Is anyone getting new WU's or am I having the problem?

Message boards : Number crunching : Is anyone getting new WU's or am I having the problem?

To post messages, you must log in.

1 · 2 · Next

AuthorMessage
gibsonrr

Send message
Joined: 10 Dec 05
Posts: 2
Credit: 11,514,122
RAC: 0
Message 46243 - Posted: 15 Sep 2007, 2:25:56 UTC

I seem to be having a problem getting new WU's. When I try to do an "update", it responds with "(not requesting new work or reporting results)". However, the queue is empty. I have upgraded to BOINC 5.10.20.

Looking for suggestions.

Thanks,
Rob Gibson

ID: 46243 · 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 46244 - Posted: 15 Sep 2007, 2:56:44 UTC

I take it that you also have other projects running? And that they do have work available to do?

This thread has some details and explanations for what you are observing.
Rosetta Moderator: Mod.Sense
ID: 46244 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
gibsonrr

Send message
Joined: 10 Dec 05
Posts: 2
Credit: 11,514,122
RAC: 0
Message 46247 - Posted: 15 Sep 2007, 3:46:06 UTC - in response to Message 46244.  

I take it that you also have other projects running? And that they do have work available to do?

This thread has some details and explanations for what you are observing.


I have stopped other projects from running. I am only trying to crunch R@H WU's.

Here is what I am seeing:

2007-09-14 23:27:45 [rosetta@home] Sending scheduler request: Requested by user
2007-09-14 23:27:45 [rosetta@home] (not requesting new work or reporting completed tasks)
2007-09-14 23:27:50 [rosetta@home] Scheduler RPC succeeded [server version 509]
2007-09-14 23:27:50 [rosetta@home] Deferring communication for 4 min 2 sec
2007-09-14 23:27:50 [rosetta@home] Reason: requested by project
2007-09-14 23:37:06 [rosetta@home] Sending scheduler request: Requested by user
2007-09-14 23:37:06 [rosetta@home] (not requesting new work or reporting completed tasks)
2007-09-14 23:37:11 [rosetta@home] Scheduler RPC succeeded [server version 509]
2007-09-14 23:37:11 [rosetta@home] Deferring communication for 4 min 2 sec
2007-09-14 23:37:11 [rosetta@home] Reason: requested by project

Thanks,
Rob Gibson

ID: 46247 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Greg_BE
Avatar

Send message
Joined: 30 May 06
Posts: 5658
Credit: 5,670,291
RAC: 2,328
Message 46259 - Posted: 15 Sep 2007, 8:55:03 UTC - in response to Message 46247.  

I dont remember where it is in the boards, but SETI may be owed some processing time so BOINC feels that RAH will have to wait until SETI's debt is settled.
Also you may have enough work stacked up from RAH that you don't need any more work for the time being. Check the total number of work units you have and devide that by the processing time you set for RAH and see if that matches
you settings for the amount of work your settings allow for.

[
I take it that you also have other projects running? And that they do have work available to do?

This thread has some details and explanations for what you are observing.


I have stopped other projects from running. I am only trying to crunch R@H WU's.

Here is what I am seeing:

2007-09-14 23:27:45 [rosetta@home] Sending scheduler request: Requested by user
2007-09-14 23:27:45 [rosetta@home] (not requesting new work or reporting completed tasks)
2007-09-14 23:27:50 [rosetta@home] Scheduler RPC succeeded [server version 509]
2007-09-14 23:27:50 [rosetta@home] Deferring communication for 4 min 2 sec
2007-09-14 23:27:50 [rosetta@home] Reason: requested by project
2007-09-14 23:37:06 [rosetta@home] Sending scheduler request: Requested by user
2007-09-14 23:37:06 [rosetta@home] (not requesting new work or reporting completed tasks)
2007-09-14 23:37:11 [rosetta@home] Scheduler RPC succeeded [server version 509]
2007-09-14 23:37:11 [rosetta@home] Deferring communication for 4 min 2 sec
2007-09-14 23:37:11 [rosetta@home] Reason: requested by project

Thanks,
Rob Gibson


ID: 46259 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
FoldingSolutions
Avatar

Send message
Joined: 2 Apr 06
Posts: 129
Credit: 3,506,690
RAC: 0
Message 46288 - Posted: 15 Sep 2007, 17:38:53 UTC - in response to Message 46259.  

