Message boards :
Number crunching :
Panic Mode On (18) Server problems
Message board moderation
Previous · 1 . . . 7 · 8 · 9 · 10 · 11 · Next
Author | Message |
---|---|
Grant (SSSF) Send message Joined: 19 Aug 99 Posts: 13765 Credit: 208,696,464 RAC: 304 |
How did it clear magically like that? Enough work had been returned that the servers weren't being hammered by clients trying to return work. Also the download traffic has almost dropped to nothing, so that frees up some bandwidth & server & scheduler resources as well. Grant Darwin NT |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14655 Credit: 200,643,578 RAC: 874 |
When a host's upload queue drops below 2*nCPUs, it's away - there's nothing to stop it. Once the ninth-but-last, or seventeenth-but-last, or whatever, upload goes through, new work requests start up with a vengeance, and it can download hundreds in the next few minutes. That could drain any ready-so-send queue very quickly indeed: and with Matt's recent change to a temporary table for ready-to-send, instead of a free-running query from each splitter, it could take a while for new work generation to pick up. |
Vistro Send message Joined: 6 Aug 08 Posts: 233 Credit: 316,549 RAC: 0 |
So there is no longer an upload crisis? |
Fred W Send message Joined: 13 Jun 99 Posts: 2524 Credit: 11,954,210 RAC: 0 |
So there is no longer an upload crisis? Correct. (Until the next one...) Note that if the downloads remain borked for a significant length of time, then when they come back the resulting pressure on bandwidth could cause another upload crisis; and so it goes on... F. |
Grant (SSSF) Send message Joined: 19 Aug 99 Posts: 13765 Credit: 208,696,464 RAC: 304 |
So there is no longer an upload crisis? Not at the moment, all mine are uploading as they finish. It'll probably be a few hours though before all the outstanding work has been uploaded. Grant Darwin NT |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14655 Credit: 200,643,578 RAC: 874 |
So there is no longer an upload crisis? There was never a crisis. There was a bit of a squash, like fans leaving a football ground at the end of a match, but game's over now and the freeway is flowing again. As Ned keeps (very properly) reminding us, BOINC was designed to cope with these ups and downs. It quietly goes about its business, backing off at the first sign of trouble, and biding its time with saintly patience. It's only us humans that use words like 'panic' and 'crisis'. |
Vistro Send message Joined: 6 Aug 08 Posts: 233 Credit: 316,549 RAC: 0 |
So I can tell my friend (who somehow got 109 tasks in under 10 hours) to run an update command now? (Update also forces the uploads, right? |
Grant (SSSF) Send message Joined: 19 Aug 99 Posts: 13765 Credit: 208,696,464 RAC: 304 |
Just tell them to make sure network access is enabled & then just let it do it's thing. They'll all upload the next time they try. Grant Darwin NT |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14655 Credit: 200,643,578 RAC: 874 |
So I can tell my friend (who somehow got 109 tasks in under 10 hours) to run an update command now? Wrong, I'm afraid. When the job is finished, two things happen, in this order: First, the data has to be uploaded. That's a straight file transfer. Once the data resides on the server's disk (following a successful transfer), the job can be reported. You can't force anything. You can instigate a 'please try again' by visiting the Transfers tab (in advanced view), selecting a pending upload, and clicking the 'Retry now' button. If that seems to work consistently, you can re-submit the entire queue by choosing "Retry communications" from the Advanced drop-down menu. Once all the uploads have completed, then you can use the Update button to instigate the second, reporting stage. But you may find that BOINC relieves you of this minor chore by requesting new work, and reporting what it's done already while it's at it. |
Vistro Send message Joined: 6 Aug 08 Posts: 233 Credit: 316,549 RAC: 0 |
K I found it strange that it reported so many tasks then just stopped, I assumed that all of the tasks it had were stuck in the random back offs. Unfortunatley, I can't seem to get into his client through the select computer command. |
1mp0£173 Send message Joined: 3 Apr 99 Posts: 8423 Credit: 356,897 RAC: 0 |
So it's all over now? No more failed uploads? Because there is a huge difference between running at 95% capacity, and running at 105% capacity. The moment we dipped under about 97% we went from most uploads failing to most uploads working, and then the need went away. |
Bernie Vine Send message Joined: 26 May 99 Posts: 9954 Credit: 103,452,613 RAC: 328 |
So there is no longer an upload crisis? Well you might say they wasn't a "crisis" but I had 4 crunchers that I wanted to reduce to 2 due to the heat in my flat, but they couldn't upload for DAYS and DAYS, I didn't want to abort for the sake of the wingman as it isn't their fault. I have finally decide to stop ALL seti@home as I cannot control it. The current temperature in my flat is 33c (91F) at 0:48am!! with only 2 machines running. No, seti@home is out of user control, and whilst I am crunching WCG, I will finish this week end and my 2 redundant crunchers are destined for the council recycle bin. Bernie PS If anyone is interested they are BAV-3 and BAV-4 in my list |
Vistro Send message Joined: 6 Aug 08 Posts: 233 Credit: 316,549 RAC: 0 |
ILL TAKE THOSE CRUNCHERS OFF OF YOUR HANDS!!!! (seriously, we need to find a way to reroute that garbage truck to Illinois) |
Bernie Vine Send message Joined: 26 May 99 Posts: 9954 Credit: 103,452,613 RAC: 328 |
ILL TAKE THOSE CRUNCHERS OFF OF YOUR HANDS!!!! LOL don't think dustcarts (what we call em in the UK) do transatlantic crossings really well!! Bernie |
Vistro Send message Joined: 6 Aug 08 Posts: 233 Credit: 316,549 RAC: 0 |
FedEx work for you? |
Alinator Send message Joined: 19 Apr 05 Posts: 4178 Credit: 4,647,982 RAC: 0 |
So there is no longer an upload crisis? LOL... I don't need anymore old timers, got enough already! ;-) Just an FYI: Only you can see the host names you have for your machines. All the rest of us can only see their HID. Alinator |
Grant (SSSF) Send message Joined: 19 Aug 99 Posts: 13765 Credit: 208,696,464 RAC: 304 |
Hmm. Something's borked. Assimilator queue is draining, uploads are tapering off. But the splitters still haven't kicked in. Grant Darwin NT |
Vistro Send message Joined: 6 Aug 08 Posts: 233 Credit: 316,549 RAC: 0 |
Assimilator... what exactly is getting absorbed? |
Grant (SSSF) Send message Joined: 19 Aug 99 Posts: 13765 Credit: 208,696,464 RAC: 304 |
Assimilator... what exactly is getting absorbed? Results that have been validated get moved in to the Science database. Grant Darwin NT |
Vistro Send message Joined: 6 Aug 08 Posts: 233 Credit: 316,549 RAC: 0 |
Yeah... that table is a little confusing... I just want to know the following, really, at any given time: How many people are waiting to upload How many tasks are ready to be shipped out right now How many are in the feeder ready to be put into the scheduler What day are we at |
©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.