SETI@Home Wow!-Event 2015

[biodoc]

It doesn't look like there will be any astropulse WUs for my linux GPUs this year so I moved my hardware around to run my GTX 780TI and 2x GTX 980s on my 2 machines that can run windows. I finally have all 3 GPUs running using the lunatics cuda 5.0 app with 2 WUs per GPU.

No cpus in play. It's hot enough in our family room!

Hopefully this will help.

 

[Pokey]

It doesn't look like there will be any astropulse WUs for my linux GPUs this year so I moved my hardware around to run my GTX 780TI and 2x GTX 980s on my 2 machines that can run windows. I finally have all 3 GPUs running using the lunatics cuda 5.0 app with 2 WUs per GPU.

No cpus in play. It's hot enough in our family room!

Hopefully this will help.

Great biodoc! Yeah, that should definitely help. :thumbsup:

 

[TennesseeTony]

Anyone care to post the (app_config.xml?) text needed to run dual WU per GPU? Single tasks are already utilizing my GPU at 91%, but there's that 9% sitting there, looking all lonely and unwanted.....

Finally hacked apart another 280X with a bad fan, and made it look just like it's twin. Will be adding it soon, since GLeeM's likely to pass me otherwise.

 

[biodoc]

Anyone care to post the (app_config.xml?) text needed to run dual WU per GPU? Single tasks are already utilizing my GPU at 91%, but there's that 9% sitting there, looking all lonely and unwanted.....

If you used the lunatics installer, you'll find their app_info.xml in the Seti project directory. \ProgramData\BOINC\projects\setiathome.berkeley.edu.

edit the file and change the "1" to "0.5" in <count>1</count> wherever you find it.

<coproc>
<type>CUDA</type>
<count>1</count>
</coproc>

change to:

<coproc>
<type>CUDA</type>
<count>0.5</count>
</coproc>

 

[TennesseeTony]

I found and changed all THIRTY entries to .5, no change even after a restart of BOINC.

Then I edited the app_config.xml to .5 and it worked. Not sure if both edits are required, but thanks for pointing me in the right direction! :thumbsup:

EDIT: But it did work on the 2nd computer, which did not have a app_config.xml. So you're 100% right. I guess Mjölnir had some left-over configuration, prior to the lunatics app installation.

 

[Pokey]

This will do it. If you do this, you don't have to worry about it any more because it doesn't get updated if you should update lunitics again.

------------------------------------------------------------------------
<app_config>
<app>
<name>setiathome_v7</name>
<gpu_versions>
<gpu_usage>0.50</gpu_usage>
<cpu_usage>0.50</cpu_usage>
</gpu_versions>
</app>
<app>
<name>astropulse_v6</name>
<gpu_versions>
<gpu_usage>0.5</gpu_usage>
<cpu_usage>1.0</cpu_usage>
</gpu_versions>
</app>
</app_config>
----------------------------------------------------------------------------

Then “save as”: app_config.xml and put the file in this folder:

C:\Program Data\Boinc\projects\setiathome.berkely.edu\

In BOINC Manager, find “Advanced” drop down menu and click on “read config files”.

You should be able to run three instances on the later cards. Just change the 0.5 to 0.333 in the multibeam section. Don't change Astropulse section.

 

[Orange Kid]

SETI@home Network Outage (August 17-18)
Due to network and DNS upgrades/changes we will be shutting down the projects tonight (August 17th) and we will be offline until the next morning.

Make sure you have enough to last

 

[TennesseeTony]

will be offline until the next morning

Ah ya gotta be Orange Kiddin'! Lol

 

[Pokey]

LOL
"until the next morning" is code for "whenever we get it back up".

It probably is faulty memory on my part but it seems as if something unforeseen happens each year. But it affects everyone so things have a way of working out.

I am set for 1.0 day plus .5 days, so if there are enough work units to go around.............

crunch on, crunch on

My Pisces group has sunk to last place.............that didn't take long. :\

 

[TennesseeTony]

My Pisces group has sunk to last place.............that didn't take long.

On the bright side, your TeAm has risen one place to 10th.

 

[Jalapeno]

I can't get any additional WUs in cache on one of my PCs.

I set additional work units to four days but nothing comes in. Uninstalled and re-installed BOINC - nothing.

Is there any third party software for this?

 

[TennesseeTony]

I can't get any additional WUs in cache on one of my PCs.

Quote: SETI@home Network Outage (August 17-18) Due to network and DNS upgrades/changes we will be shutting down the projects tonight (August 17th) and we will be offline until the next morning.

They are down for at least 4 hours, possibly longer, much longer. Check the server status of any project giving you trouble.

 

[Jalapeno]

They are down for at least 4 hours, possibly longer, much longer. Check the server status of any project giving you trouble.

I'm aware of that Tony.

My problem has been ongoing for at least a week. Any ideas?

 

[Pokey]

Jalapeno, I am not aware of any third party software. Sometimes they just run out of work suitable for your equipment. I ran into a similar situation myself getting set up for the event. I finally gave up and let things take their course and eventually work started coming in. The software is surprisingly self sufficient.
Of course right now, with them being down, is not a good time to work on problems.

Which computer are you having trouble with? Maybe we can tell something when their servers come back on line.

 

[biodoc]

Hopefully they will be back up soon 'cause I'm out of work. I had my boinc cache set for 2 days but perhaps GPU WUs are capped at a certain number?

 

[Pokey]

Now that you mention it, I seem to recall reading somewhere that they dictate how much cache you get now. Don't quote me though.

 

[TennesseeTony]

I was set for 10days and was out by this am on most machines.

 

[Rattledagger]

Last time I heard the limit was 100/computer + 100/gpu, meaning especially GPU you can quickly run-out of cached work.

 

[Orange Kid]

:whiste::whiste::whiste:

 

[biodoc]

Last time I heard the limit was 100/computer + 100/gpu, meaning especially GPU you can quickly run-out of cached work.

Sounds like what I have ready to report. 100 WUs on rig with 1 GPU and 200 WUs on rig with 2 GPUs.

The Servers are accepting uploads but task reporting and downloading new WUs are not possible yet.

Server

 

[biodoc]

Getting some work now. Hopefully more to come.

 

[Jalapeno]

They are out of WUs again...

 

[Pokey]

I am back in business now, hope everyone else is too.

 

[Jalapeno]

Which computer are you having trouble with? Maybe we can tell something when their servers come back on line.

It's this one: http://setiathome.berkeley.edu/show_host_detail.php?hostid=7378006

It shows me 129 tasks. I'm guessing that is current work plus cache.

The BOINC interface however shows me only the four the CPU is currently working on.


Edit: 129 is everything, (pending, etc...). I have 86 "In Progress" WUs at the moment.

 

[TennesseeTony]