Have you tried checking if the "allow new work" button has been pressed?
ID: 46288 · 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 46294 - Posted: 15 Sep 2007, 18:09:37 UTC

gibsonrr please verify the following:

Rosetta project allows new work (see projects tab in advanced view)
Rosetta project is not suspended (see projects tab in advanced view)
Other projects are "suspended" (see projects tab in advanced view)
BOINC is allowed network access (see advanced view "Activity" pulldown menu, "Network activity always available")

If you go to the projects tab, is there any status shown out to the right on the Rosetta line?
Rosetta Moderator: Mod.Sense
ID: 46294 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
TA_JC

Send message
Joined: 7 Nov 05
Posts: 13
Credit: 6,017,070
RAC: 10,870
Message 46402 - Posted: 17 Sep 2007, 2:35:36 UTC

I'm getting the same messages as the others. My other project doesn't have any pending deadlines, Rosetta is set to 71%, but it refuses to queue me any WUs.

You know, I just noticed...I changed my queue to 6 days today, my single-core box has 2 days or so of work, but both my other dual-cores have 11+ days work! These 3 machines are only crunching Rosetta. What gives???
ID: 46402 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
DJStarfox

Send message
Joined: 19 Jul 07
Posts: 145
Credit: 1,235,310
RAC: 70
Message 46423 - Posted: 17 Sep 2007, 12:37:04 UTC - in response to Message 46402.  

I'm getting the same messages as the others. My other project doesn't have any pending deadlines, Rosetta is set to 71%, but it refuses to queue me any WUs.

You know, I just noticed...I changed my queue to 6 days today, my single-core box has 2 days or so of work, but both my other dual-cores have 11+ days work! These 3 machines are only crunching Rosetta. What gives???


Well, we are not there at your machine to see exactly what you do. But I have an idea that I know works for me. The trick is to create a situation where BOINC has a spare CPU but no WU is being crunched. Internally, BOINC will sense the idle CPU and demand more work from the next project it can. Make sure the Activity menu has "Run Always" and "Network Activity Always available" while you're doing this.

Also, generally I don't suspend a project, I just set "No new work" on the project(s). Then, I will suspend individual tasks if necessary, or I let the queue empty for that project. This gracefully allows BOINC to empty a queue for one project and enqueue more WU from another.

Hope that helps you.
ID: 46423 · 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 46432 - Posted: 17 Sep 2007, 13:37:15 UTC - in response to Message 46402.  

I'm getting the same messages as the others. My other project doesn't have any pending deadlines, Rosetta is set to 71%, but it refuses to queue me any WUs.

You know, I just noticed...I changed my queue to 6 days today, my single-core box has 2 days or so of work, but both my other dual-cores have 11+ days work! These 3 machines are only crunching Rosetta. What gives???


Is the estimated runtime shown in line with your runtime preference? If your last task took longer then the preference, perhaps the estimates are all reflecting that. Let it crunch through. You never know, sometimes tasks finish in advance of the preference too, so most likely you will have it all done before the 10 day deadline.

Oh, and don't forget, a dual core will be crunching two tasks at a time. So, 48hrs is one full day of work, for two cores.
Rosetta Moderator: Mod.Sense
ID: 46432 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Greg_BE
Avatar

Send message
Joined: 30 May 06
Posts: 5658
Credit: 5,670,291
RAC: 2,328
Message 46464 - Posted: 17 Sep 2007, 18:36:02 UTC

Something is weird, I am set for communicate every 1 day and load additional 7 days work. Run time is set for 6 hrs and I am now down to 2.83 days of work with nothing new loading in. Whats going on?
ID: 46464 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
TA_JC

Send message
Joined: 7 Nov 05
Posts: 13
Credit: 6,017,070
RAC: 10,870
Message 46496 - Posted: 18 Sep 2007, 2:10:14 UTC

Well, this morning I suspended my other projects, all the settings were right, and Rosetta said there was no work :( Then about 9AM the floodgates opened, and now I'm full up on WUs :)
ID: 46496 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Greg_BE
Avatar

