Message boards :
Number crunching :
Problems...
Message board moderation
Previous · 1 · 2 · 3 · 4 · 5 · 6 · 7 · 8 . . . 13 · Next
Author | Message |
---|---|
kittyman Send message Joined: 9 Jul 00 Posts: 51478 Credit: 1,018,363,574 RAC: 1,004 |
What cooling systems have the server? Perhaps even more cooling systems use? One of the many overhead services, light the power to run them. Supplied by Berkeley, out of 'their cut' of donations. Not a small thing......I know what it costs to run AC 24/7. "Time is simply the mechanism that keeps everything from happening all at once." |
Raistmer Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121 |
05/03/2010 20:59:48 SETI@home Reporting 3 completed tasks, requesting new tasks for GPU 05/03/2010 20:59:50 Project communication failed: attempting access to reference site 05/03/2010 20:59:52 Internet access OK - project servers may be temporarily down. 05/03/2010 20:59:53 SETI@home Scheduler request failed: Server returned nothing (no headers, no data) Does SETI's inet connection still overloaded to lead to such server reply or something another wrong here? |
Fred W Send message Joined: 13 Jun 99 Posts: 2524 Credit: 11,954,210 RAC: 0 |
05/03/2010 20:59:48 SETI@home Reporting 3 completed tasks, requesting new tasks for GPU This doesn't line up with traffic overload on the Cricket graphs. More likely problems accessing the database I would guess. F. |
Julie Send message Joined: 28 Oct 09 Posts: 34060 Credit: 18,883,157 RAC: 18 |
|
dnolan Send message Joined: 30 Aug 01 Posts: 1228 Credit: 47,779,411 RAC: 32 |
05/03/2010 20:59:48 SETI@home Reporting 3 completed tasks, requesting new tasks for GPU Yeah, I think something else is going on. Last night I watched my Cuda machine request work and get 4 tasks ten times in a row, 4 tasks sent each time, each request about a minute after the last, then it got the above message about 4 times in a row, then got some more work and laid off the requests... The traffic at the time didn't look that heavy to me. -Dave |
Pappa Send message Joined: 9 Jan 00 Posts: 2562 Credit: 12,301,681 RAC: 0 |
anakin The scheduling Server has an Apache Web Server that recieves your file "sched_request_setiathome.berkeley.edu.xml" Those vary in size from about 1K to several hundred K. Your Boinc Client fires up a connection to anakin "http://setiboinc.ssl.berkeley.edu/sah_cgi/cgi" which in its config has a Max number of allowed connections. If it transfering requests from several other computers "Max Connections" You are not going to get a connection. You of course know what is the sched_request file: It is your account information. What Application and version that are on that particular machine. A copy of the Preferences (and venues that you have configured). The contents of what was in the "stderr.txt" for any completed workunits. A list results in progress. A list of workunits to be ran (including Deadlines and estimates of time to complete). So when the Scheduler recieves this from you, It then gets to decide Are you reporting completed work? If Yes fire up information to the Validator (transitioners also on anakin). Is Your Cache Full or do you need work (count up all those times and see how many seconds of work you require). If You need work, check the Feeder (anakin) for Results Ready to send and assign them to that computer. Reply with the "sched_reply_setiathome.berkeley.edu.xml" which defines where to go get your assigned Workunits. It can also tell your Boinc Client to remove outdated files or Workunits that have been canceled. So it is a "busy" machine! It has a LOT to Decide. Regards Please consider a Donation to the Seti Project. |
JohnDK Send message Joined: 28 May 00 Posts: 1222 Credit: 451,243,443 RAC: 1,127 |
anakin Problem just is that it happened after the big outage, before there's wasn't any major problem. |
Highlander Send message Joined: 5 Oct 99 Posts: 167 Credit: 37,987,668 RAC: 16 |
anakin That can i fully sign. Something on 20 Feb 2010 around 18:28 UTC has definitely changed in the behaviour of the scheduler. Is it possible to trace back the changes round this time (last changes before this time)? (Meanwhile i have upgraded from 6.10.18 to 6.10.36 -> no change in this specific point) - Performance is not a simple linear function of the number of CPUs you throw at the problem. - |
Highlander Send message Joined: 5 Oct 99 Posts: 167 Credit: 37,987,668 RAC: 16 |
Seems, you have found it!? Last time i got the (no headers, no data) failure, was on 10:46 UTC, since then, no update-request had this msg. Can others confirm? - Performance is not a simple linear function of the number of CPUs you throw at the problem. - |
W-K 666 Send message Joined: 18 May 99 Posts: 19402 Credit: 40,757,560 RAC: 67 |
Seems, you have found it!? Last time i got the (no headers, no data) failure, was on 10:46 UTC, since then, no update-request had this msg. Still getting that "no headers, no data" msg at 17:33 UTC. |
perryjay Send message Joined: 20 Aug 02 Posts: 3377 Credit: 20,676,751 RAC: 0 |
Also updated to 6.10.36. Still getting the message now and then. PROUD MEMBER OF Team Starfire World BOINC |
zoom3+1=4 Send message Joined: 30 Nov 03 Posts: 66349 Credit: 55,293,173 RAC: 49 |
