| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Support
Forum: BOINC Agent Support
Thread: Servers don't give me the numbers of WUs I asked |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 26
|
|
| Author |
|
|
Sekaiji
Cruncher France Joined: Dec 15, 2005 Post Count: 20 Status: Offline Project Badges: 
|
Hello, I'm requesting for some help.
I've got BOINC 5.10.20, a C2D 4300, and I work only and WGC projects (no others projects in BOINC). The past week, when I had the 5.10.13 I modified in BOINC the "connect everey about : 0.9 days" and "additional work buffer : 0.25" to 0.9/2. It was ok, I get something like 14 more WUs queued. 5 minutes later, I put 0.9/4 and get somthing like 14 others units. 5 minutes later, I put 0.9/7 and get the total of 39 WUs queued. Finaly I put the defaut settings back : 0.9/0.25 Everything was ok. Some hours after, I installed the 5.10.20 Three days later (approximatively), I had almost finished my list of tasks, and since that, I'm unable to get more works by myself with the update button. BOINC asks itself for more works and the server gives it only 1 by 1 WU. So, I've got in my list of tasks only 3 WUs, 2 are working and 1 is waiting. Everytime I push the update button, in the message window I've got this : 12/09/2007 01:32:26|World Community Grid|Sending scheduler request: Requested by user 12/09/2007 01:32:26|World Community Grid|(not requesting new work or reporting completed tasks) 12/09/2007 01:32:31|World Community Grid|Scheduler RPC succeeded [server version 509] 12/09/2007 01:32:31|World Community Grid|Deferring communication for 5 min 3 sec 12/09/2007 01:32:31|World Community Grid|Reason: requested by project To solve the problem, I rebooted my pc, I uninstalled BOINC 5.10.20 to retry with the 5.10.13, that didn't work, so I reinstalled the 5.10.20. When I choose the preferences frome WGC, it's the same thing. 4 days later, I'm here, posting this message and asking for help. I'd like to get 20 WUs in my queue. Why ? Because Somedays I have not Internet, so BOINC won't work and that is not an option for me  Could you, please, tell me what I must do ? Thanks |
|||
|
||||
|
Sgt.Joe
Ace Cruncher USA Joined: Jul 4, 2006 Post Count: 7849 Status: Offline Project Badges: 
|
See this thread for an explanation from Ingleside:
----------------------------------------https://secure.worldcommunitygrid.org/forums/wcg/viewthread?thread=15587#120592 Cheers
Sgt. Joe
*Minnesota Crunchers* |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hello Sekaiji,
Finaly I put the defaut settings back : 0.9/0.25 I do not understand your settings. In order for BOINC 5.10 to meet your situation, you should connect fairly often (0.1 to 0.3 days) while requesting a fairly large cache (about 1.5 days). Instead, your settings are the opposite. Lawrence |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
The rapid changing of connect 0.9 with the different additional buffer settings, particular the reduction of values with lots of tasks in buffer has upset the BOINC scheduler. Assuming you are permanently connected, but want a safety in case any connection goes wrong, your best bet is WCG default connect of 0.3 (7.2 hours) and 1.5 days additional buffer. For a 4300 that will give approx.20, probably more. It changes all the time as the durations of jobs vary.
----------------------------------------To my own dismay discovered that the ',' locale is ignored (cont.Europe). The values have to be entered with a dot as decimal. Using a comma as decimal may store an entirely different value. At any rate, getting 1 at the time suggests you managed to confused things enough, that you best clear all the local settings (there's a button for that in the local prefs screen), so it can use the website profile again and after modify the local values for connect and additional buffer again as per earlier recommendation. After, leave it alone and it will find its balance, which can take a few days to a week. If you see 'high priority' in the Tasks window or a Duration Correction Factor in the client_state file of greater than 1.5, those are more indicators, a 'hands-off' period needs consideration.
