| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Retired Forums
Forum: Member-to-Member Support [Read Only]
Thread: Software blocked |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 10
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Here it is my problem
I run the WCG Agent for months. One or two weeks ago, I had a problem. For example, i'll say that the Task run time was 29h 54 min and 25 sec when i started my computer (you'll see it's important) and the Task execution was 93.3% Ok so my computer is on and the Task Run time was running, normally, when i clicked to the application button, the graphics were evolving normally too, BUT, the Task execution progress was like blocked. The percentage remained the same. So when i realized that problem after an hour (so 30h 54min of running time) I restarted my computer. The task run time was back to 29h54 min 25 sec. And for the rest, it was the same thing, the program was running, the time was running, the graphics too, but the percentage stayed the same. I uninstalled the software, and reinstalled it. It worked well again. Today, it does the same thing. The percentage stay blocked. What can be the problem ?  Thanks in advance for your answers Cédric. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Cédric --
I have not seen anything like this reported before. I will say that it does not surprise me that your task run time goes back to what it had been when you restart your computer. If the percentage completion does not change, no real processing is happening and no checkpoints will be taken, so you are loading from the same checkpoint as before when you restart. Choosing to reinstall as you did earlier was the right choice to make then. Since it has happened again, we need to find out the cause so that it will not happen again for you. Could you load the Task Manager (right click on the system tray bar at the bottom of the screen) and go to Processes and see if WCGrid_Rosetta.exe appears in the list and if so, what percentage utilization does it show? Have you noticed anything else strange about your system when this is happening? I think this will not be an easy problem to resolve. We Community Admins do not have access to the program code, so this will probably require the assistance of the WCG Technical Staff. I will notify them of your problem, but unfortunately, this is a holiday weekend here in the US, so it may be Tuesday before any of them are available. Perhaps another of the Admins might have some thoughts on this as well. Best Regards, |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Me neither I've never heard something like that before, even if my team (GrenouilleTeamFrance) counts more than 100 members and that we're in touch to resolve our problems and answer our questions by ourselves.
If the task manager is what I think (yep sorry, but my windows is in french, so ...), I can obtain it also by ctrl+alt+suppr (or del maybe for english keyboards). Well, anyway, among the processes running WCGrid_Rosetta.exe is there, in the list, and his percentage utilisation is normal: 40% (One of the Admins told our team about Thread Master, used to reduce the utilization of the CPU, and I use it, so that's why my CPU is used by WCG at 40% and not 99%). So everything is in order. Nothing strange happened. The only common point that might exist between today and the last time (but I'm not absolutely sure that it was the case the last time), is that I runned my anti-spyware softwares (Spybot and Ad-Aware). But that's not new, I mean I used them before and there wasn't any problem. Another common point, is that the task executing was a "long" one in both cases and that I was almost at the end of them (today: 93.3% in 29h54, so it would be more than 30hrs to be at 100%, and the last time, I remember it was a same sort of task a more than 30hrs one, and that it was over 90%). I don't see anything else in common. Should I reinstall the software, or wait for the Technical staff ? I mean, would/could they try to "unblock" it, would they try to make me do some operations on my computer, or it doesn't matter if i reinstall WCG ? Well, happy 4th of July for the Independence Day for you and all the Staff, and every American  |
||
|
|
RT
Master Cruncher USA - Texas - DFW Joined: Dec 22, 2004 Post Count: 2636 Status: Offline Project Badges: 
|
I would almost bet this is a computer problem as opposed to a software problem.
----------------------------------------I would check my disk drive for an error. I would run disk clean up and then defrag it. then run this command: cipher /w:c:\ (which writes stuff in all the free areas). This might not help at all but at least you will have a clean disk. We need to get the specifics on that work unit so some of the rest of us can try running it and see what happens. I am sure the WCG Tech guys will respond, They were notified by Lewis Carroll. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hello Cédric,
What you are reporting sounds like a non-convergent work unit. If so, after some time (perhaps several hours), the program should stop trying to fold it and return the partial results to the server and get another work unit. If it does not, then there is a possible program error or hardware error. In that case, use 'Contact Us' to report the problem (including your machine ID and time) and then use Task Manager to kill WCGrid_Rosetta.exe. This will download a new Work Unit. Since several people are working on your work unit, if all have the same problem, then the WCG Staff will know that the problem is in either the work unit or the program. If only you have the problem, then . . . . we'll discuss it. mycrofth |
||
|
|
Alther
Former World Community Grid Tech United States of America Joined: Sep 30, 2004 Post Count: 414 Status: Offline Project Badges: 
|
This doesn't sound like a problem at all. It just sounds like you've run across a very tough portion of the protein.
