Message boards :
Number crunching :
Not getting tasks, even though the server has 10000 available.
Message board moderation
Author | Message |
---|---|
Send message Joined: 5 Jul 11 Posts: 990 Credit: 376,142,956 RAC: 2 |
It would appear the server refuses to give me tasks until I've been inactive for about 15 minutes. I get 600 tasks at a time, and Boinc gives the results back every minute and a half as requested by the server (see log below). But since Boinc reports every minute and a half, the server won't give me any more tasks until Boinc stops pestering the server. Is there some kind of hidden timer that the server isn't telling my client about? Why is the server saying "wait 1.5 minutes" when it really needs me to wait 15 minutes? 887 Milkyway@Home 29/10/2019 11:43:01 PM [sched_op] Starting scheduler request 888 Milkyway@Home 29/10/2019 11:43:01 PM Sending scheduler request: To fetch work. 889 Milkyway@Home 29/10/2019 11:43:01 PM Reporting 1 completed tasks 890 Milkyway@Home 29/10/2019 11:43:01 PM Requesting new tasks for AMD/ATI GPU 891 Milkyway@Home 29/10/2019 11:43:01 PM [sched_op] CPU work request: 0.00 seconds; 0.00 devices 892 Milkyway@Home 29/10/2019 11:43:01 PM [sched_op] AMD/ATI GPU work request: 29694.34 seconds; 0.00 devices 893 Milkyway@Home 29/10/2019 11:43:04 PM Scheduler request completed: got 0 new tasks 894 Milkyway@Home 29/10/2019 11:43:04 PM [sched_op] Server version 713 895 Milkyway@Home 29/10/2019 11:43:04 PM Project requested delay of 91 seconds 896 Milkyway@Home 29/10/2019 11:43:04 PM [sched_op] handle_scheduler_reply(): got ack for task de_modfit_83_bundle4_4s_south4s_bgset_2_1570151403_12432789_1 897 Milkyway@Home 29/10/2019 11:43:04 PM [sched_op] Deferring communication for 00:01:31 898 Milkyway@Home 29/10/2019 11:43:04 PM [sched_op] Reason: requested by project |
Send message Joined: 16 Mar 10 Posts: 208 Credit: 105,447,129 RAC: 36,580 |
This is a currently-known issue which probably relates to some server-side configuration issue, as other projects seem capable of issuing work on a combined "reporting" and "request new work" contact. There are already existing threads on this subject - see Delay in getting new work units untill all work units have cleared and WUs not downloaded in time - rig is idling - doing no work ..., which you may [or may not :-)] find informative. Cheers - Al. |
Send message Joined: 5 Jul 11 Posts: 990 Credit: 376,142,956 RAC: 2 |
Thanks, the search function for these forums is terrible, I can never find a discussion already in progress. |
Send message Joined: 8 May 09 Posts: 3315 Credit: 519,946,492 RAC: 22,331 |
Thanks, the search function for these forums is terrible, I can never find a discussion already in progress. Change the box to 'newest post first' and the latest ones should be up top just below the permanent threads. |
Send message Joined: 5 Jul 11 Posts: 990 Credit: 376,142,956 RAC: 2 |
I mean if I do a search for keywords in the forum. Instead of a list of titles like google would do for a web search, everything is expanded and takes up many screens full. It's quite weird. Just tried again, I never noticed the small section above of title matches! Because most of the screen is taken up with expanded messages where someone happened to mention that keyword in a reply. Anyway, it never seems to match what I'm looking for, even if I search "GPU" which should have 1000s of results, I get ONE! I think I sorted it, I should go into the advanced search, as the default is only the last month of results, which is rather small. |
Send message Joined: 8 May 09 Posts: 3315 Credit: 519,946,492 RAC: 22,331 |
I mean if I do a search for keywords in the forum. Instead of a list of titles like google would do for a web search, everything is expanded and takes up many screens full. It's quite weird. Yes the advanced search is MUCH better, but as you say if you are too specific you don't get what you want and if you are too generic you get thousands of what you don't want. |
Send message Joined: 5 Jul 11 Posts: 990 Credit: 376,142,956 RAC: 2 |
I've got used to Google just finding everything by itself. Remember the days before Google when we used various search engines which always gave the wrong results? |
Send message Joined: 8 May 09 Posts: 3315 Credit: 519,946,492 RAC: 22,331 |
Yes I do!! |
Send message Joined: 7 Dec 09 Posts: 11 Credit: 637,858,888 RAC: 0 |
Well, my local bank has millions of euros available, but they aint gonna give'm to me. I guess that's life, usually we dont get what we want, I'm used to it by now, i just ignore it and move on. |
Send message Joined: 5 Jul 11 Posts: 990 Credit: 376,142,956 RAC: 2 |
You clearly don't belong to a loud mouthed union then :-) |
Send message Joined: 18 Nov 08 Posts: 291 Credit: 2,461,693,501 RAC: 0 |
Just went through this with LHC. They had plenty of atlas work units, most 1gb in size (VBox image) but only my Linux system could get work. Turned out the checkbox "Run native if available" should have been worded "Don't run anything except native code" After fixing the problem I got 5 huge withdrawals. The 1gb+ withdrawals from the :"bank" on two system tied up my network for 20 minutes. Atlas tasks can take up to 8 cpus which is nice to use with Milkyway GPU tasks. |
Send message Joined: 5 Jul 11 Posts: 990 Credit: 376,142,956 RAC: 2 |
I thought you could get 64 core ones? Somebody mentioned a Ryzen threadripper doing it. I just had to rearrange how my systems are set up, due to Atlas tasks misbehaving in two ways. I did have it with Boinc set to use all 6 cores, but with the Einstein GPU tasks set to require a certain amount of CPU, so Boinc would reduce the number of CPU tasks accordingly and never throttle the GPU. This works fine with multiple single core CPU tasks eg. Rosetta. But when Boinc downloads Atlas, it gets a WU to use all the cores I've allocated, ignoring the fact that the GPU will always need some. And then, while it's running, Virtualbox gets set to normal priority instead of idle, so it takes the CPU in preference to the GPU getting to use it. I've fixed this by just finding out the most CPU cores ever needed by anything I run on the GPU, and setting Boinc to have that as how many are used (and telling it the GPU tasks need 0.01 CPUs - zero confuses it), now I get Atlas tasks using less cores. |
Send message Joined: 7 May 14 Posts: 57 Credit: 201,100,267 RAC: 23,304 |
hi all made vid on youtube for multiple instances instruction's and at full load on a Radeon VII RADEON VII GIGABYTE// 3 Instances_ Milkyway@home WUs BOINC_ 3_instances https://www.youtube.com/watch?v=4xKy9wGKmz4 all the best and welcome to earth |
©2024 Astroinformatics Group