Message boards :
Number crunching :
Anyone else not getting work???
Message board moderation
Previous · 1 · 2 · 3 · 4 · Next
Author | Message |
---|---|
Donald L. Johnson Send message Joined: 5 Aug 02 Posts: 8240 Credit: 14,654,533 RAC: 20 |
As I recall from previous discussions, if you have ANY tasks Suspended or running in High Priority mode, BOINC will NOT request new work from ANY project. Appreciate the correction. But with 17 S@H Cuda Tasks suspended, his BOINC should still not request new S@H work, not even CPU Tasks, right? Donald Infernal Optimist / Submariner, retired |
Claggy Send message Joined: 5 Jul 99 Posts: 4654 Credit: 47,537,079 RAC: 4 |
As I recall from previous discussions, if you have ANY tasks Suspended or running in High Priority mode, BOINC will NOT request new work from ANY project. True, Claggy |
BigWaveSurfer Send message Joined: 29 Nov 01 Posts: 186 Credit: 36,311,381 RAC: 141 |
Awwww....what happened to all the AP ATI/AMD tasks?! I was receiving a lot of them, spitting them out quickly, then all of sudden....no more :( I guess the AP tasks go as quickly as they come for the GPU's. Hopefully after the long weekend we will get treated to some more. Is there any way to configure the prefs for each individual computer for what tasks you want to accept, not just the general overall ones? |
Claggy Send message Joined: 5 Jul 99 Posts: 4654 Credit: 47,537,079 RAC: 4 |
Awwww....what happened to all the AP ATI/AMD tasks?! I was receiving a lot of them, spitting them out quickly, then all of sudden....no more :( I guess the AP tasks go as quickly as they come for the GPU's. Hopefully after the long weekend we will get treated to some more. All the AP tasks have been split, now you're just got to wait for new tapes to be loaded, Is there any way to configure the prefs for each individual computer for what tasks you want to accept, not just the general overall ones? No, you can't configure individual hosts for what apps they run (via the website), But you have four venues/locations, so you can setup four different sets of preferences, but that isn't always good enough, (you can't exclude AP from the CPU and only send them to the GPU, and the same with MB) If you do an app_info you can have any combination you want, ie AP only on the CPU, and MB only on the GPU, or vice versa. Claggy |
BigWaveSurfer Send message Joined: 29 Nov 01 Posts: 186 Credit: 36,311,381 RAC: 141 |
[/quote] No, you can't configure individual hosts for what apps they run (via the website), But you have four venues/locations, so you can setup four different sets of preferences, but that isn't always good enough, (you can't exclude AP from the CPU and only send them to the GPU, and the same with MB) If you do an app_info you can have any combination you want, ie AP only on the CPU, and MB only on the GPU, or vice versa. Claggy[/quote] Thanks. I'm not looking to get too picky, but I would like the abilty to run more GPU wu's on the stronger machines. Let the computer 'stock-up' on those so it has a nice batch so it does not run out. Oh well, might just leave it all auto and let seti do the brain work for me. |
Worf Send message Joined: 17 May 99 Posts: 1 Credit: 41,381,790 RAC: 35 |
Dont know if this is the right thread, but I am only running one W/U on my GPU while my 6 cores & 12 threads sit idle. I should be firing on all 12 threads. I think this has been for about 2 weeks now. Am I not getting additional W/U or did something else creep in?? Anyone? thanks gary |
Donald L. Johnson Send message Joined: 5 Aug 02 Posts: 8240 Credit: 14,654,533 RAC: 20 |
Dont know if this is the right thread, but I am only running one W/U on my GPU while my 6 cores & 12 threads sit idle. I should be firing on all 12 threads. I think this has been for about 2 weeks now. Am I not getting additional W/U or did something else creep in?? If you are talking about this machine, you have 128 GPU tasks erroring out on 197 Exit_Time_Limit_Exceeded, because of a problem with the GPU not being usable and the Task falling back to the CPU. Others more familiar with GPUs than I will be along to offer more specific help, but you should probably check that you have the right NVidia (NOT MS) drivers for that card (check this sticky thread), and that it is not overheating. Donald Infernal Optimist / Submariner, retired |
Fred E. Send message Joined: 22 Jul 99 Posts: 768 Credit: 24,140,697 RAC: 0 |
Dont know if this is the right thread, but I am only running one W/U on my GPU while my 6 cores & 12 threads sit idle. I should be firing on all 12 threads. I think this has been for about 2 weeks now. Am I not getting additional W/U or did something else creep in?? Agree w/Donald, the errors are probably from the sleep bug in your 295.73 drivers as described in Claggy's sticky thread. . Upgrade or roll back. You'll continue to get gpu tasks until BOINC thinks it has reached your settings for days on hand and quits asking for gpu work You can get some cpu work by temporarily turning off "use GPU" in your Seti@Home preferences on the website. That won't impact crunching what you have, it just tells BOINC on the next update to not request gpu work. When you have enough cpu tasks, turn the gpu option back on. Worf, welcome to the forums - don't wait 13 years to post again! Another Fred Support SETI@home when you search the Web with GoodSearch or shop online with GoodShop. |
Paul van Dijken Send message Joined: 19 Jun 10 Posts: 4 Credit: 1,676,276 RAC: 1 |
I am not getting new work on one of my systems. Not for SETI not for any of the other projects. It looks like to be an issue with BOINC, although I have the same version on both HW. Just the one not receiving is 3 years old |
Nemesis Send message Joined: 14 Mar 07 Posts: 129 Credit: 31,295,655 RAC: 0 |
