Message boards :
Number crunching :
The 400 and 50 WU limits are way too small
Message board moderation
Previous · 1 · 2 · 3 · 4 · Next
Author | Message |
---|---|
Claggy Send message Joined: 5 Jul 99 Posts: 4654 Credit: 47,537,079 RAC: 4 |
I just wish the GPU limit was a little higher than 400, though... Shorties are a killer. :( What i do is Bump the 'Maintain enough tasks to keep busy for at least --- days' setting up high enough so shorties are processed first, (or 'connect about every' setting in older Boinc Managers) with some suspending of CPU shorties, so too many don't run at once, i then restore the previous settings once the GPU shorties are complete, that has enabled me to build quite a few days cache up, it also helps if you can do Astropulse on CPU and GPUs, each Wu will fill your cache up just a bit more than a MB Wu does. (It also helps that i have a ATI GPU fitted, this doubles my GPU limit, so i get more Wu's for my Nvidia GPU, and a few AP Wu's for the ATI GPU, which then goes and does work elsewhere most of the time) Claggy |
red-ray Send message Joined: 24 Jun 99 Posts: 308 Credit: 9,029,848 RAC: 0 |
Wow, I just got a CPU WU, that makes a total of 3 in 24 hours! 20/02/2012 15:30:21 | SETI@home | Reporting 1 completed tasks, requesting new tasks for CPU 20/02/2012 15:30:25 | SETI@home | Scheduler request completed: got 1 new tasks but I noticed the system is not even asking for GPU WUs. After some digging I found this is down to an "NVIDIA GPU work fetch deferral interval" of 00:40:00. I guess this is great for the servers but means once a client has run dry the chance of new WUs is even less. Is this another 6.12 "feature" I wonder. Pressing [Update] seems to make it ask so I feel a .BAT file is needed. |
red-ray Send message Joined: 24 Jun 99 Posts: 308 Credit: 9,029,848 RAC: 0 |
Lookes like my .BAT file did the trick! 20/02/2012 16:49:14 | SETI@home | Sending scheduler request: Requested by user. 20/02/2012 16:49:14 | SETI@home | Requesting new tasks for CPU and NVIDIA GPU 20/02/2012 16:49:35 | SETI@home | Scheduler request completed: got 33 new tasks |
kittyman Send message Joined: 9 Jul 00 Posts: 51468 Credit: 1,018,363,574 RAC: 1,004 |
Eric and I have spoken about this. If he gets the AP validation algorithms corrected and successfully tested on Beta, the limits may be raised. Not before then. DA's attempt at 'fixing' certain other 'problems' skewed things way off kilter.......so far as to not allow the usual workfetch to work properly. Boinc properties of DCF and WTF have not worked properly since he screwed with them. It has taken much more time to recover from this than expected. But, it shall happen. It's particularly the WTF factor that I am waiting to see fixed......LOL. Best wishes and kitty kisses, DA. "Freedom is just Chaos, with better lighting." Alan Dean Foster |
Josef W. Segur Send message Joined: 30 Oct 99 Posts: 4504 Credit: 1,414,761 RAC: 0 |
"The 400 and 50 WU limits are way too small" for some computers, way too big for most. We need BOINC to provide a time-based limits option, restricting each host to about 2 days of work "in progress" would be appropriate for this project currently. Joe |
HAL9000 Send message Joined: 11 Sep 99 Posts: 6534 Credit: 196,805,888 RAC: 57 |
The idea for a super host has been around for some time: http://boinc.berkeley.edu/trac/wiki/SuperHost. I am well aware of how my machines act. That is why I have removed some of them the past few years. So I wouldn't be pounding the servers even more than I am now. I take advantage of those fancy network tools people have made to manage everything. Until we got some new network equipment I did have to get up and walk to the other side of the room to check on a few of the boxes, but now I can do it all without even rolling my chair 2 feet. It is nice to be able to see all of my machines at once. So I can see how the limits are effecting their queue. My faster ones end up with just a few days work. While the slower boxes hit their queue limit before the task limit. The least I heard the feeder held 100 tasks and refilled once a minute. SETI@home classic workunits: 93,865 CPU time: 863,447 hours Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[ |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14650 Credit: 200,643,578 RAC: 874 |
"The 400 and 50 WU limits are way too small" for some computers, way too big for most. We need BOINC to provide a time-based limits option, restricting each host to about 2 days of work "in progress" would be appropriate for this project currently.Joe Provided that the project servers - which is where such a limit would have to be applied - have an accurate estimate of the runtime of tasks cached. At the moment, the servers ignore DCF entirely when calculating their own idea of how long newly-allocated work will run during work allocation, and the well-rehearsed APR capping problem means that hosts are running some extreme DCFs, far from the 1.0000 taken for granted by CreditNew. (I've got a 0.0691) |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14650 Credit: 200,643,578 RAC: 874 |
The least I heard the feeder held 100 tasks and refilled once a minute. I thought it was 100 tasks too, but it has to be faster than once per minute. That would limit the project to 6,000 tasks per hour - we know the regular throughput is far, far higher than that. The feeder refresh interval will be every second or two. |
SciManStev Send message Joined: 20 Jun 99 Posts: 6652 Credit: 121,090,076 RAC: 0 |