Also updated to 6.10.36. Still getting the message now and then. I also get the same message and I'm using 6.10.36 too. 3/6/2010 10:41:04 AM SETI@home Scheduler request failed: Server returned nothing (no headers, no data) I pretty much don't much attention to It though. Savoir-Faire is everywhere! The T1 Trust, T1 Class 4-4-4-4 #5550, America's First HST |
JohnDK Send message Joined: 28 May 00 Posts: 1222 Credit: 451,243,443 RAC: 1,127 |
It often gets my attention since I often have to do manual updates, Boinc often wont do it by itself. 06/03/2010 20:10:34 SETI@home update requested by user |
Dave Send message Joined: 29 Mar 02 Posts: 778 Credit: 25,001,396 RAC: 0 |
I pretty much don't much attention to It though. Same here. All working so far. |
perryjay Send message Joined: 20 Aug 02 Posts: 3377 Credit: 20,676,751 RAC: 0 |
I've found that if I just let it go it will go through when it wants to. PROUD MEMBER OF Team Starfire World BOINC |
JohnDK Send message Joined: 28 May 00 Posts: 1222 Credit: 451,243,443 RAC: 1,127 |
*Cut* |
Raistmer Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121 |
Still needed to ask twice... 07/03/2010 03:28:46 SETI@home update requested by user 07/03/2010 03:28:50 SETI@home Sending scheduler request: Requested by user. 07/03/2010 03:28:50 SETI@home Reporting 20 completed tasks, not requesting new tasks 07/03/2010 03:28:52 Project communication failed: attempting access to reference site 07/03/2010 03:28:54 Internet access OK - project servers may be temporarily down. 07/03/2010 03:28:55 SETI@home Scheduler request failed: Server returned nothing (no headers, no data) 07/03/2010 03:28:56 SETI@home update requested by user 07/03/2010 03:29:00 SETI@home Sending scheduler request: Requested by user. 07/03/2010 03:29:00 SETI@home Reporting 20 completed tasks, not requesting new tasks 07/03/2010 03:29:10 SETI@home Scheduler request completed |
Pappa Send message Joined: 9 Jan 00 Posts: 2562 Credit: 12,301,681 RAC: 0 |
I have been in and out of my house and on the road for the last 6 plus days. Other than the other evening when I had the chance to install 6.10.36 on my various machines and spend a few hours "Alpha" testing. Boinc and Seti have been running in AUTO! I have not seen a large backlog of "results" ready to report. As I just sat down again my Seti Machines are happy... Other projects have a few issues that I need to look at. That said, the message is a "GENERIC Message reported by the BOINC Client that there WAS an ISSUE contacting the Scheduler." It does Not report what actually happened during the process of contact/acknowledgement and handling of Your "sched_request_setiathome.berkeley.edu.xml" A user with a 10 Cache has a "sched_request_setiathome.berkeley.edu.xml" that is probably somewhere close to or above a Half a Megabyte. That will tie up the scheduler for quite a FEW Seconds while the file is being UPLOADED (with everything else going on {Uploads/Downloads/Other Scheduler attempts}). For the record I have not asked how many active scheduler sockets (listeners) are available. I do suspect that it is fairly large to the extent that an actual connection "could time out," then you would also get the same message from the Boinc Client "SETI@home Scheduler request failed: Server returned nothing (no headers, no data)" Once again, if "anakin" is processing even 10 "power users" with a 10 day cache (over several thousands workunits and results) and communicating with the transitioner and the feeder queue to send work. It is going to be working as fast as it can matching things. If it drops a connection not really a big deal. Users with older Boinc Clients are probably going to have the Libecurl issues where you "might" need to restart the Boinc client. Lastly and I continue to read this, "if you push the button and it succeeds." Then I would suspect the Boinc Client more than I would the server. The only thing that happened is the User, pushed a Button on the Boinc client. The Server has not changed (other than handling requests it gets). Yes, sometimes I think I really have been doing this Too Long. Regards Please consider a Donation to the Seti Project. |
W-K 666 Send message Joined: 18 May 99 Posts: 19402 Credit: 40,757,560 RAC: 67 |
Al, I think it has been observered and reported because in recent weeks it has been a common occurance. Since the last outage until about 18:00 UTC yesterday it was the response to at least 60% of my attempts to contact Berkeley on the two computers I control. At times, for periods of several hours, usually when the USA is awake, it was the only response. Looking through my logs the most successful time for me to connect was between 05:00 UTC ( midnight NY) until 10:00 UTC. Andy |
Pappa Send message Joined: 9 Jan 00 Posts: 2562 Credit: 12,301,681 RAC: 0 |
Thank You Andy After looking at my logs I see a couple connection attempts that failed but no reason as to why (2:00am PST on the 6th of Mar). With the current "shorty storm" this machine has been asking for work about every ten minutes for the last 18+ hours (successfully). My current "sched_request_setiathome.berkeley.edu.xml" is at 170K at the moment (492 WU's or 5000 lines in the file) . As it will make my logs a bit chatty, I have enabled <http_debug>1</http_debug> in my cc_config.xml file. That will give the best indication of what might be happening. If someonce can cut and paste the failed scheduler conversation that will help. Regards Please consider a Donation to the Seti Project. |
©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.