Pirates Ahoy!

platypus said:
Does it need 1gb of free ram to even start running? In which case either not bother with the linux box, or just wack more ram in it?
Click to expand...
I'd recommend 256MB for Linux, 256MB for X (if you're using it), and 256MB for each CPU core for Rosetta@home. Any CPU 1.6GHz with 512KB L2 cache or better will run Rosetta@home, but you should bear in mind the cost/performance - a PC at that spec will net you 80 points/day. A Q6600 will get you at least 1200.

platypus said:
Also, (sorry to keep asking!), does disabling the graphical aspect turn around quicker result yields?
Click to expand...
It'll help a little, but not much.

(EG I know that running SETI@Home via the command line is a lot faster then running the gui + screensaver).
Click to expand...
Used to be true in the bad old days, but since the screensaver got DirectXified, there's no longer that much difference (unless of course you're using the CUDA optimized client).
 
Yeah, watch out. Even well-built laptops can get silly hot. I have a laptop 'in the game', but it's set up to use 66% of one core. That's as high as it'll go without the fan going crazy.
 
I've just quickly ran some numbers. So far today, the gap has widened, and if we can keep this up, we could unstomp ourselves from my old team (The Planetary Society). On a personal level, that's be rather sweet. :D
 
It'll sort itself out if left alone, but you can 'kick' it if you so desire. Go to the 'projects' tab, click on the project, click 'update'.
 
It does. The tasks should be gone and the timer will run down on its own. That always happens and as said it's to stop people 'spamming' the projects.
 
ChrissyT88 said:
Okey dokey, got it running i think. I must admit i quite like the client interface - F@H really needs a piece of managemnet software like this! I have no idea what im doing though! I want to run two cores of Rosetta and two for F@H for the time being, but Rosetta wants them all! I have suspended some of the processes (and all the others it tries to start when i suspend some), and just have two running on the cores i chose for now. Is there any way to set it to just use two cores by default and not download anymore work until i have finished the unit like F@H? Also, what should i set the target CPU run time to?

Cheers

EDIT: I think i have found some of the options in the preferences section, but still not sure how i get it to download work only when i finnish the old WU...?
Click to expand...
Go into computing preferences under 'your account' on the site and limit number of CPUs to two (then 'update' the project from within BOINC Manager to register the change), or from within BOINC Manager you can set local preferences to use only 50% of available CPUs. You can then unsuspend those WUs.
 
ChrissyT88 said:
It says this in the event logs:

The VMware Server Web Access service terminated unexpectedly. It has done this 1 time(s). The following corrective action will be taken in 60000 milliseconds: Restart the service.

And thats it... The same message, lots of times!
Click to expand...
What OS is the host running?
 
ChrissyT88 said:
Sorry - the host is running Vista 64. I'll try your link biffa, cheers.

EDIT: Thanks for the input guys, but i really need to go to bed! This will have to wait until tomorrow... I can see a dreaded OS reinstall on the horizon...!
Click to expand...
After trying Biffa's solution, have a look at recent log files in C:\ProgramData\VMWare\VMWare Server - and especially the ones in the tomcat-logs folder. They might shed some light on the problem.

If you can't get to the bottom of which logs you should look at, delete the old ones and try starting the service again. It might create new logs for you to look at.
 
I'd encourage people to hang around a while. The reason is simple. If people leave en masse we'll be back where we started within a week or two.

I'm not asking people to hang around forever, but we shouldn't capitulate so easily. :)
 
ChrissyT88 said:
Aside: Anyone have a paid for copy of workstation...? I can get it as a student, though i am not sure its worth all the money... My guess is probably not?
Click to expand...
It's only worth it for not having that horrendous web interface. Otherwise I've never done anything with Workstation that can't be done with Server (there are features - I just haven't found use for them).
 
Just underclocked my laptop (by setting power management options that enforce a SpeedStep upper limit of 80%), and let BOINC use 100% of one core. So, I've gained 15% extra performance and the temps are 2C lower. Result.

Most of what I do is disk-limited rather than CPU-limited, so I don't mind leaving it set that way. I might even fire up the other core and see how hot it gets (I suspect 10-15C cooler than it was).
 