Send message
Joined: 30 May 06
Posts: 5658
Credit: 5,670,291
RAC: 2,328
Message 46502 - Posted: 18 Sep 2007, 5:43:30 UTC - in response to Message 46496.  

i'm soley rosie and still no new work, there is nothing in messages saying no new work, just doesnt download. only upload.
Well, this morning I suspended my other projects, all the settings were right, and Rosetta said there was no work :( Then about 9AM the floodgates opened, and now I'm full up on WUs :)


ID: 46502 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
tomba

Send message
Joined: 29 May 06
Posts: 43
Credit: 1,558,972
RAC: 0
Message 46623 - Posted: 19 Sep 2007, 16:48:15 UTC

Another "me too".

I've changed nothing in preferences but I'll be out of work some time tonight.
ID: 46623 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Greg_BE
Avatar

Send message
Joined: 30 May 06
Posts: 5658
Credit: 5,670,291
RAC: 2,328
Message 46628 - Posted: 19 Sep 2007, 17:45:37 UTC - in response to Message 46623.  

odd...I see you have 5.10.20 and 5.80 on your system.
i would think a reset might do the trick, wait till you run out of work and try that. if that doesn't cure your problem, download a fresh copy of BOINC and do a repair of the program. I ran some disk diagnostics and cleaned up the old files on my drive as well. Were you running CAPRI at all?

I think that is what is messing up some of us.


Another "me too".

I've changed nothing in preferences but I'll be out of work some time tonight.


ID: 46628 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
tomba

Send message
Joined: 29 May 06
Posts: 43
Credit: 1,558,972
RAC: 0
Message 46629 - Posted: 19 Sep 2007, 17:57:53 UTC - in response to Message 46628.  

odd...I see you have 5.10.20 and 5.80 on your system.

I've no idea what that means...
i would think a reset might do the trick, wait till you run out of work and try that. if that doesn't cure your problem, download a fresh copy of BOINC and do a repair of the program. I ran some disk diagnostics and cleaned up the old files on my drive as well.

How does one do a repair??
Were you running CAPRI at all?

No idea what that is so I quess not!!

Thanks for the response. Tom
ID: 46629 · 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 46639 - Posted: 19 Sep 2007, 18:50:05 UTC - in response to Message 46244.  

Let's not go off resetting projects and relating how the BOINC scheduler works to specific Rosetta work units.

Review this info (link that follows). If your client does not request xxx seconds of work, it will not receive any. Simple as that.
This thread has some details and explanations for what you are observing.

Rosetta Moderator: Mod.Sense
ID: 46639 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Greg_BE
Avatar

Send message
Joined: 30 May 06
Posts: 5658
Credit: 5,670,291
RAC: 2,328
Message 46641 - Posted: 19 Sep 2007, 20:17:35 UTC
Last modified: 19 Sep 2007, 20:19:59 UTC

interesting, but....he is running out of work on rosie
i was in the same place...all CAPRI and no new work and rosie is my sole project.
changed settings localy and online. that didn't work.
so what was i supposed to do????
I had to go the hard way since nothing else worked.
I was set for report everyday and keep 7 days in extra work and nothing happened.

perhaps you can add the way to get your project to try and respond when it won't download any new work to your FAQ list or point me to it.
ID: 46641 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
The_Bad_Penguin
Avatar

Send message
Joined: 5 Jun 06
Posts: 2751
Credit: 4,271,025
RAC: 0
Message 46642 - Posted: 19 Sep 2007, 21:26:19 UTC - in response to Message 46639.  
Last modified: 19 Sep 2007, 21:29:09 UTC

What did I miss from that post?

I had Rosie 100% on my quadcore, until "the problems". Since then, Rosie at 90%, and about five other back-up projects at 10%.

From this post, I would expect the result from "the problems" to mean that Rosie would be taking over nearly 100%.

Right now, with the quadcore at 90% Rosie, I have no Rosie wu's, and when I update, it does NOT request new work.

Currently have 30 hours (one 13 hr wu / 14 one hour wu's) of non-Rosie wu's running / waiting. With four cores, that would be about 7 hours worth of crunchin' time.

