
10) Message boards : Number crunching : ATI 7950 and 7850 not getting GPU WU's (Message 63082)
Posted 28 Jan 2015 by europa
Post:
You might try reading this thread over at Einstein too:
http://einstein.phys.uwm.edu/forum_thread.php?id=11057

They had similar problems but were talking about an Nvidia card and a special command that was needed now.[/quote]

Thanks for the heads-up. I just saw it and will give it a try. I also have a new Nvidia Boinc machine I'd like to try it on.

The thing I don't understand with this ATI setup is why E@H has been crunching away with no problems while MW@H and Seti won't send me any GPU WU's. That would seem to rule out a problem on the ATI side.

Regards,
Steve
11) Message boards : Number crunching : ATI 7950 and 7850 not getting GPU WU's (Message 63076)
Posted 25 Jan 2015 by europa
Post:
Are you getting any error messages about why you are not receiving GPU work?


Arkayn,

I checked the event log and there aren't any error messages. I went onto my MW@H and Seti preferences pages and even de-selected the option for cpu units to try to force some gpu WU's but no luck.

I'm using the Catalyst 14.10 drivers and E@H seems happy as a clam. But, Seti is like MW; no gpu units.

I just had an idea. I'll switch to a different HD and re-install the OS (Linux Mint 17.1) and then install the Catalyst 13.x drivers that I have archived instead of the Catalyst 14.10 and see if that makes a difference. If it doesn't, I don't lose anything.

Regards,
Steve
12) Message boards : Number crunching : ATI 7950 and 7850 not getting GPU WU's (Message 63062)
Posted 23 Jan 2015 by europa
Post:
Here's the app_config.xml

<app_config>
<app>
<name>milkyway</name>
<gpu_versions>
<gpu_usage>.5</gpu_usage>
<cpu_usage>.05</cpu_usage>
</gpu_versions>
</app>
<app>
<name>milkyway_separation__modified_fit</name>
<gpu_versions>
<gpu_usage>.5</gpu_usage>
<cpu_usage>.05</cpu_usage>
</gpu_versions>
</app>
</app_config>

Thanks,
Steve
13) Message boards : Number crunching : ATI 7950 and 7850 not getting GPU WU's (Message 63061)
Posted 23 Jan 2015 by europa
Post:
I just converted a machine with an ATI HD7950 and HD7850 GPUs over to BOINC but am not receiving any GPU work units. I checked and both are DP-capable so that's not it. I've checked my MW computing prefs and everything looks fine. The machine in question is M6400T.

The machine is happily crunching E@H GPU WU's so it's not a question of BOINC not seeing the GPUs.

All of my other machines are NVIDIA and are doing fine.

Any help would be appreciated.

Regards,
Steve


Can you post the first 20 lines of your startup log?


You bet. Here they are.

