Panic Mode On (78) Server Problems?

No, I don't. Like I said, it happened on the server side! My host didn't even succeed in making a scheduler request. This is the second time it happened. I'm not the only one this happened to; there were other users reporting the same thing.

No, I don't. Like I said, it happened on the server side! My host didn't even succeed in making a scheduler request. This is the second time it happened. I'm not the only one this happened to; there were other users reporting the same thing.

What version of BOINC?
Which OS & version?

EDIT- ie, were they the same as yours or different?

The last few days I have been getting a few "error while downloading"
Any ideas as to what is causing this?

Frank

Frank, I've been having a lot of this too and all my desktops with Nvidia GTX 550 ti graphics processors have been slowly running out of work. I have used the No new tasks option and then update followed by the allow new tasks and update a few minutes later after tasks upload routine again and again over the last week or so with gradually decreasing success. As of about 6 a.m. this morning I was out of all work on both desktops. I've switched over to my old standby's gpugrid, world community grid, and a few other cosmology projects. Some of us believe the problem is with the scheduler not being able to keep track of tasks completed and sending out way to many tasks. Many of us have many dozens and many hundreds of ghost processes in the system. There also seems to be an association between Astro Pulse work being split and sent out which may be fouling the rest of the scheduling work. The internet bandwidth of the 100 meg line from the lab to BOINC seems to be far beyond capacity. Some of us have noticed this big problem of getting work and reporting work after a maintenance downtime about 3 or 4 weeks ago. I noticed that the system seemed to come out of that maintenance fine and all my computers were happily send working out and getting new work without time outs or failures. My recollection is that it (Seti at Home) stopped running well after just a few hours.

My notebooks, even one with an Nvidia 525m graphics processor and an i7 2670 qm core processor is still getting some work and reporting out, but having dozens and dozens of reporting failures and time outs every day. My i3 notebook with intel integrated graphics is doing fine. My little core duo notebook whose Radeon 2600 series processor doesn't qualify to run jobs is working fine too. As I wrote earlier, I am gradually switching over to alternative projects with my desktops beginning today.

I have never seen it this bad in my year here and am rethinking my priorities. Right now I am thinking a new project mix will be 2 parts medical discovery and disease fighting projects along with some Seti at Home work with my notebooks. I have fought this kind of chronic frustration at work before and it is too stressful for many people to handle well. The low limits on work per cpu and graphics processor are already hitting my desktops even though they were both down to just a few dozen short work units each. The limits will not help with the problem at all according to what I have read here. From my point of view, the download/upload issue became much more severe after that maintenance downtime around 4 weeks ago. I remember it all happened not too long after my wife and I got back from a memorial service for a close family member around mid October. We had just returned from visiting our families for an extended period and were building RAC up again slowly. Momentum stopped. Sorry, I don't have enough data to track it back to an exact date. I hope the Seti folks address this very, very soon. They are way understaffed, but I know they have the project's interests at heart. I know too that there are times when a project just has to step back and solve serious issues that may negatively affect project morale if left without at least a partial or interim solution. Brother Frank on Seti at Home.

probably you have a nvidia graphic card and the error is on cuda_fermi packages. You must download this: https://developer.nvidia.com/cuda-downloads

fscheel wrote:

The last few days I have been getting a few "error while downloading"
Any ideas as to what is causing this?

Frank

Example:
http://setiathome.berkeley.edu/result.php?resultid=2715242784

<core_client_version>7.0.28</core_client_version> <![CDATA[ <message> WU download error: couldn't get input files: <file_xfer_error> <file_name>01se12ac.21111.2934.140733193388046.10.203</file_name> <error_code>-200</error_code> </file_xfer_error> </message> ]]>

This could be a problem with the S@h server.

For a few weeks we had the same problem at S@h-Beta.

Maybe an admin should look to this problem.


* Best regards! :-) * Sutaru Tsureku, team seti.international founder. * Optimize your PC for higher RAC. * SETI@home needs your help. *

No, I don't. Like I said, it happened on the server side! My host didn't even succeed in making a scheduler request. This is the second time it happened. I'm not the only one this happened to; there were other users reporting the same thing.

What version of BOINC?
Which OS & version?

EDIT- ie, were they the same as yours or different?

Linux/x64, BOINC 7.0.39
Others have had it happen on Windows (don't know about BOINC version, most definitely not 7.0.39).
I don't think it's client's fault but server's - I assume it receives a malformed request (due to networking problems) and thinks project was reset.
Client wasn't even notified about that and continued to crunch already abandoned tasks so I had to manually abort them (ok, I accidentaly aborted a few more tasks that weren't "abandoned"). I had to use proxy for that otherwise that computer on that ISP rarely manages to contact scheduler without timing out.

Are you using a proxy- guess what? I just got my first abandoned tasks, 200 of them.
Only occurred since using a proxy.

Totally empty que. Have gotten only some 20 WU today. They were consumed in a matter of few tens of minutes. Is there no work out there?

Hey this is like Unemployment for the clients. No work.... We want more work....
____________

I have been running just fine since I fixed my networking, have as much work as the project will send me. And reporting is going though just about always on the first shot. Downloads have still been a little slow but moving along.

Now I just need to replace one of my nvidia cards that is no longer working right on my old crunch box.


____________

Seti@Home 11/15/2012 8:15:00PM work fetch resumed by user
Seti@Home 11/15/2012 8:15:01PM update requested by user
Seti@Home 11/15/2012 8:15:02PM Sending scheduler request. Requested by user.
Seti@Home 11/15/2012 8:15:02PM Requesting new tasks for CPU.
Seti@Home 11/15/2012 8:21:25PM Scheduler request failed. Timeout was reached
Seti@Home 11/15/2012 8:21:29PM Project communication failed. attempting access to reference site
Seti@Home 11/15/2012 8:21:31PM Internet access ok - Project servers may be temporarily down

This is what I keep getting. Any word as to when this all may be rectified??

Seti@Home 11/15/2012 8:15:00PM work fetch resumed by user
Seti@Home 11/15/2012 8:15:01PM update requested by user
Seti@Home 11/15/2012 8:15:02PM Sending scheduler request. Requested by user.
Seti@Home 11/15/2012 8:15:02PM Requesting new tasks for CPU.
Seti@Home 11/15/2012 8:21:25PM Scheduler request failed. Timeout was reached
Seti@Home 11/15/2012 8:21:29PM Project communication failed. attempting access to reference site
Seti@Home 11/15/2012 8:21:31PM Internet access ok - Project servers may be temporarily down

This is what I keep getting. Any word as to when this all may be rectified??

You should try using a proxy. It seems that the scheduller is not having issues doing its work, but it fails to comunicate with the clients when its contacted directly. When a proxy is used the comunications fail a lot less... Not a confirmed theory, but everyone using a proxy is now getting work regularly.

You should try using a proxy. It seems that the scheduller is not having issues doing its work, but it fails to comunicate with the clients when its contacted directly. When a proxy is used the comunications fail a lot less... Not a confirmed theory, but everyone using a proxy is now getting work regularly.

Ok. So where is the list of proxies? And how to configure it in Boinc?

Not that astute to initiate a proxy connection without a little assistance. I follow directions well. Anyone offer assistance so I can help us meet ET a little sooner?