| Info | Message |
|---|---|
| 1) Message boards : Server programs : Server reissues task more then "total" specified
Message 10921 Posted 15 Jun 2007 by feet1st |
Make sure you are logged in. I know I was logged in. I'm on another PC today and now clearly see a "new ticket" button up top. I'm guessing I just needed to refresh the display. Browser must have cached the menu bar without the "new" option on it. edit: I had found last night in the wiki that you go to /newticket to add a ticket, so I was signed on and attempted the correct URL with /trac/newticket and received a specific message that my profile did not have sufficient authority. Perhaps it just takes a while for the profile to be fully functional. I had just created it. Thanks for opening the ticket. |
| 2) Message boards : Server programs : Server reissues task more then "total" specified
Message 10903 Posted 15 Jun 2007 by feet1st |
I wasn't aware of the system. Thanks for the link. I've registered there, but still do not appear to have authority to create a new ticket. I found in the Wiki there that the field name in the configuration files that is related to this issue is called: max_total_results Can someone open a ticket for this issue please? |
| 3) Message boards : Server programs : Server reissues task more then "total" specified
Message 10894 Posted 14 Jun 2007 by feet1st |
I've seen this on a number of different projects. So it's not unique to Rosie. ...which is why I posted the issue here. Yes. There's apparently a scenario in the server code where issued tasks are miscounted. |
| 4) Message boards : Server programs : Server reissues task more then "total" specified
Message 10890 Posted 13 Jun 2007 by feet1st |
Rosetta is seeing cases where 3 results get sent out for the same WUID, even though the project is set up to have a total of at most 2. This seems to be possible when the first result is returned with errors and the second misses the deadline. In that case, a third result is issued for the WUID and this exceeds the total allowed. Is there something about how the missed deadline is being counted? Reference this Rosetta thread. |
| 5) Message boards : Web interfaces : View user's General Preferences
Message 9196 Posted 29 Mar 2007 by feet1st |
My point is simply that no series of steps, regardless of how easy or how documented and how widely known... is going to make BOINC seem easier to use. If we make it easy to access such information, then responders (including the Skype team) can speak with confidence about how the user's settings are causing the behavior the user is observing. Or, if you prefer the wording, how the user might modify their settings to better suite BOINC to the environment of their machine. |
| 6) Message boards : Web interfaces : View user's General Preferences
Message 9194 Posted 29 Mar 2007 by feet1st |
Clearly I am being misunderstood here. None of the information I wish to see is personal, nor can it lead to any spamming. I just want to see the basic general preferences, % of memory to use while idle, % of CPU to use, % of disk allowed, do work while computer in use, whether the person has multiple locations configured... THAT stuff. ...and if you didn't wish others to be able to see that, well the ability to view the preferences could be controlled by whether your computers are hidden or not. And only show the preference info. if they are not hidden. The problems with asking users to copy the information are numerous.... The reason I want the info. is to assist new users in setting things up and getting BOINC running. As new users, they don't necessarily KNOW how to view it, and aren't necessarily adept at using the msg boards to post the response. And to document all of that detail on every reply requesting the information would really be a pain. Asking users to go read about how to get the info. does not work well either. I think by now we realize that people don't read. With that being the ONLY means of obtaining such basic information, you are adding about a day longer to that user getting their question answered. And that's assuming they are able to figure out how to post the info. |
| 7) Message boards : Web interfaces : View user's General Preferences
Message 9191 Posted 29 Mar 2007 by feet1st |
I guess my point is that I'd like to request that such a function be available via the web interfaces. Presumably each specific project would have to enable this for it to work. Perhaps a link from the user profile page of the project that would go retrieve the user's General Preferences... project preferences would be nice to see too. To alter them I'd need the user's password or account key. I just want to view them. |