Fri 23 Jan 2015 02:15:17 PM EST | | Starting BOINC client version 7.2.42 for x86_64-pc-linux-gnu
Fri 23 Jan 2015 02:15:17 PM EST | | log flags: file_xfer, sched_ops, task
Fri 23 Jan 2015 02:15:17 PM EST | | Libraries: libcurl/7.35.0 OpenSSL/1.0.1f zlib/1.2.8 libidn/1.28 librtmp/2.3
Fri 23 Jan 2015 02:15:17 PM EST | | Data directory: /home/tangosite/BOINC/BOINC
Fri 23 Jan 2015 02:15:17 PM EST | | OpenCL: AMD/ATI GPU 0: Tahiti (driver version 1214.3 (VM), device version OpenCL 1.2 AMD-APP (1214.3), 2949MB, 2949MB available, 2072 GFLOPS peak)
Fri 23 Jan 2015 02:15:17 PM EST | | OpenCL: AMD/ATI GPU 1: Pitcairn (driver version 1214.3 (VM), device version OpenCL 1.2 AMD-APP (1214.3), 1929MB, 1929MB available, 1648 GFLOPS peak)
Fri 23 Jan 2015 02:15:17 PM EST | | OpenCL CPU: AMD Phenom(tm) II X6 1090T Processor (OpenCL driver vendor: Advanced Micro Devices, Inc., driver version 1214.3 (sse2), device version OpenCL 1.2 AMD-APP (1214.3))
Fri 23 Jan 2015 02:15:17 PM EST | | Host name: M6400T
Fri 23 Jan 2015 02:15:17 PM EST | | Processor: 6 AuthenticAMD AMD Phenom(tm) II X6 1090T Processor [Family 16 Model 10 Stepping 0]
Fri 23 Jan 2015 02:15:17 PM EST | | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm 3dnowext 3dnow constant_tsc rep_good nopl nonstop_tsc extd_apicid aperfmperf pni monitor cx16 popcnt lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs skinit wdt cpb hw_pstate npt lbrv svm_lock nrip_save pausefilter
Fri 23 Jan 2015 02:15:17 PM EST | | OS: Linux: 3.13.0-37-generic
Fri 23 Jan 2015 02:15:17 PM EST | | Memory: 15.67 GB physical, 16.00 GB virtual
Fri 23 Jan 2015 02:15:17 PM EST | | Disk: 277.29 GB total, 255.41 GB free
Fri 23 Jan 2015 02:15:17 PM EST | | Local time is UTC -5 hours
Fri 23 Jan 2015 02:15:17 PM EST | SETI@home | Found app_config.xml
Fri 23 Jan 2015 02:15:17 PM EST | SETI@home | Your app_config.xml file refers to an unknown application 'setiathome_v7.01'. Known applications: 'setiathome_v7'
Fri 23 Jan 2015 02:15:17 PM EST | SETI@home | Your app_config.xml file refers to an unknown application 'setiathome_v7.04'. Known applications: 'setiathome_v7'
Fri 23 Jan 2015 02:15:17 PM EST | SETI@home | Your app_config.xml file refers to an unknown application 'astropulse_v7.04'. Known applications: 'setiathome_v7'
Fri 23 Jan 2015 02:15:17 PM EST | SETI@home | Your app_config.xml file refers to an unknown application 'astropulse_v7.1'. Known applications: 'setiathome_v7'
Fri 23 Jan 2015 02:15:17 PM EST | Milkyway@Home | Found app_config.xml
Fri 23 Jan 2015 02:15:17 PM EST | | Config: use all coprocessors
Fri 23 Jan 2015 02:15:17 PM EST | SETI@home | URL http://setiathome.berkeley.edu/; Computer ID 7481448; resource share 33
Fri 23 Jan 2015 02:15:17 PM EST | Einstein@Home | URL http://einstein.phys.uwm.edu/; Computer ID 11723989; resource share 33
Fri 23 Jan 2015 02:15:17 PM EST | Milkyway@Home | URL http://milkyway.cs.rpi.edu/milkyway/; Computer ID 606499; resource share 100

Also, here's the app_info.xml that I'm using successfully with the Nvidia machines, but not with the ATI.

Thanks for the help.

Regards,
Steve
14) Message boards : Number crunching : ATI 7950 and 7850 not getting GPU WU's (Message 63059)
Posted 23 Jan 2015 by europa
Post:
I just converted a machine with an ATI HD7950 and HD7850 GPUs over to BOINC but am not receiving any GPU work units. I checked and both are DP-capable so that's not it. I've checked my MW computing prefs and everything looks fine. The machine in question is M6400T.

The machine is happily crunching E@H GPU WU's so it's not a question of BOINC not seeing the GPUs.

All of my other machines are NVIDIA and are doing fine.

Any help would be appreciated.

Regards,
Steve
15) Message boards : Number crunching : Please assist - can't get Radeon HD 4850 recognized in Linux (Message 59386)
Posted 17 Jul 2013 by europa
Post:
Shodan,

Yep, that's the thing about pain.....it feels so good when it stops!! :)

Regards,
Steve
16) Message boards : Number crunching : Please assist - can't get Radeon HD 4850 recognized in Linux (Message 59265)
Posted 7 Jul 2013 by europa
Post:
Shodan,

I'm running Mint 14 and 15 on my machines but with Nvidia cards. See if this helps.