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Sekaiji
Cruncher France Joined: Dec 15, 2005 Post Count: 20 Status: Offline Project Badges:
|
Thank you all
Sgt.Joe, I already saw that post https://secure.worldcommunitygrid.org/forums/wcg/viewthread?thread=15587#120592 but it did'nt help me enough lawrencehardin yes your right, but I also tried value like 0.1/0.45 and it doesn't work either Sekerob "The rapid changing of connect 0.9 with the different additional buffer settings, particular the reduction of values with lots of tasks in buffer has upset the BOINC scheduler. Assuming you are permanently connected, but want a safety in case any connection goes wrong, your best bet is WCG default connect of 0.3 (7.2 hours) and 1.5 days additional buffer. For a 4300 that will give approx.20, probably more. It changes all the time as the durations of jobs vary." I tried your settings 0.3/1.5 but not more success "To my own dismay discovered that the ',' locale is ignored (cont.Europe). The values have to be entered with a dot as decimal. Using a comma as decimal may store an entirely different value." I tried to put values with a dot and it didn't work, it put some stranges values when I reopen the preferences window to verify, but when I put a coma, the value are ok. "At any rate, getting 1 at the time suggests you managed to confused things enough, that you best clear all the local settings (there's a button for that in the local prefs screen), so it can use the website profile again and after modify the local values for connect and additional buffer again as per earlier recommendation. After, leave it alone and it will find its balance, which can take a few days to a week." I reseted many times local preferences to take again the WGC value, but it's the same problem of "not requesting new work". If you see 'high priority' in the Tasks window or a Duration Correction Factor in the client_state file of greater than 1.5, those are more indicators, a 'hands-off' period needs consideration. You're right, my tasks since I've got the problem always are in "running, high priority". My Duration Corection Factor is 1.009874 So what I can do now ? Just to wait ? Actually my settings are 0.2/0.25 from WGC. And if I change the value of the duration to a minor value ? Actually, I've got a C2D 4300 but @3Ghz so, I do something like 12 or 14 results per day (my computer works 16 hours a day) Thanks |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
These are the 2 values that can appear in both the global_prefs.xml and global_prefs_override.xml
----------------------------------------<work_buf_min_days>0.300000</work_buf_min_days> <work_buf_additional_days>1.500000</work_buf_additional_days> The global_prefs has a default section and also venue sections, so the values can appear multiple times. To set them, use an text editor and modify the global_prefs_override.xml. After, go to Advanced menu and take the "Read local preferences" Further check in activity menu: 1. network according preferences - Important!!!!!! 2. work according preferences If you set it to working off-line, yet you got 0.3 / 1.5 the system will think it wont be able to report in time. Each time you hit the Update button, it opens the line for 5 minutes. The high priority is a confirmation that some setting has put your client in panic mode. When in panic mode it will only fetch 1 job at a time. Your DCF is perfectly fine Yep, the comma/dot conflict is not resolved yet and am still testing, hence why i have reverted to manual editing of the pref files to ensure i got the right values stored.
WCG
----------------------------------------Please help to make the Forums an enjoyable experience for All! [Edit 1 times, last edit by Sekerob at Sep 12, 2007 3:45:07 PM] |
||
|
|
Sekaiji
Cruncher France Joined: Dec 15, 2005 Post Count: 20 Status: Offline Project Badges:
|
I'm not sure of what I must do now.
I change values in global_prefs_override.xml to 0.3/1.5, ask it to read my local preferences. I'm in network according preferences and work according preferences. And, I'm still in panic mode, and when I hit the update button : the same result "not requesting work" You said : some setting has put your client in panic mode. Wich settings can I touch now to peacefull my agent ? At this state my settings are : -preferences local 0.3/1.5 -panic mode (running high priority) -1 waiting WUs, 2 working (deadlines near 22 sept) So what can I do now ? Hope a new version of boinc will include a function "how much WUs do you want for this project ?" and if people doesn't respect deadlines, servers will give them less maximum tasks the next time they'll ask. |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
For now, just let it sit. As long as it fetches work 1 by 1, you're okay. Eventually it should work it self out of the state of emergency.
----------------------------------------No there's wont be a X WU buffering. All is time driven and with multiple projects on a single client those can vary tremendously.
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Sekaiji
Cruncher France Joined: Dec 15, 2005 Post Count: 20 Status: Offline Project Badges:
|
Ok, hope it won't be long to have a normal behaviour.
Thnaks |
||
|
|
Sekaiji
Cruncher France Joined: Dec 15, 2005 Post Count: 20 Status: Offline Project Badges:
|
Hello, 15 days later, I've still have the same problem.
Always running in high priority mode and still don't have WUS in advance. I let the settings you advised me Sekerob : 0,2/1,5 in BOINC (BOINC 5.10.20 C2D 4300) |
||
|
|
|