Message boards :
Number crunching :
Panic Mode On (97) Server Problems?
Message board moderation
Previous · 1 . . . 22 · 23 · 24 · 25 · 26 · 27 · 28 . . . 33 · Next
Author | Message |
---|---|
Speedy Send message Joined: 26 Jun 04 Posts: 1643 Credit: 12,921,799 RAC: 89 |
I think we are about to go from a Shorty Storm to a VLAR Storm. All I'm getting for CPU tasks are VLARs. A VLAR Storm is Bad for GPUs because VLARs are not sent to GPUs which means if the Feeder is Full of VLARs when you ask for GPU tasks...you get Nothing. All you get is a confusing reply about not having any work when you know there is work. Well, there is work, it just can't be sent to your GPU. I know that the GPU cache gets filled first so maybe you cannot get CPU work with both the GPU and CPU options selected. Best thing to do if you want to get CPU work is to deselect the GPU option in your preferences. I just looked at your atom computer and I see you currently 53 tasks in progress. |
Grant (SSSF) Send message Joined: 19 Aug 99 Posts: 13746 Credit: 208,696,464 RAC: 304 |
I think we are about to go from a Shorty Storm to a VLAR Storm. All I'm getting for CPU tasks are VLARs. A VLAR Storm is Bad for GPUs because VLARs are not sent to GPUs which means if the Feeder is Full of VLARs when you ask for GPU tasks...you get Nothing. All you get is a confusing reply about not having any work when you know there is work. Well, there is work, it just can't be sent to your GPU. No. Grant Darwin NT |
rob smith Send message Joined: 7 Mar 03 Posts: 22220 Credit: 416,307,556 RAC: 380 |
First, there are no "GPU tasks", all tasks are essentially the same. A task is dedicated to either a CPU or GPU at the time of distribution. However there is a class of task that is never sent to a GPU because it is known to cause Nvidia GPUs problems. When the feed channel is empty you will get no tasks, normally this will be a rare event as it refills every few seconds with another 100 Work Units (200 tasks). Sometimes you can be unlucky and have to wait out a few attempts to get new work, but generally the next call will be a hit. Bob Smith Member of Seti PIPPS (Pluto is a Planet Protest Society) Somewhere in the (un)known Universe? |
Claggy Send message Joined: 5 Jul 99 Posts: 4654 Credit: 47,537,079 RAC: 4 |
When the feed channel is empty you will get no tasks, normally this will be a rare event as it refills every few seconds with another 100 Work Units (200 tasks). I think your statement saying that the feeder holds 100 Wu's (200 tasks) is wrong, In the past my hosts have been allocated well over 100 tasks at a time, If there were only 200 tasks being held by the feeder then I would be receiving replications of the same Wu' ie. both the _0 and the _1, so there must be some thing like 200 Wu's available, and you can't assume each Wu has two tasks available to send, the feeder will also send resends out, ie _3, _4, _5 etc, there is likely to be only one of those being needed to be resent at a time. Claggy |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14653 Credit: 200,643,578 RAC: 874 |
The feeder will only have knowledge of individual tasks (pulled from the 'result' table in the database). It won't have a list of WUs, and I don't know whether the current number is 100, 200, or something else. Once a provisional task has been selected for allocation, it will be tested for feasibility - that will include the VLAR/GPU test, whether any replications have been allocated to the same host or user, whether the deadline can be met, etc. etc. |
Wild6-NJ Send message Joined: 4 Aug 99 Posts: 43 Credit: 100,336,791 RAC: 140 |
Uh oh, the replica db just went offline again. |
Wild6-NJ Send message Joined: 4 Aug 99 Posts: 43 Credit: 100,336,791 RAC: 140 |
Uh oh, the replica db just went offline again. But the data driven pages are still current. I guess plan B is in effect. |