1. Before you even start up the Boinc client, open the terminal window and type:

sudo service boinc-client restart

It may take a few seconds but then you should get 3 status lines all ending in "OK"

Then close the terminal and start your Boinc client and look under the event log tab. One of the early entries will tell you if your graphics cards are being seen.

Also, in the cc_config.xml after <options> be sure to add
<use_all_gpus>1</use_all_gpus>

Hope that helps.

Regards,
Steve


17) Message boards : Number crunching : MilkyWay not up loading new workunits (Message 59175)
Posted 30 Jun 2013 by europa
Post:
I'm still having a problem on two of my three machines (ID: 518375 and ID: 518536). I have to keep hitting "update" every hour to get any new WU's and even then I only receive 5 or 6 which I burn right through. My global settings are all for 5 days + 1.



If it hasn't been suggested, try changing your cache settings to 5 days

+ max additional work buffer 0.01 . That way it will ask


for work more often.


Bif,

Thanks, my settings have been 5 days + 1 day. They just seem to be ignored. I'll try changing that second setting and see if it helps.

Regards,
Steve
18) Message boards : Number crunching : MilkyWay not up loading new workunits (Message 59169)
Posted 29 Jun 2013 by europa
Post:
I believe what you are seeing is the following:

1) The Milkway applications only allow a small number of tasks in progress.
2) Often a request for new tasks results in a no tasks available message.
3) The scheduler always tries to keep the cache full on the client.

As a result, when no tasks are available, the scheduler will fill the queue with tasks from the other projects, hence the don't need any message. On my systems this always results in Milkway getting swamped out by the other projects. My solution is to run each project solo. Einstein and Seti used to run fairly well together but, as the GPU apps have matured, even that is getting more complicated. One app seems to get preference over the other when running more than 1x.

Just some observations from my systems.


Gary,

Thanks for your insights. It sounds like this will only get worse as projects add more GPU apps. It also sounds as if this is something that can only be fixed by the people at BOINC Central or MW Central, probably on the server side. It kind of sounds like we're beyond app_config.xml tweaks.

I had been thinking about adding a second graphics card to the machine with the Fermi 550TI but I'm now questioning if I should waste the money on what will be an underutilized card.

Thanks again for the feedback.

Regards,
Steve
19) Message boards : Number crunching : MilkyWay not up loading new workunits (Message 59164)
Posted 29 Jun 2013 by europa
Post:
Well, II'm still having the problem in varying degees on all of my machines. Now on machine ID: 518884, when I hit "update" it says "no work sent, don't need it" even though I don't have any WU's at all.

On machine ID: 518375, I'm getting the same message even though I have only 1 task.

And on machine ID: 518536, I'm back to getting only a couple of tasks at a time, even though I only have about 7 in the cache.

In all cases, if I don't hit "update", they just sit on my machine until I do.

The only solution (and it's a temporary one at that) is to run down the cache, delete everything MilkyWay in the Boinc and Projects directories, re-start the service, re-start the client and hit "synchonize" (with BAM). That seems to get me a good fill but then things start to run down over time again.

Since Seti, Cosmology, and Einstein aren't having this problem, I don't think it's a BOINC problem.

I don't even have a clue as to what to try next.

Am I the only person still having this problem?

Regards,
Steve
20) Message boards : Number crunching : MilkyWay not up loading new workunits (Message 59093)
Posted 25 Jun 2013 by europa
Post:
[/quote]

You are using an app_config file, not a cc_config file, your list above shows that. But actually since you are getting some files then nothing is wrong with what you have, your problem, now that I reread it, is a lack of enough units, not lack of any units. And that I have no idea about, I just checked your pc's and all but one have a small cache of units on them.[/quote]

Mikey,

That's it exactly.

Although, now when I checked on startup this morning, the machine that was light now has a larger cache of WU's. Hopefully, it will increase further.

Regards,
Steve


Next 20

©2024 Astroinformatics Group