 |
|
Message boards :
Number crunching :
Problem uploading final results
Message board moderation
| Author | Message |
|---|---|
|
Send message Joined: 26 Aug 04 Posts: 14 Credit: 123,062 RAC: 0 |
I've been having problems uploading the zip result files for the past 24 hours. One of the files has uploaded, but the other 4 keep failing. Is there a problem with the uploader here is where it is pointed to upload: http://phkup21.unibe.ch/boinc/file_upload_handler Please advise. Thanks for your time |
|
Send message Joined: 5 Sep 04 Posts: 7629 Credit: 24,240,330 RAC: 0 |
Hi Jason, 'unibe' is University of Bern. 'ch' is Switzerland. So they may be having a problem, which the Oxford people may not know about yet. Les |
|
Send message Joined: 26 Aug 04 Posts: 14 Credit: 123,062 RAC: 0 |
More information: climateprediction.net - 2005-03-14 13:31:48 - Started upload of 2vpt_100155773_1_3.zip climateprediction.net - 2005-03-14 13:31:53 - Temporarily failed upload of 2vpt_100155773_1_3.zip climateprediction.net - 2005-03-14 13:31:53 - Backing off 2 hours, 44 minutes, and 45 seconds on transfer of file 2vpt_100155773_1_3.zip climateprediction.net - 2005-03-14 13:33:04 - Started upload of 2vpt_100155773_1_2.zip climateprediction.net - 2005-03-14 13:33:09 - Temporarily failed upload of 2vpt_100155773_1_2.zip climateprediction.net - 2005-03-14 13:33:09 - Backing off 3 hours, 3 minutes, and 30 seconds on transfer of file 2vpt_100155773_1_2.zip They seem to kick out after 5 seconds, please advise. Thanks |
|
Send message Joined: 5 Sep 04 Posts: 7629 Credit: 24,240,330 RAC: 0 |
Sorry, should have added: Don't worry about it. BOINC is designed to continue attempting to upload. Last December, there was a problem with the Bern server which kept it off air for about 2 weeks. (First a faulty power supply; then it wouldn't reboot; then a security issue was discovered.) But eventually it was back up, and the large number of people with waiting uploads were taken care of, myself included. I would suggest though, that you tick 'Disable BOINC network access' for a while, just trying once a day perhaps. This will prevent so many messages, and some worry. Les |
|
Send message Joined: 26 Aug 04 Posts: 14 Credit: 123,062 RAC: 0 |
Does anybody know if there is an issue with uploader currently? http://phkup21.unibe.ch/boinc/file_upload_handler |
 Thyme LawnSend message Joined: 5 Aug 04 Posts: 1283 Credit: 15,824,334 RAC: 0 |
> Does anybody know if there is an issue with uploader currently? > http://phkup21.unibe.ch/boinc/file_upload_handler None of the models I've uploaded this week (or am due to upload in the next couple of days) used that server, but traceroute definitely reaches it. Can you use a packet sniffer (e.g. <a href="http://www.ethereal.com/">Ethereal</a>) to trap the traffic for an upload attempt? "The ultimate test of a moral society is the kind of world that it leaves to its children." - Dietrich Bonhoeffer |
|
Send message Joined: 26 Aug 04 Posts: 14 Credit: 123,062 RAC: 0 |
> None of the models I've uploaded this week (or am due to upload in the next > couple of days) used that server, but traceroute definitely reaches it. Can > you use a packet sniffer (e.g. <a> href="http://www.ethereal.com/">Ethereal</a>) to trap the traffic for an > upload attempt? But how do you explain how one of the 5 zip files did upload while the others have continued to fail? |
|
Thyme Lawn Send message Joined: 5 Aug 04 Posts: 1283 Credit: 15,824,334 RAC: 0 |
> But how do you explain how one of the 5 zip files did upload while the others > have continued to fail? I can't Jason :( There are various reasons why file uploads can fail, and the diagnostics at the client end could definitely be improved to make it easier for participants to self-diagnose problems like this. If none of the stdout/stderr files on your system give an indication why the upload is failing the only way you're likely to find out what's going on is by using a packet sniffer to trap the HTTP messages. The only other possibility would be if one of the project team could check the server log files, but they're short-staffed at the moment and I doubt if they'd be able to spare the time to do it with all the other things that are going on. "The ultimate test of a moral society is the kind of world that it leaves to its children." - Dietrich Bonhoeffer |
|
Send message Joined: 30 Jan 05 Posts: 16 Credit: 65,203 RAC: 0 |