----------------------------------------When you restart your computer or exit and restart the agent, it restarts at the last checkpoint. This is why you restarted at 29h and the same percentage when you restarted your computer. The software checkpoints after a successful fold. Sometimes this is takes just a few minutes, but sometimes it can take a few hours. If it hasn't been able to successfully fold after a few hours, we call this a "non-convergant" workunit and "abort" it. By aborting I mean we just finish it early and it submits a result containing what you've completed so far. You still get full credit and non-convergant workunits are just as important as converging ones. The simple way to see if Rosetta is running correctly is to check the graphics panel. If the graphs are all drawing then Rosetta is running properly. If it's been a while since the percentage increased (check the graphics screen as it's in tenths of a percent), but the graphs are drawing, just give it time. It's working properly. Again, sounds like everything is running along fine, you just need to let it get over this rather large hump. Happy crunching!
Rick Alther
Former World Community Grid Developer |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
@RT: I didn't mean that I doubted that the Tech Staff would respond (I'm sure they will, even if it's only on Tuesday), I just asked if it would matter if I reinstalled the software.
Well anyway, I'm gonna do what you say, it can only be good to my computer to be cleaned. We need to get the specifics on that work unit so some of the rest of us can try running it and see what happens. How can I know and give you what you're asking for ? (above all now that the partial results are sent)@mycrofth then use Task Manager to kill WCGrid_Rosetta.exe. This will download a new Work Unit I didn't even know that "killing" it like that made it download a new unit (well it did work though  The new task's running normally.) |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
The simple way to see if Rosetta is running correctly is to check the graphics panel. If the graphs are all drawing then Rosetta is running properly. If it's been a while since the percentage increased (check the graphics screen as it's in tenths of a percent), but the graphs are drawing, just give it time. It's working properly. Again, sounds like everything is running along fine, you just need to let it get over this rather large hump. Happy crunching! Well, I didn't give it a lot of time today, but I know that the last time I did let it time (if i remember well, about 7 hrs or so, and it didn't evolved of 0.1%). I'll see if it happens again, and if it does happen again, I'll let it run to see what's going on. And yes the graphics were drawing normally, that was just the rest that didn't evolve. Thank you everyone. |
||
|
|
Viktors
Former World Community Grid Tech Joined: Sep 20, 2004 Post Count: 653 Status: Offline Project Badges: 
|
When this happens, it is often useful to use the task manager to see if there is another process using the majority of the cpu time, which would cause Rosetta to slow down its progress.
Also, we have been getting a few longer running work units lately due to exceptionally difficult to fold proteins. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Yeah I know a little bit how WCG works and stuffs about it, my team's working on a document about that to help our future (and present as well) members. We help eachother since december '04, so some of us know well about it, either because they're good in computer's stuff or because they read your site and follow the experiences of members of our team (we don't know everything, that's sure, but we know a little bit about it). So I know that when another process uses a lot of the CPU power, WCG won't get "its" 40%, and that some units are longer than other (I already had some units needing less than 10 hrs, and others needing more than 40 hrs, and i know one of my team who had some 100hrs units). There's no problem about that. If the percentage runs slowly, but almost always at the same speed, that's ok, but when it turns normally during more than 90% of the unit, and suddenly it stays hours with the same percentage, I begin to wonder what happens.
I killed WCGrid_Rosetta.exe as mycrofth said, and I'll see if this happen another time. And if it does happen a third time, I'll let it run to see what finally happen after hours. I'll let you know if it happens again. Thanks a lot again for your help  |
||
|
|
|