No new work

Message boards : Number crunching : No new work

To post messages, you must log in.

Previous · 1 · 2

AuthorMessage
Profile Hammer

Send message
Joined: 7 Mar 07
Posts: 4
Credit: 13,686,257
RAC: 14,206
Message 64668 - Posted: 31 Dec 2009, 3:20:53 UTC - in response to Message 64667.  
Last modified: 31 Dec 2009, 3:36:56 UTC

All the computing power that has been donated freely, going unused. I guess I'm missing the reason behind the apparent lack of urgency to fix whatever is broken.


It's not going to "waste" if you run multiple projects.

If you're not on multiple projects you're just a credit-obsessive, fanboy that cares only for points and not the science.

Just my opinion.


That was uncalled for.
ID: 64668 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile tiger

Send message
Joined: 16 Jul 06
Posts: 17
Credit: 1,083,385
RAC: 0
Message 64669 - Posted: 31 Dec 2009, 4:31:39 UTC - in response to Message 64668.  

Heck, I'm not even sure what he said (re: "fanboy") other than it sounded immature.

[/quote]

That was uncalled for.[/quote]

ID: 64669 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile banditwolf

Send message
Joined: 10 Jan 06
Posts: 28
Credit: 139,737
RAC: 0
Message 64673 - Posted: 31 Dec 2009, 15:07:29 UTC - in response to Message 64670.  

All the computing power that has been donated freely, going unused. I guess I'm missing the reason behind the apparent lack of urgency to fix whatever is broken.


It's not going to "waste" if you run multiple projects.

If you're not on multiple projects you're just a credit-obsessive, fanboy that cares only for points and not the science.

Just my opinion.


I don't think that is the problem. Rosetta is not a project where you can earn 'the big bucks'.


Nope, sure isn't.
ID: 64673 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Mod.Sense
Volunteer moderator

Send message
Joined: 22 Aug 06
Posts: 4018
Credit: 0
RAC: 0
Message 64676 - Posted: 31 Dec 2009, 16:53:43 UTC - in response to Message 64663.  

Mod.Sense: Are you in a position to confirm that no-one will be able to address this issue before Monday after the New Year? ie Jan 4th 2010? Given that nothing's changed since Xmas that's the assumption I'm now making, so I'm going to ramp up the runtime on my remaining WUs (desktop quad) to the maximum 24 hours.


No, I'm not in a position to confirm such an assumption. So, I'll just be presuming with you. Meanwhile, I allowed new work from another project. This will help me leave R@h servers alone when work becomes available as well, thus leaving work for others to get and help the system reattain a normal state sooner.
Rosetta Moderator: Mod.Sense
ID: 64676 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile jaxom1
Avatar

Send message
Joined: 5 Jun 06
Posts: 180
Credit: 1,586,889
RAC: 0
Message 64678 - Posted: 31 Dec 2009, 17:45:43 UTC
Last modified: 31 Dec 2009, 17:50:00 UTC

I find it funny that this is the week I decided to start running Rosseta again. I haven't been doing this for a while, but spun a server up only to find out it cannot get work! lol. I just added another project, but still find it funny...

ID: 64678 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Chilean
Avatar

Send message
Joined: 16 Oct 05
Posts: 711
Credit: 26,694,507
RAC: 0
Message 64679 - Posted: 31 Dec 2009, 18:10:40 UTC
Last modified: 31 Dec 2009, 18:43:44 UTC

Well, I joined POEM@Home in the meanwhile.
ID: 64679 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Dave Mickey

Send message
Joined: 29 Dec 07
Posts: 33
Credit: 4,136,957
RAC: 0
Message 64683 - Posted: 31 Dec 2009, 23:56:50 UTC

my take on this is that work is not shut off, but rather only dribbling
out (so far), since in the last 24H, each of my 2 active machines got a WU.
So I've cranked up the runtime to make the most of each one I can get.

Normally, I keep runtimes in the 8-12 hour range so that credit comes in
more smaller chunks, just to smooth it out. But if work is rare, then I'd
rather make the most of it to keep multi-project debt and what not in
smoothness (compared to r@h running dry, and then debt values flailing about).

If not, boinc will just crunch some other stuff if r@h does run dry.

Dave
ID: 64683 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile robertmiles

Send message
Joined: 16 Jun 08
Posts: 1223
Credit: 13,824,497
RAC: 2,340
Message 64687 - Posted: 1 Jan 2010, 4:23:16 UTC

If you look at the server status, you'll see that only one of the two workunit generator programs is still running.

I can't tell if one of the programs needed for downloading workunits is among the approximately half of the other server programs not running.

So I'd guess that the server just can't keep up with the demand for new workunits at present. Likely to improve soon after someone able to fix problems with server programs not running comes back from the holidays, though.
ID: 64687 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile robertmiles

Send message
Joined: 16 Jun 08
Posts: 1223
Credit: 13,824,497
RAC: 2,340
Message 64688 - Posted: 1 Jan 2010, 4:36:58 UTC - in response to Message 64683.  
Last modified: 1 Jan 2010, 5:10:09 UTC

If not, boinc will just crunch some other stuff if r@h does run dry.

Dave


I've seen one way to handle situations where your preferred BOINC project stops generating enough workunits for a while: Enable at least one more BOINC project, but with a share of the CPU low enough that it could normally barely finish workunits from that project by their deadlines. This soon allows your machine to execute workunits from the second project only when the first project has none to offer.