I'm not getting any CPU work, and have not for a couple of weeks. I recently upgraded my video from an "old" 8800GTX Ultra to a new 680GTX. Seems that since around that time I only get GPU work and the 4 CPU's are idle... Played around with the cache settings but nothing changed other than a bunch more DL's of GPU work...grrr. Any suggestions or advice would be appreciated! |
rob smith Send message Joined: 7 Mar 03 Posts: 22190 Credit: 416,307,556 RAC: 380 |
Until your nice shiny new 680 has its belly full you will not get any more CPU tasks - its the way BOINC works, feed the fastest workers first.... Bob Smith Member of Seti PIPPS (Pluto is a Planet Protest Society) Somewhere in the (un)known Universe? |
Mike Send message Joined: 17 Feb 01 Posts: 34256 Credit: 79,922,639 RAC: 80 |
Uncheck use nvidia GPU on your preferences page. Increase cache settings to 7 days. After your cache built up enable use nvidia GPU again. With each crime and every kindness we birth our future. |
Cherokee150 Send message Joined: 11 Nov 99 Posts: 192 Credit: 58,513,758 RAC: 74 |
This is an idea for all of us who have been frustrated by the cheaters, i.e., the projects that play with the rules (usually deadlines) to push to the front of the queue and keep all the other projects from running. I have always had a special dislike for anyone who attempts to take control of my computers from me, even if it is for a good cause (my main unruly project's name starts with an "E" and ends with "n"...I'll let you figure it out ;) ), so I set out to find a solution. By trial and error I found a quick and simple way to keep them under control. I keep the offending project set to "No new tasks". When I decide it's time to let that project process some, I change it to "Allow new tasks", then tell the project to "Update". Usually the project will send you a new batch of tasks right away. As soon as the line with "...got xx new tasks" pops up I quickly set the project back to "No new tasks". Then I let those tasks run out and repeat the process when I feel like letting that project have some more cycles. If the offender should send a few too many units in the batch, I just "Suspend" the project somewhere in the middle of processing the current batch. When I want to let the batch have some more time, I just "Resume" the project. It works great, and only takes a few seconds of my time every now and then. The other projects are basically unaffected by this and continue to request work as needed. As a -very- long-time programmer, I realize that sometimes simple manual intervention works better than all the programming switches and logic. ;) Let's keep those pc hijackers under control, eh? :) |
Keith Myers Send message Joined: 29 Apr 01 Posts: 13164 Credit: 1,160,866,277 RAC: 1,873 |
This too has been my solution for E@H. That project just didn't want to play by the rules .... never obeyed my disk space limitations or resource share and always downloaded way more work than it reasonably could finish before deadlines and always forced itself into Priority mode and chucked my other projects in the round file. Now I just get a under a dozen tasks with an Update and NNT and let it run roughshod over my other projects like usual and then put in back in the closet for a week until I feel like my other projects can take the RAC hit for a day. Keith Seti@Home classic workunits:20,676 CPU time:74,226 hours A proud member of the OFA (Old Farts Association) |
Matt Send message Joined: 6 Jun 10 Posts: 18 Credit: 1,434,540 RAC: 0 |
No new work for several weeks. Nothing for CPU, nothing for GPU. I have SETI@Home set for a majority share of my work allotment, but still nothing... |
Grant (SSSF) Send message Joined: 19 Aug 99 Posts: 13731 Credit: 208,696,464 RAC: 304 |
Work has been flowing for several weeks. If you post the first 30 or so lines from your message log, people might be able to help you figure out why you're not getting any work. EDIT- you're also subscribed to a lot of projects, so if you could post exactly what your work allocation is that would help as well. Grant Darwin NT |
kittyman Send message Joined: 9 Jul 00 Posts: 51468 Credit: 1,018,363,574 RAC: 1,004 |
It's a little tight right now......AP is splitting again and taking up a good portion of the bandwidth. But the kitties have had little trouble 'caching up' when AP runs out to keep things rolling here. "Freedom is just Chaos, with better lighting." Alan Dean Foster |
Guilherme Tomazini Send message Joined: 15 Jul 11 Posts: 1 Credit: 166,816 RAC: 0 |
Asking only for ATI work. I don't get any work here either. |
bill Send message Joined: 16 Jun 99 Posts: 861 Credit: 29,352,955 RAC: 0 |
There is probably 4 ATI requesters for every 1 ATI AP work unit produced. |
rob smith Send message Joined: 7 Mar 03 Posts: 22190 Credit: 416,307,556 RAC: 380 |
There is NO SUCH THING as an "ATI" work unit. All work units are either "Astropulse", or "Multibeam", and as such they are available to any processor that is capable of processing them. At the time of a processor making a request the server looks to see if it has any work units available in the delivery queue, and if there are any restrictions on them being delivered to that processor. These restrictions may be "human", in the form of "don't send my any Astopulses", or "machine", processor is judged to be incapable of handling a particular work unit. Provided the work unit and the limitations are OK, then the work unit is set as a task for that particular processor. The number Asropluse work units that can be split from a tape is significantly smaller than the number of Multibeam work units that can be split for two reasons: There is no overlap between Astopulses, but there is an overlap on Multibeams; and the bandwidth of an Astropule is considerably larger than that of a Multibeam. Each tape is split twice, once to give Astropulses, and once to give Multibeams. (In this context "processor" is either the CPU, or GPU on a given cruncher) Bob Smith Member of Seti PIPPS (Pluto is a Planet Protest Society) Somewhere in the (un)known Universe? |
©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.