1) Where does the GUI identify where the uploads are being directed to, or where the downloads are coming from? 2) I've finished a project, but the file uploads are slow and erratic and downloads worse for the next project coming -- having to replace 4.04 with 4.10 as well as get new models. 3) the only way I can use the web is to suspsend BOINC file transfers, and with it, the project still 6 hours from completion. I don't mind keeping my CPU occupied, but I frankly cannot allow my web access to be compromised by file transfers. IF they ran cleanly -- fine -- but if a new project ain't up and running and uploads complete by Monday, I'm outta here. Now I'm gonna un-suspend the project and go read a book while upload proceed at about 200 bytes per second on average. |
|
Send message Joined: 5 Sep 04 Posts: 7629 Credit: 24,240,330 RAC: 0 |
1) in the client_state.xml file. 2) there are about 7 megs in 5 zips. Your listed upload rate is faster than mine. There's not much that others can do about your comms speed. A 3D climate model over 45 years is a big project. Perhaps some of the other dc projects would suit your hardware better. Lots of people drop out for one reason or another. Les |
|
Honza Send message Joined: 5 Aug 04 Posts: 390 Credit: 2,475,242 RAC: 0 |
> 1) Where does the GUI identify where the uploads are being directed to, or > where the downloads are coming from? I would look in client_state.xml in BOINC folder for section upload_when_present - should be followed by url of upload server. Sorry, not sure why your upload/download is slow at the moment... <i>phpBB forum for CPDN, all are </i><a href="http://www.climateprediction.net/board">invited</a> |
|
Send message Joined: 30 Jan 05 Posts: 16 Credit: 65,203 RAC: 0 |
Well -- it only took about 8-9 hours to get results uploaded and new work units started. The new display doesn't seen to change significantly, except the "F"ilter toggle doesn't work -- filter seems always on. Actually, on inspection, it seems more than that -- like some kind of modeling based on cell values, not just a graphical edge-smoothing. The "W" view is quite nice, but doesn't seem to be scalable like the globe view, and it resets the globe view to center on Greenwich Meridian when you toggle back. One problem, which I hope is only temporary, it the the last trickle of my second model to finish, in the midst of the all the file transfer activity, seems to still be unrecorded. So, I can't see yet the Phase III changes. --- Addendum: At the start of new work units, I rebooted right away, since my first models messed up and new ones had to be loaded the first time I rebooted after starting CP... It was some kind of corruption of some of the hadsm3 files. Anyhow -- caluclations started again cleanly -- and right away it communicated to host to send the last trickle and do whatever "ready to report" wants to report. So all is well, and 3 months from now, 2 more work units and hopefully broadband by then. --- Did 4.10 fix the color anomaly on temperatures > 40C? I'm glad that file transfers support resuming. I had about 20 re-connects in the course of finishing, during 2 of which, the bulk of the transfers were made. Did finally get about 2kbps transfer, in brief intervals. Damned dial-up! |
|
geophi Send message Joined: 7 Aug 04 Posts: 2169 Credit: 64,555,907 RAC: 5,858 |
> units started. The new display doesn't seen to change significantly, except > the "F"ilter toggle doesn't work -- filter seems always on. Actually, on > inspection, it seems more than that -- like some kind of modeling based on > cell values, not just a graphical edge-smoothing. You can use the "9" key to go between the contoured and blocky cell displays. |
|
Send message Joined: 26 Aug 04 Posts: 14 Credit: 123,062 RAC: 0 |
> > But how do you explain how one of the 5 zip files did upload while the > others > > have continued to fail? > > I can't Jason :( > > There are various reasons why file uploads can fail, and the diagnostics at > the client end could definitely be improved to make it easier for participants > to self-diagnose problems like this. If none of the stdout/stderr files on > your system give an indication why the upload is failing the only way you're > likely to find out what's going on is by using a packet sniffer to trap the > HTTP messages. The only other possibility would be if one of the project team > could check the server log files, but they're short-staffed at the moment and > I doubt if they'd be able to spare the time to do it with all the other things > that are going on. > Don't have permissions to install this type of software. I still haven't been able to get the remaining 4 zip files to upload. Has anyone else had a problem uploading to http://phkup21.unibe.ch/boinc/file_upload_handler? Could there also be a size limitation on the server side since my _5.zip was only 216K and the other 4 zips are above 1M? I check the maxnbytes and they were at 5000000. Thanks for your time |