"The 400 and 50 WU limits are way too small" for some computers, way too big for most. We need BOINC to provide a time-based limits option, restricting each host to about 2 days of work "in progress" would be appropriate for this project currently.Joe That was well spoken. In my situation I would like to see 2 days of CPU work, and two days of GPU work if that were possible, which still amounts to 2 days of work. That would be a huge improvement over current conditions. The next problem is just to get back to the limits we currently have. That alone can be no easy task, especially durinig a shortie storm. Steve Warning, addicted to SETI crunching! Crunching as a member of GPU Users Group. GPUUG Website |
HAL9000 Send message Joined: 11 Sep 99 Posts: 6534 Credit: 196,805,888 RAC: 57 |
The least I heard the feeder held 100 tasks and refilled once a minute. Perhaps I was just remembering "minute" instead of "second". As they say the memory is the 2nd thing to go, but no one can remember the first. Perhaps a variable "Max allocated tasks" like the "Max tasks per day" thing would be the best solution. SETI@home classic workunits: 93,865 CPU time: 863,447 hours Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[ |
Josef W. Segur Send message Joined: 30 Oct 99 Posts: 4504 Credit: 1,414,761 RAC: 0 |
The least I heard the feeder held 100 tasks and refilled once a minute. As of December 2010 the -sleep_interval command line value for the Feeder was 3 seconds. But that sleep time is only used if nothing was done during a pass through the array. The default array size is 100, and the default query to the BOINC database to enumerate the next tasks is sized 200. I think the 100 matches observed effects here, but the query size could be larger than default. The "nothing was done" could be because no requests were successful since the last Feeder pass or because there were no tasks ready to send. My guess is the usual pattern is the Feeder makes two or three passes through the array, doing something on the first, very little or nothing on the second, etc. That's assuming the database enumeration happens quickly each time, though. Joe |
kittyman Send message Joined: 9 Jul 00 Posts: 51468 Credit: 1,018,363,574 RAC: 1,004 |
I am wondering if the GPU limits could be raised, like, right NOW. What I understand is that the current holdup is fixing the AP validator. As no stock app exists for AP on GPUs.....and whack me if I am wrong...why can't the GPU limits be raised right now without much risk of damage? Leave the CPU limits where they are at if you must. It's the GPUs on the power hosts that are what is starved when anything goes slightly south with the servers. EDIT... I just emailed Eric.....we'll see what he thinks. "Freedom is just Chaos, with better lighting." Alan Dean Foster |
Lionel Send message Joined: 25 Mar 00 Posts: 680 Credit: 563,640,304 RAC: 597 |
it is rare for me to get up to the limits at any time ... in fact it is hard to get work at any time (wiggo keeps sucking the pipe dry) ... they need to be lifted otherwise we are all going to be knocking on the door all the time ... |
kittyman Send message Joined: 9 Jul 00 Posts: 51468 Credit: 1,018,363,574 RAC: 1,004 |
For now, I just hope the kitties can gather enough kibble overnight to run through tomorrow's outage. And hope they can get the deleter process back in order. I dunno where they are getting the disc space right now. "Freedom is just Chaos, with better lighting." Alan Dean Foster |
kittyman Send message Joined: 9 Jul 00 Posts: 51468 Credit: 1,018,363,574 RAC: 1,004 |
me i m getting I am nowhere near that, but the kitties have been gathering some kibbles into their bowls!! "Freedom is just Chaos, with better lighting." Alan Dean Foster |
Wiggo Send message Joined: 24 Jan 00 Posts: 34744 Credit: 261,360,520 RAC: 489 |
it is rare for me to get up to the limits at any time ... in fact it is hard to get work at any time (wiggo keeps sucking the pipe dry) ... they need to be lifted otherwise we are all going to be knocking on the door all the time ... It really isn't my fault if I'm closer to the point where those work units land in this country. ;) me i m getting Yes I'm getting very sick of seeing that message all the time. :( Cheers. |
kittyman Send message Joined: 9 Jul 00 Posts: 51468 Credit: 1,018,363,574 RAC: 1,004 |
it is rare for me to get up to the limits at any time ... in fact it is hard to get work at any time (wiggo keeps sucking the pipe dry) ... they need to be lifted otherwise we are all going to be knocking on the door all the time ... Yeah, I wish we could get things sorted and get back to Boinc calling the shots. So depressing when I can't have the rigs weather a day or two outage. "Freedom is just Chaos, with better lighting." Alan Dean Foster |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14650 Credit: 200,643,578 RAC: 874 |
Yeah, I wish we could get things sorted and get back to Boinc calling the shots. I hope Eric can get that assimilator sorted out for when the new download server arrives - after all, we'll want massive download congestion so we can see what it can do under real load (and give them a chance to fine tune the software). |
Wiggo Send message Joined: 24 Jan 00 Posts: 34744 Credit: 261,360,520 RAC: 489 |
I hope Eric can get that assimilator sorted out for when the new download server arrives - after all, we'll want massive download congestion so we can see what it can do under real load (and give them a chance to fine tune the software). Well at least the number to be assimilated has dropped but I wonder why the AP creation rate is so low. Cheers. |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14650 Credit: 200,643,578 RAC: 874 |
I hope Eric can get that assimilator sorted out for when the new download server arrives - after all, we'll want massive download congestion so we can see what it can do under real load (and give them a chance to fine tune the software). Ooops, sorry - it's a validator that Eric is working on, of course. 'Not enough coffee' error. |
©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.