| 8) Message boards : Web interfaces : View user's General Preferences
Message 9188 Posted 29 Mar 2007 by feet1st |
Is there any means of viewing the General Preferences of another BOINC user?? It would be extremely helpful in trying to assist other users if we could view their preference settings. |
| 9) Message boards : BOINC Manager : BOINC has CLIMATE PREDICTION
Message 3770 Posted 4 Apr 2006 by feet1st |
I think thr max WU-runtime nowadays is 12hr, The max is presently 24hrs until that bug is nailed. Each WU is 2-3MB, regardless of how long you chose to crunch it. The results are generally 300K or less. So, you probably had seen the R@H default of 2hr WUs. But you can set the preference in the Rosetta preferences to up to 24hrs, thus using about a tenth of the bandwidth you may have observed. Once the rare 1% bug is nailed, they'll up the limit to 4 days again. They're just trying to avoid anyone spinning for 4 days on a WU that's not getting anywhere. So, crunching for 4days per WU would cut network bandwidth by 75% AGAIN! But you'd still have the 3MB download when you do get a new WU. But if you're modem runs 56kb that should still be <15 min. There's a nice chart in the wiki that might help you chose the projects that will work for you. |
| 10) Message boards : BOINC client : Finish WU before reschedule
Message 3769 Posted 4 Apr 2006 by feet1st |
How (if possible) can I fine tune the client I believe you are saying your objective is simply to complete the work unit that's already in progress before switching. Bottom line, BOINC doesn't work that way... unless you only have one project installed with work. But that's ok, the WUs are made to be interrupted and take the CPU time they get from the scheduling algorythms. You can change your general preferences to "Leave applications in memory while preempted? YES". This will minimize any efficiency that may be lost when a WU is suspended without a recent checkpoint. You can also minimize the effect a bit if you increase your "Switch between applications every... minutes" setting to exceed your WU size preference. Rosetta allows a project specific preference for your preferred WU size. If you've got high bandwidth available, and want to keep a tidy WU list, perhaps a short 2hr WU is for you. Switch between apps every 120 minutes or a bit more, and you'll be closer or totally completed more frequently. I have mine set to switch every 360min. Also, every time you "connect to network" (in your case every 2.4hrs, .1 days) the scheduler comes in and makes a new assessment as to who should be running next. So, if you'd increase that value to connect say every .25 or .5 days, then you'd be reducing the number of interruptions from connections, and therefore reducing the number of times the scheduler makes a new assessment. |
| 11) Message boards : BOINC client : Does BOINC affect a network?
Message 3768 Posted 4 Apr 2006 by feet1st |
Actually, Rosetta now has a preference where you designate how long you'd like to crunch on a given WU. So you can download one WU, and crunch for 24hrs if you like to minimize bandwidth. The prior post here (note the ~4 month time difference) relates to R@H prior to implementation of this setting, and they were all running 2hrs I believe. It appears part of the question was about network security "affect". I believe the initial posters were focused on network bandwidth. But I wanted to point out there is some BOINC security information here: Security issues in volunteer computing |
| 12) Message boards : BOINC client : req: new feat. to avoid BOINC clogging networks
Message 3767 Posted 4 Apr 2006 by feet1st |
Carlos, it sounds like you should increase your "connect to network ever xx days" in your general preferences. That way you can use the free access for ALL of your connections. And I believe it is correct to say BOINC will get only two connections, but, I concur with Carlos, it's two PER PROJECT. Also, I've tried throttling back my upload and download rates, all the way to 1KB/sec. It seems to have no effect. My transfers still seem to take all the bandwidth they can get. My messages consistently show large transfers occuring at 50 or even 100,000 bytes (i.e. 100KB) per second. And I've updated the preferences. Does anyone know if the upload and download rates work? I've seen multiple projects (all that I've observed so far) running dramatically faster than my preference. |
Copyright © 2025 University of California.
Permission is granted to copy, distribute and/or modify this document
under the terms of the GNU Free Documentation License,
Version 1.2 or any later version published by the Free Software Foundation.