|
Thyme Lawn Send message Joined: 5 Aug 04 Posts: 1283 Credit: 15,824,334 RAC: 0 |
> Has anyone else had a problem uploading to http://phkup21.unibe.ch/boinc/file_upload_handler? I've just successfully uploaded to that server Jason. > Could there also be a size limitation on the server side since my _5.zip was > only 216K and the other 4 zips are above 1M? I check the maxnbytes and they > were at 5000000. It's not going to be file size (you'd get a -131 error if it was). One thought. Your post suggests that the '_5' is the one that was uploaded. Does your firewall limit the size of file you're allowed to upload? "The ultimate test of a moral society is the kind of world that it leaves to its children." - Dietrich Bonhoeffer |
|
Send message Joined: 23 Sep 04 Posts: 7 Credit: 341,690 RAC: 0 |
I did start colaborating with this proyect right from the beginning. Two finished models ago it took 2 days to upload the results and download a new model. The last model took about a month (connected 24h. a day) for the same task. My actual model did finish (100.00%) 5 weeks ago and still shows the same message "climateprediction.net - 2005-03-22 11:15:19 - Deferring communication with project for ..." In the last week I have tried with and without the firewall, stopping and starting the program, the computer... Taking longer (will it success this time?)to comunicate than calculating the model it is not worthy to continue. So if there's no solution for this problem, I'm giving up. |
|
Thyme Lawn Send message Joined: 5 Aug 04 Posts: 1283 Credit: 15,824,334 RAC: 0 |
> I did start colaborating with this proyect right from the beginning. Two > finished models ago it took 2 days to upload the results and download a new > model. The last model took about a month (connected 24h. a day) for the same > task. My actual model did finish (100.00%) 5 weeks ago and still shows the > same message "climateprediction.net - 2005-03-22 11:15:19 - Deferring > communication with project for ..." > In the last week I have tried with and without the firewall, stopping and > starting the program, the computer... Taking longer (will it success this > time?)to comunicate than calculating the model it is not worthy to continue. > So if there's no solution for this problem, I'm giving up. Hi Jose, Your problem isn't the same as Jason's. From your stats page it looks like you're trying to upload all the trickles from the model at the same time (38 got through to the server on Feb 25 and another one yesterday). Unfortunately there's currently a problem with uploading lots of trickles in one go. The symptoms of the problem are: 1) Every CPDN scheduler request from the host gets a <b>No schedulers responded</b> message. 2) You'll have lots of <b>trickle_up_*.xml</b> files (in your case over 39) in your <b>projects/climateprediction.net</b> directory. The work around is to create a directory and move all but 30 of the trickle_up_*.xml files into that directory, do an update (which should succeed and remove the files you left behind), then move the other files back in batches of 30 followed by an update each time. As you're trying to download another WU at the same time as uploading the completed one the scheduler is allocating a new one (but never sending it to you) each time you contact it. There are currently at least 30 WUs stuck in limbo because of this. You might also want to consider merging your 66 comupters into a single one! "The ultimate test of a moral society is the kind of world that it leaves to its children." - Dietrich Bonhoeffer |
|
Send message Joined: 7 Aug 04 Posts: 187 Credit: 44,163 RAC: 0 |
> You might also want to consider merging your 66 > comupters into a single one! There are currently 121 computers! :o Whatcha doing Jose? |
|
Send message Joined: 23 Sep 04 Posts: 7 Credit: 341,690 RAC: 0 |