In utterly selfish news, I'm currently #1 in the team by RAC. Go me! \o/

I rather like this graph too...

cpartrank1.php


Now that's a good ol' stompin' :D
 
Yeah, the workunit crashed with a segmentation fault...

http://boinc.bakerlab.org/rosetta/result.php?resultid=234622721 (short-lived link - will disappear in 24 hours).

Code: 
<core_client_version>5.10.45</core_client_version>
<![CDATA[
<stderr_txt>
BOINC:: Initializing ... ok.
[2009- 3-11  0:49:28:] :: BOINC :: boinc_init()
BOINC:: Setting up shared resources ... ok.
BOINC:: Setting up semaphores ... ok.
BOINC:: Updating status ... ok.
BOINC:: Registering timer callback... ok.
BOINC:: Worker initialized successfully. 
Registering options.. 
Registered extra options.
Initializing core...
Initializing options.... ok 
Loaded options.... ok 
Processed options.... ok 
Initializing random generators... ok 
Initialization complete. 
Setting WU description ...
Unpacking zip data: ../../projects/boinc.bakerlab.org_rosetta/minirosetta_database_rev26003.zip
<unzip> <-oq> <../../projects/boinc.bakerlab.org_rosetta/minirosetta_database_rev26003.zip> <-d./> 
Firstarg=true; pp=-d./ 
firstarg: <-d./> 
End of unzipping.
Unpacking WU data ...
Unpacking data: ../../projects/boinc.bakerlab.org_rosetta/lrfrag_t328_.zip
<unzip> <-oq> <../../projects/boinc.bakerlab.org_rosetta/lrfrag_t328_.zip> <-d./> 
Firstarg=true; pp=-d./ 
firstarg: <-d./> 
End of unzipping.
Setting database description ...
Setting up checkpointing ...
Setting up folding (abrelax) ...
Beginning folding (abrelax) ... 
BOINC:: Worker startup. 
Starting watchdog...
Watchdog active.
Starting work on structure: _00001
# cpu_run_time_pref: 14400
====>
called boinc_finish
[color=yellow][b]SIGSEGV: segmentation violation[/b][/color]
Stack trace (17 frames):
[0x8f88f07]
[0x8fb3778]
[0xb7ffe420]
[0x8fe74d4]
[0x804adf8]
[0x865972e]
[0x8663a21]
[0x8628e86]
[0x813a5a8]
[0x8137206]
[0x844b39b]
[0x843ef88]
[0x8192ffc]
[0x81a2a17]
[0x80498ea]
[0x900f84c]
[0x8048111]

Exiting...

</stderr_txt>
]]>

Could be random bad luck, bad hardware design, or just old age. It'd be worth checking ventilation and trying to blow away any dust build-up.

Other things I spotted:

It didn't finish even the first simulation, so it's doubtful you'd have got much credit anyway. It may just be too slow. You could try setting it to run for 24 hours, but it's doubtful you'll see more than 50 PPD from it, so you might just want to turn it off.

If you do decide to continue, 5.10.45 is a somewhat old BOINC build. It could be that. I'd recommend using 6.2.15 (6.4.5 works but I've noticed a few scheduler bugs with that one).
 
Last edited:
I'll give him his due - he's doing that with less systems than me - but I did pass him to get where I am now.

It's that Biffa chap that might give me a Nightmare. :)
 
http://boinc.bakerlab.org/rosetta/show_host_detail.php?hostid=864190

That's equivalent to a 1.92GHz C2D laptop switched on 12 hours/day.

In fact, the laptop contains a 2.4GHz C2D, but I wasn't comfortable with the heat it was putting out, so I've capped it at 1.92GHz using SpeedStep and limited BOINC to one core only, which yields a 30C temperature drop - rather more comfortable if somewhat less efficient.
 
I was doing maintenance on Team OcUK early yesterday and didn't get to switch the updates back on until early afternoon. For some projects that makes no difference, and for others it does. Rosetta appears to be in the latter category. :eek:

Nothing to worry about. Move along. Nothing to see here. Programmer blaming his tools. Etc. :D

PS - I do this once a month now. After the catastrophe that happened last year when it ran out of disk space and promptly crashed and burned with no recent backup in sight, I'm a little more careful. :)
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom