Info | Message |
---|---|
1) Message boards : Questions and problems : Your computer has been measured at x.xx KB/sec. (Ibercivis)
Message 36841 Posted 13 Feb 2011 by manuel oliveira |
The above (31583) is not true. It happens with both Linux and Microsoft OS's. |
2) Message boards : Questions and problems : Starting New Tasks with Many Tasks in "waiting to run" State
Message 35565 Posted 31 Oct 2010 by manuel oliveira |
Thank you for your reply. I would like to add that this is happening even in computers running just wu's of one application of "just one" project, so the nature of this issue may be other than those described in http://boinc.berkeley.edu/trac/wiki/ClientSchedOctTen Regards ------- Part of http://boinc.berkeley.edu/trac/wiki/ClientSched This looks like a problem somewhere in... CPU scheduling policy The CPU scheduler uses an earliest-deadline-first (EDF) policy for results that are in danger of missing their deadline, and weighted round-robin among other projects if additional CPUs exist. This allows the client to meet deadlines that would otherwise be missed, while honoring resource shares over the long term. The scheduling policy is: 1. Set the 'anticipated debt' of each project to its short-term debt 2. Let P be the project with the earliest-deadline runnable result among projects with deadlines_missed(P)>0. Let R be P's earliest-deadline runnable result not scheduled yet. Tiebreaker: least index in result array. 3. If such an R exists, schedule R, decrement P's anticipated debt, and decrement deadlines_missed(P). 4. If there are more CPUs, and projects with deadlines_missed(P)>0, go to 1. 5. If all CPUs are scheduled, stop. 6. If there is a result R that is currently running, and has been running for less than the CPU scheduling period, schedule R and go to 5. 7. Find the project P with the greatest anticipated debt, select one of P's runnable results (picking one that is already running, if possible, else the one received first from the project) and schedule that result. 8. Decrement P's anticipated debt by the 'expected payoff' (the scheduling period divided by NCPUS). 9. Go to 5. The CPU scheduler runs when a result is completed, when the end of the user-specified scheduling period is reached, when new results become runnable, or when the user performs a UI interaction (e.g. suspending or resuming a project or result). CPU schedule enforcement The CPU scheduler decides what results should run, but it doesn't enforce this decision. This enforcement is done by a separate scheduler enforcement function, which is called by the CPU scheduler at its conclusion. Let X be the set of scheduled results that are not currently running, let Y be the set of running results that are not scheduled, and let T be the time the scheduler last ran. The enforcement policy is as follows: 1. If deadline_missed(R) for some R in X, then preempt a result in Y, and run R (preempt the result with the least CPU wall time since checkpoint). Repeat as needed. 2. If there is a result R in Y that checkpointed more recently than T, then preempt R and run a result in X. (...something wrong in the scheduler enforcement function?) |
3) Message boards : Questions and problems : Starting New Tasks with Many Tasks in "waiting to run" State
Message 35562 Posted 31 Oct 2010 by manuel oliveira |
Unfortunately, after some time working well, this fault returned both on windows and linux OS machines. So I am now using 6.10.17 / 18 without issues, 100% OK. Regards. |
4) Message boards : Questions and problems : Starting New Tasks with Many Tasks in "waiting to run" State
Message 35330 Posted 21 Oct 2010 by manuel oliveira |
Using now version 6.10.58 and all is Ok! Regards. |
5) Message boards : Questions and problems : Starting New Tasks with Many Tasks in "waiting to run" State
Message 35093 Posted 4 Oct 2010 by manuel oliveira |
In my opinion this is a bug. As I have already mentioned somewhere in this forum, it is not normal to start work units after wu's, without finishing them, reaching deadline, a complete mess... Downgrading to 6.10.17, the same work is orderly performed(FIFO). This is easily seen when crunching very small work units. This happens to me while working for EDGeS@home and Ibercivis projects using 6.10.5x's both linux and microsoft OS's. Regards. |
6) Message boards : News : BOINC 6.10.56 released
Message 33414 Posted 16 Jun 2010 by manuel oliveira |
Just to summarize, when upgraded from 61017 to 61056, many tasks were started but left incomplete to waiting to run status. Then, after a reset to the Ibercivis project, new tasks have been received and everything was (FIFO) working perfectly. Some hours later, attached to EDGeS project. The problem started again and worse is that several Ibercivis sanidad processes (60MiB+ per processs) and EDGeS fusion processes (400MiB+ per processs) were sleeping but occupying system resources. A quad desktop had 4 running EDGeS processes, 4 sleeping EDGeS processes plus 4 sleeping Ibercivis processes - sleeping processes occupying about 2GiB memory. Resource share doesn't work but it works with the combination of Ibercivis and ABC projects. It would be nice to totally separate prefs, no general prefs from one project to all, just to manage projects individually. Is it possible? Back to 61017 and no problems. (Sanidad tasks, 10 minutes each and EDGeS tasks, 10 to 40 minutes each.) Tks. |
7) Message boards : News : BOINC 6.10.56 released
Message 33379 Posted 15 Jun 2010 by manuel oliveira |
I have seen this happening after two different type of events. 1- Boinc is now "Suspending computation - CPU usage is too high" and 2- Scheduled update of project with a switch to high priority mode. If event 1 occurs frequently, as BOINC loads new wus, a lot of them are in waiting to run status. Solution: "Suspend work if non-BOINC load exceeds 0%" I dont remember of so frequent changes of priority mode before, no idea yet... Tks |
8) Message boards : News : BOINC 6.10.56 released
Message 33371 Posted 15 Jun 2010 by manuel oliveira |