Cactus Bob Send message Joined: 19 May 99 Posts: 209 Credit: 10,924,287 RAC: 29 |
Replica seconds behind master 17,088 As of 31 May 2015, 16:45:00 UTC (estimated as I didn't look at this) Replica seconds behind master 16,354 As of 31 May 2015, 17:00:03 UTC Looks like the replica DB is back online and catching up. not sure what plan B is ------------------- Bob Sometimes I wonder, what happened to all the people I gave directions to? ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ SETI@home classic workunits 4,321 SETI@home classic CPU time 22,169 hours |
kittyman Send message Joined: 9 Jul 00 Posts: 51468 Credit: 1,018,363,574 RAC: 1,004 |
Well, sumpthin's having a hard time of it. Task pages are once again taking forever to load and the kitties are slowly losing caches. "Freedom is just Chaos, with better lighting." Alan Dean Foster |
Akio Send message Joined: 18 May 11 Posts: 375 Credit: 32,129,242 RAC: 0 |
Well, sumpthin's having a hard time of it. +1 |
kittyman Send message Joined: 9 Jul 00 Posts: 51468 Credit: 1,018,363,574 RAC: 1,004 |
Yup, this is bad. The splitters seems to be pumping air now... Actually, over the last half an hour or so, things seem to have righted themselves. Task pages are responding, and have recovered almost 400 tasks to the caches. Might have been the daily task dump hanging things up, perhaps. "Freedom is just Chaos, with better lighting." Alan Dean Foster |
Grant (SSSF) Send message Joined: 19 Aug 99 Posts: 13746 Credit: 208,696,464 RAC: 304 |
MB splitters still having issues. Grant Darwin NT |
S@NL Etienne Dokkum Send message Joined: 11 Jun 99 Posts: 212 Credit: 43,822,095 RAC: 0 |
only 48 channels left... we're about to run out of work all together PANIC! |
Brent Norman Send message Joined: 1 Dec 99 Posts: 2786 Credit: 685,657,289 RAC: 835 |
I'm not panicking, I'm anticipating that AP files may follow :D |
kittyman Send message Joined: 9 Jul 00 Posts: 51468 Credit: 1,018,363,574 RAC: 1,004 |
Just sent word to Eric and Matt about the splitter queues running out. It's only 7:00AM in Berkeley. Hope one of them gets the message and can drum up some more work to load later this morning. "Freedom is just Chaos, with better lighting." Alan Dean Foster |
S@NL Etienne Dokkum Send message Joined: 11 Jun 99 Posts: 212 Credit: 43,822,095 RAC: 0 |
It's OK. The worst thing that can happen, is that we run out of work. I think I can survive that too :-) you're still grinding through trees with your teeth for the lack of AP's ;-) |
kittyman Send message Joined: 9 Jul 00 Posts: 51468 Credit: 1,018,363,574 RAC: 1,004 |
Just got a reply from Matt..... He said thanks, and more work is coming down the pike!!! I did not inquire as to whether any of it might be worthy of AP splitting. I am guessing not, due to what has been split for a while now, but we'll just have to wait and see. The main thing is that at least MB should pick back up again, as RTS just dried up. "Freedom is just Chaos, with better lighting." Alan Dean Foster |
Ulrich Metzner Send message Joined: 3 Jul 02 Posts: 1256 Credit: 13,565,513 RAC: 13 |
APs are splat, but as usual... 01/06/2015 18:59:36 | SETI@home | Sending scheduler request: To fetch work. 01/06/2015 18:59:36 | SETI@home | Requesting new tasks for CPU and NVIDIA GPU 01/06/2015 18:59:39 | SETI@home | Scheduler request completed: got 0 new tasks 01/06/2015 18:59:39 | SETI@home | Project has no tasks available...no APs for Uli. :/ Aloha, Uli |
kittyman Send message Joined: 9 Jul 00 Posts: 51468 Credit: 1,018,363,574 RAC: 1,004 |
APs are splat, but as usual... As I pretty much expected... The first new dataset has already been processed for AP with nothing coming out. "Freedom is just Chaos, with better lighting." Alan Dean Foster |
Brent Norman Send message Joined: 1 Dec 99 Posts: 2786 Credit: 685,657,289 RAC: 835 |
Well is official, I'm 100% dry now :( |
©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.