Message boards :
Number crunching :
Is there a way ro remove hands from BOINC?
Message board moderation
Author | Message |
---|---|
Marek Majewski Send message Joined: 26 Nov 00 Posts: 31 Credit: 18,604,752 RAC: 147 |
Management tasks multiply when you are running multiple computers. I like to simplify these tasks. Therefore I have two questions: 1. Is there a way to make the "update" automatic so "ready to report" WUs go by themselves? 2. Is there a way (a third-party-ad-on, perhaps) to see from a central location the status and progress on each BOINC computer in the farm? With SETI classic I use MSETIMON and SETIQUEUE - they give me a nice view into the current state of each computer/WU and a nice way to look at historical and statistical data for each machine? Thoughts? -mm- |
CyberGoyle Send message Joined: 2 Jun 99 Posts: 160 Credit: 3,622,756 RAC: 26 |
memnoch beat me to the link for BOINCview. It will do the job nicely. <a> |
Mike Send message Joined: 17 Feb 01 Posts: 34258 Credit: 79,922,639 RAC: 80 |
hi Here is the Link of Boincview. http://boincview.amanheis.de greetz Mike |
EclipseHA Send message Joined: 28 Jul 99 Posts: 1018 Credit: 530,719 RAC: 0 |
You can make it a bit more "hands free" by using the "-return_results_immediately" command line parameter with boinc. Things don't get stuck in a "Ready to Report" mode, unless the schedulers are down. I know this works with the CLI versions (windows and Linux) but have not tried it with the windows GUI. With this parameter, result is uploaded and the scheduler is contacted right away. |
ric Send message Joined: 16 Jun 03 Posts: 482 Credit: 666,047 RAC: 0 |
> You can make it a bit more "hands free" by using the > "-return_results_immediately" command line parameter with boinc. Woody, everybody has other experiences using the above parameter let me run into an ugly proglem: I'm using 2 Management tools, one is BV. With both Management tools I DID NOT GET every time the statistical data, cumulated into BoincLog. This becomes a greater issue, running a couple of clients. THIS parameter has to be used in a carefull manner (my opinion) there will be to less time left, for the Management Tool to collect all the nice Datas as cputime completed an much much more. If you don't need/whant this, it's up to you. I made the experience to not use this parameter. friendly ric |
Christopher Hauber Send message Joined: 10 Feb 01 Posts: 196 Credit: 71,611 RAC: 0 |
You should note that just because it says "Ready to report" in the work tab, doesn't mean it hasn't already sent. Completed workunits are generally sent immediately (if your settings allow). As long as network access is not disabled, completed workunits will appear in the transfer tab until they are actually sent. Once they are sent, they stay in the work tab saying "ready to report" until you update the project. I don't know why that is, but just the way of things. You can confirm it by looking through the log and finding where things have been uploaded. > Management tasks multiply when you are running multiple computers. I like to > simplify these tasks. Therefore I have two questions: > > 1. Is there a way to make the "update" automatic so "ready to report" WUs go > by themselves? > > 2. Is there a way (a third-party-ad-on, perhaps) to see from a central > location the status and progress on each BOINC computer in the farm? With > SETI classic I use MSETIMON and SETIQUEUE - they give me a nice view into the > current state of each computer/WU and a nice way to look at historical and > statistical data for each machine? > > Thoughts? > > -mm- > > |
EclipseHA Send message Joined: 28 Jul 99 Posts: 1018 Credit: 530,719 RAC: 0 |
> > You can make it a bit more "hands free" by using the > > "-return_results_immediately" command line parameter with boinc. > > Woody, > > everybody has other experiences > using the above parameter let me run into an ugly proglem: > > I'm using 2 Management tools, one is BV. I've been using "-return_results_immediately" since the public launch (win and Linux), with things like boincview, and the only thing I've seen is that results don't sit in a "ready to report" state for a few hours (they are reported right after upload). If BV or other tools have a problem, it's probably a bug in the tool, and I'd report it to the tool developers. As the tools don't really impact the crunch/report cycle, I'll say that from my experience, reporting sooner, means credits sooner (you cant see it on seti, but you can on the other projects) > THIS parameter has to be used in a carefull manner (my opinion) > there will be to less time left, for the Management Tool to collect all the > nice Datas as cputime completed an much much more. All it "seems" to do is to contact the scheculer right after the result has been uploaded, vs waiting to contact the scheduler at a later time (the next completed WU in most cases), and therefore means a "tick" of CPU time now or hours from now.. No difference is the overall process.. BTW, I have BV set to use RPC and a second monitor which uses NFS (file) access, and not seen a single hickup..... |