For those of you not ready to participate in beta test projects, I'd recommend either the protein folding subproject at World Computing Grid, or the separate POEM@home project, as the closest to what Rosetta@home is now doing. For those of you who are ready to participate in beta test projects, I expect you to already know a good selection of other BOINC projects you could choose instead.

A note to any of you with a GPU that you want to start participating in the PrimeGrid project - their estimates for how many workunits your GPU can handle in a certain length of time currently have enough problems that you'll need to keep the amount of time you're requesting workunits for very low until you've successfully returned at least one workunit with each of your machines that you allow to participate there.
ID: 64688 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
amgthis

Send message
Joined: 25 Mar 06
Posts: 81
Credit: 203,879,282
RAC: 0
Message 64689 - Posted: 1 Jan 2010, 5:15:03 UTC

My boxes are running out of work.
ID: 64689 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
mikey
Avatar

Send message
Joined: 5 Jan 06
Posts: 1894
Credit: 8,541,841
RAC: 10,179
Message 64690 - Posted: 1 Jan 2010, 10:55:50 UTC - in response to Message 64688.  

If not, boinc will just crunch some other stuff if r@h does run dry.

Dave


I've seen one way to handle situations where your preferred BOINC project stops generating enough workunits for a while: Enable at least one more BOINC project, but with a share of the CPU low enough that it could normally barely finish workunits from that project by their deadlines. This soon allows your machine to execute workunits from the second project only when the first project has none to offer.

For those of you not ready to participate in beta test projects, I'd recommend either the protein folding subproject at World Computing Grid, or the separate POEM@home project, as the closest to what Rosetta@home is now doing. For those of you who are ready to participate in beta test projects, I expect you to already know a good selection of other BOINC projects you could choose instead.

A note to any of you with a GPU that you want to start participating in the PrimeGrid project - their estimates for how many workunits your GPU can handle in a certain length of time currently have enough problems that you'll need to keep the amount of time you're requesting workunits for very low until you've successfully returned at least one workunit with each of your machines that you allow to participate there.


Just to expand on this a little bit, here is a link to all the current Distributed Computing Projects out there, the Boinc ones are noted:
http://www.distributedcomputing.info/projects.html

As for GPU projects Collatz http://boinc.thesonntags.com/collatz/ lets you use both your ATI and Nvidia cards and gives excellent credits.

But there are many other projects you can use to keep your pc's happy and contributing, as Robert said just assign them a low percentage, says 5% or less and they will basically only crunch so infrequently that you won't even notice them.
ID: 64690 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
AMD_is_logical

Send message
Joined: 20 Dec 05
Posts: 299
Credit: 31,460,681
RAC: 0
Message 64697 - Posted: 1 Jan 2010, 15:51:12 UTC - in response to Message 64687.  

So I'd guess that the server just can't keep up with the demand for new workunits at present.

There is no NEW work being generated at all right now. My machines sometimes get a WU, but it is always an old one that was resent after someone else didn't return it.

I notice on the Server Status page that one of the programs not working is "db_purge". I would guess that there is a safety that stops WUs from being generated if there are already too many in the database. Maybe that's the problem.

ID: 64697 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
JohnG

Send message
Joined: 22 Jul 07
Posts: 3
Credit: 1,076,801
RAC: 0
Message 64698 - Posted: 1 Jan 2010, 16:16:35 UTC

Ran out of WU's about an hour ago for my CPU.I guess I am on to Einstein for something to do.
ID: 64698 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile tiger

Send message
Joined: 16 Jul 06
Posts: 17
Credit: 1,083,385
RAC: 0
Message 64700 - Posted: 1 Jan 2010, 16:38:47 UTC - in response to Message 64679.  

Me too because I DO care about the science. Such an incredible waste.

Well, I joined POEM@Home in the meanwhile.


ID: 64700 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Sid Celery

Send message
Joined: 11 Feb 08
Posts: 1965
Credit: 38,174,417
RAC: 10,123
Message 64715 - Posted: 2 Jan 2010, 2:45:10 UTC - in response to Message 64676.  

Mod.Sense: Are you in a position to confirm that no-one will be able to address this issue before Monday after the New Year? ie Jan 4th 2010? Given that nothing's changed since Xmas that's the assumption I'm now making, so I'm going to ramp up the runtime on my remaining WUs (desktop quad) to the maximum 24 hours.

No, I'm not in a position to confirm such an assumption. So, I'll just be presuming with you. Meanwhile, I allowed new work from another project. This will help me leave R@h servers alone when work becomes available as well, thus leaving work for others to get and help the system reattain a normal state sooner.

Ok, thanks anyway. It looks like my guess won't be far off. I've grabbed the odd single WU over the last few days but otherwise I'm out like everyone else.

The plus side is that running 24 hour WUs has been reasonably successful (bar hitting the 100 decoy limit I think), so I'll probably drop down to 8 or 10 hours when things come back on stream rather than my previous 4, which'll help the servers in my own small way. Some small dividend and more individual WUs for others.

Thanks to Mikey for the link to other projects. POEM looks a little too similar to here, so I'm going to try and give WCG a go - I've been interested in their work for a while now.
ID: 64715 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Previous · 1 · 2

Message boards : Number crunching : No new work



©2024 University of Washington
https://www.bakerlab.org