What am I not understanding?

Review this info (link that follows). If your client does not request xxx seconds of work, it will not receive any. Simple as that.
This thread has some details and explanations for what you are observing.

ID: 46642 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Greg_BE
Avatar

Send message
Joined: 30 May 06
Posts: 5658
Credit: 5,670,291
RAC: 2,328
Message 46644 - Posted: 19 Sep 2007, 21:42:36 UTC - in response to Message 46642.  
Last modified: 19 Sep 2007, 21:48:49 UTC

now you see what I am getting at.
since i did the reset and got rid of all CAPRI and 5.78 stuff i was able to download 5.80 and got my 7 days of work and its chugging along just fine now.
It reports each work unit and downloads one in its place. so i feel the threads pointed out so far just don't match the actual problem being presented here.
there's work to be had by our machines, but BOINC isn't letting it in. how do you explain Tomba's problem of being out of work tonight when he has 5.80 and BOINC 5.10.20?

The thread quoted here is useless to this discussion, nothing is being taken over, its the lack of new work coming in. By theory, as the CAPRI work units were being finished they should have been replaced with the stuff that 5.80 is running now or at the time it should have downloaded 5.80 CAPRI until they were complete or aborted. This did not happen. Also if there was debt to be paid back by Tomba's machine to Rosie then he should be getting loads of work and his other project should be scaled back until there is no debt owed to rosie. That is not the case. He is getting no work from rosie. For me,there is/was no debt to be paid by my machine as i am dedicated rosie cruncher. So the thread does not speak to my situation.

What did I miss from that post?

I had Rosie 100% on my quadcore, until "the problems". Since then, Rosie at 90%, and about five other back-up projects at 10%.

From this post, I would expect the result from "the problems" to mean that Rosie would be taking over nearly 100%.

Right now, with the quadcore at 90% Rosie, I have no Rosie wu's, and when I update, it does NOT request new work.

Currently have 30 hours (one 13 hr wu / 14 one hour wu's) of non-Rosie wu's running / waiting. With four cores, that would be about 7 hours worth of crunchin' time.

What am I not understanding?

Review this info (link that follows). If your client does not request xxx seconds of work, it will not receive any. Simple as that.
This thread has some details and explanations for what you are observing.



ID: 46644 · 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 46647 - Posted: 19 Sep 2007, 23:08:46 UTC

My point is simply that you are taking actions on Rosetta, like resetting the project, aborting tasks etc. when you are concerned about actions of the BOINC scheduler. So, when your son is misbehaving, you are scolding your daughter. It may have some side effects, but it's not directly going to improve your son's behavior.

The key phrase in the description of the problem is that the machine is "ALMOST" out of work, and that it is NOT requesting new work. There's nothing Rosetta's programs can do about such a situation. These decisions are handled by BOINC. Looking back at it now, I guess the original post did say "OUT OF work". But that was before we knew anything else about the machine in question. And since that time others have presented other situations, so it has created somewhat a moving target here.

I pointed back to the other thread, because noone was describing the message about the work being requested, and whether xxx seconds were requested or not. And until we know that, we don't know which child is acting up.

Rosetta will not send you work until your BOINC client requests it. It doesn't mean Rosetta is out of work or malfunctioning. If your client does not request xxx seconds of work from Rosetta, then we need to dig in to BOINC, and which version, and what settings and debts it is using in it's scheduler's decision making.

Tomba said "...I'll be out of work some time tonight", and said nothing about how many days of work BOINC is configured to request, other projects, resource shares, and nothing about the message requesting work from the project.

I'm not trying to contradict what others are saying about what worked for them. We each have our own experience with what passes for normal. And sometimes scolding the daughter DOES result in the son cleaning up his act. I'm just trying to avoid unnecessary downloads, manual efforts, and the potential for losing useful work (by doing project reset), when it is the "son" that we need to be observing and questioning here to get to the bottom of things.
Rosetta Moderator: Mod.Sense
ID: 46647 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
1 · 2 · Next

Message boards : Number crunching : Is anyone getting new WU's or am I having the problem?



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