EclipseHA Send message Joined: 28 Jul 99 Posts: 1018 Credit: 530,719 RAC: 0 |
> You should note that just because it says "Ready to report" in the work tab, > doesn't mean it hasn't already sent. Completed workunits are generally sent > immediately (if your settings allow). As long as network access is not > disabled, completed workunits will appear in the transfer tab until they are > actually sent. Once they are sent, they stay in the work tab saying "ready to > report" until you update the project. I don't know why that is, but just the > way of things. You can confirm it by looking through the log and finding where > things have been uploaded. Exactly why "-return_results_immediately" is a cool thing.. You dial-in to upload a result, and it get reported at the same time! Try it, you'll like it! (and then ask yourself "Why isn't this the default?") |
ric Send message Joined: 16 Jun 03 Posts: 482 Credit: 666,047 RAC: 0 |
>seen a single hickup..... be happy Woody, just be happy. It's fine if it's work yourside, that's basically important I don't know how many clients you have to "manage", myself, I had to reduce BV cpu usage, the p2400 512MB (m$, XP) as an monitoring station had to put 40-70%, sometime til 90% of the resources. It's not the "problem" of BV, it's the problem to have high load on BV. Always speaking for my environment, perhaps not the best settings, I use it on a 35 sek intervall, this for every client. this could reduce the cpu demand in a massive way. Now a question is there, do I want to use all cpu cycles to monitor or wouldn't I spend the boinc client as much as possible of the cycles? Really cool would be, having a tool doing the Update, or better spoken, the "ask for communication" (both directions) in an automated way, the tool would should be able to manage all clients of the farm. Without a user intervention. something like sending an Update every hour or so while we are sleeping this is missing.. this would be cool. Well, sorry no alfa, no beta, no zeta, no gamma, just a *ugly* user, came over on june 24. from seticlassic. I do have very low experience and I do have "nearly" never seen a wu.... ;-) |
Toby Send message Joined: 26 Oct 00 Posts: 1005 Credit: 6,366,949 RAC: 0 |
> You should note that just because it says "Ready to report" in the work tab, > doesn't mean it hasn't already sent. Completed workunits are generally sent > immediately (if your settings allow). As long as network access is not > disabled, completed workunits will appear in the transfer tab until they are > actually sent. Once they are sent, they stay in the work tab saying "ready to > report" until you update the project. I don't know why that is, but just the > way of things. You can confirm it by looking through the log and finding where > things have been uploaded. Uploading results is a 2-step proces. The result XML file is uploaded as soon as the work unit finishes (if network access is available). This does not give you any credits though. The credits are claimed by communicating with the scheduler which happens when you 'update' the project. It also happens automatically if the work unit that just finished is within 24 hours of its deadline or your cache has reached the low-water mark. When either of those conditions is met, BOINC automatically contacts the scheduler and updates your claimed credit and (maybe) asks for more work so as was pointed out above, there really is no need to click 'update' on your entire farm every day. It may take a couple of hours or days (depending on your cache settings) before you claim credit for the completed work units but it will happen automatically. Actually this is all explained in Paul's BOINC user manual. I never have a link handy when I need one though. His web page should be required reading for all new members! :) --------------------------------------- - A member of The Knights Who Say NI! Possibly the best stats site in the universe: http://boinc-kwsn.no-ip.info |
Paul D. Buck Send message Joined: 19 Jul 00 Posts: 3898 Credit: 1,158,042 RAC: 0 |
> Actually this is all explained in Paul's BOINC user manual. I never have a > link handy when I need one though. His web page should be required reading > for all new members! :) Just find where I have answered a thread, bookmark my site ... :) all kinds of ways to find me ... <p> For BOINC Documentaion: Click Me! |
©2024 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.