I think you may have misunderstood. By FIFO, I mean the order in which Boinc received the WU. I have learnt that in digital electronics long ago. I have just reinstalled 6.10.56 in a quad machine and I can assure you that it didn't follow the order the tasks were received. It jumped more than 30 tasks and ignored the oldest ones(fifo ok by 6.10.17) that were already more than 30% done, and are now in waiting to run status. But I will check it better... Tks anyway. |
9) Message boards : News : BOINC 6.10.56 released
Message 33368 Posted 14 Jun 2010 by manuel oliveira |
Boinc has always run in FIFO within a project. Well, I have noticed that it is not working in FIFO basis. I had a lot of wu's -just one project, no mix- started but not completed, in waiting to run state. My wu's are usualy sorted out by Report deadline and only this version jumps forth and backwards... I had several hours of work in a lot of "waiting to run" tasks, then decided to go back and use 6.10.17 in all linux machines and I guess 6.10.18 to a pair of microsoft ones, and now no loss of work as it is now running FIFO. Tks. |
10) Message boards : News : BOINC 6.10.56 released
Message 33358 Posted 14 Jun 2010 by manuel oliveira |
Understood, tks. This version sometimes stops a wu before ending it and starts a new one, not respecting the Report deadline order. This puts some problems as you can imagine... Tks again. manuel oliveira |
11) Message boards : News : BOINC 6.10.56 released
Message 33354 Posted 14 Jun 2010 by manuel oliveira |
Hi all, I have installed 6.10.56 in some AMD 965 machines running openSuse 11.2 and I've noticed that several work units are not finished, are left "waiting to run", and others, more recent, are started. Example: a work unit is 90% done but left waiting to complete. Meanwhile it's deadline is reached and work unit - ibercivis and edges - is aborted by project servers. Back to old 6.10.17 version, and with same wu's, their report time is orderly followed, none is left before completion. The Activity feature "Use GPU: never" does not prevent pojects from asking work for GPU, even projects with no GPU applications - edges for example. |
12) Message boards : Questions and problems : Your computer has been measured at x.xx KB/sec. (Ibercivis)
Message 31583 Posted 14 Mar 2010 by manuel oliveira |
I kept testing and found out that this problem just happens to machines running linux. Microsoft OS machines using same connection never displayed this error / message. I've found other users complaining about this issue, old versions of BOINC for linux included (2008). Just while using linux OS, other OS machines sharing same internet connection, do not suffer of this. |
13) Message boards : Questions and problems : Your computer has been measured at x.xx KB/sec. (Ibercivis)
Message 30858 Posted 6 Feb 2010 by manuel oliveira |
Ok I will do so. Thanks a lot for your support. |
14) Message boards : Questions and problems : Your computer has been measured at x.xx KB/sec. (Ibercivis)
Message 30856 Posted 6 Feb 2010 by manuel oliveira |
Yes, any value above minimum required (in this case 1.96 KB/sec) is ok. Afterwards it will be adjusted. Now the part of my previous post: whenever BW measurement falls under a certain threshold, the application should not stop getting and updating the client_state.xml file with more BW measurements readings in order to enable the project to send work for that machine. What is happening now is that for some reason adsl 2+ goes down when a dw/load is about to be performed. From that moment on, no more work will be sent to that machine, as the file will not be updated anymore. How to avoid this? Is it a Boinc or a Ibercivis issue? Tks |
15) Message boards : Questions and problems : Your computer has been measured at x.xx KB/sec. (Ibercivis)
Message 30852 Posted 6 Feb 2010 by manuel oliveira |
I have same as yours so I will use your values. But it seems that any project servers can change this value "Your computer has been measured at 1.93 KB/sec." That is fine if they do perform several measurements. If they do one and get a value out of their range and simply abandon that peer, that could be altered in order to give one more chance do that pc, right? Tks. |
16) Message boards : Questions and problems : Your computer has been measured at x.xx KB/sec. (Ibercivis)
Message 30848 Posted 6 Feb 2010 by manuel oliveira |
Yes that's it. I was doing the change with boinc running. So now the value I entered, 12405.429504, allows the receipt of new tasks from the project server. All back to normal state. Tks |
17) Message boards : Questions and problems : Your computer has been measured at x.xx KB/sec. (Ibercivis)
Message 30846 Posted 6 Feb 2010 by manuel oliveira |
In the client_state.xml I found this: <net_stats> <bwup>12405.429504</bwup> ... <bwdown>1978.895233</bwdown> I have edited and corrected the bwup value but it returns to the same value. I will keep on trying... Tks. ----------------- I was doing the change with boinc running. So now the value I entered, 12405.429504, allows the receipt of new tasks from the project server. All back to normal state. Tks |
18) Message boards : Questions and problems : Your computer has been measured at x.xx KB/sec. (Ibercivis)
Message 30842 Posted 6 Feb 2010 by manuel oliveira |
Hello all, I am getting this type of error message from Ibercivis server. I've contacted isp, all running fine. Ibercivis have no idea so far. My question is: Is it possible for Boinc to double chech network bandwidth availability? My other question is: This check is controlled by BOINC or by Ibercivis servers? Thank you very much. Sat 06 Feb 2010 03:10:42 AM WET ibercivis Message from server: No work sent Sat 06 Feb 2010 03:10:42 AM WET ibercivis Message from server: No work is available for nanoluz Sat 06 Feb 2010 03:10:42 AM WET ibercivis Message from server: nanoluz requires 1.96 KB/sec download bandwidth. Your computer has been measured at 1.93 KB/sec. Sat 06 Feb 2010 03:10:42 AM WET ibercivis Message from server: adsorcion requires 1.96 KB/sec download bandwidth. Your computer has been measured at 1.93 KB/sec. Sat 06 Feb 2010 03:10:42 AM WET ibercivis Message from server: amiloide-docking requires 1.96 KB/sec download bandwidth. Your computer has been measured at 1.93 KB/sec. |
Copyright © 2024 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.