> > I did start colaborating with this proyect right from the beginning. Two > > finished models ago it took 2 days to upload the results and download a > new > > model. The last model took about a month (connected 24h. a day) for the > same > > task. My actual model did finish (100.00%) 5 weeks ago and still shows > the > > same message "climateprediction.net - 2005-03-22 11:15:19 - Deferring > > communication with project for ..." > > In the last week I have tried with and without the firewall, stopping > and > > starting the program, the computer... Taking longer (will it success > this > > time?)to comunicate than calculating the model it is not worthy to > continue. > > So if there's no solution for this problem, I'm giving up. > > Hi Jose, > > Your problem isn't the same as Jason's. From your stats page it looks like > you're trying to upload all the trickles from the model at the same time (38 > got through to the server on Feb 25 and another one yesterday). Unfortunately > there's currently a problem with uploading lots of trickles in one go. > > The symptoms of the problem are: > > 1) Every CPDN scheduler request from the host gets a <b>No schedulers > responded</b> message. > > 2) You'll have lots of <b>trickle_up_*.xml</b> files (in your case over 39) in > your <b>projects/climateprediction.net</b> directory. > > The work around is to create a directory and move all but 30 of the > trickle_up_*.xml files into that directory, do an update (which should succeed > and remove the files you left behind), then move the other files back in > batches of 30 followed by an update each time. > > As you're trying to download another WU at the same time as uploading the > completed one the scheduler is allocating a new one (but never sending it to > you) each time you contact it. There are currently at least 30 WUs stuck in > limbo because of this. You might also want to consider merging your 66 > comupters into a single one! > Sorry folks I've been travelling. 1-I thought I did nothing but leave the program to run based on preferences without changing the defoult ones. 2-I can't find those "trickle_up_*xml</b> files you mention. 3-I only have one computer and I have no idea of what are you taking about when you say that "You might also want to consider merging your 66 > comupters into a single one!" 4-As you have guess I'm a computer user but unfortunately I can't follow your expert computer jargon. So if you know what is wrong with my computer, please tell me step by step how I should mend it. Thanks |
|
Send message Joined: 23 Sep 04 Posts: 7 Credit: 341,690 RAC: 0 |
> > > I did start colaborating with this proyect right from the beginning. > Two > > > finished models ago it took 2 days to upload the results and > download a > > new > > > model. The last model took about a month (connected 24h. a day) for > the > > same > > > task. My actual model did finish (100.00%) 5 weeks ago and still > shows > > the > > > same message "climateprediction.net - 2005-03-22 11:15:19 - > Deferring > > > communication with project for ..." > > > In the last week I have tried with and without the firewall, > stopping > > and > > > starting the program, the computer... Taking longer (will it > success > > this > > > time?)to comunicate than calculating the model it is not worthy to > > continue. > > > So if there's no solution for this problem, I'm giving up. > > > > Hi Jose, > > > > Your problem isn't the same as Jason's. From your stats page it looks > like > > you're trying to upload all the trickles from the model at the same time > (38 > > got through to the server on Feb 25 and another one yesterday). > Unfortunately > > there's currently a problem with uploading lots of trickles in one go. > > > > The symptoms of the problem are: > > > > 1) Every CPDN scheduler request from the host gets a <b>No schedulers > > responded</b> message. > > > > 2) You'll have lots of <b>trickle_up_*.xml</b> files (in your case over > 39) in > > your <b>projects/climateprediction.net</b> directory. > > > > The work around is to create a directory and move all but 30 of the > > trickle_up_*.xml files into that directory, do an update (which should > succeed > > and remove the files you left behind), then move the other files back in > > batches of 30 followed by an update each time. > > > > As you're trying to download another WU at the same time as uploading > the > > completed one the scheduler is allocating a new one (but never sending it > to > > you) each time you contact it. There are currently at least 30 WUs stuck > in > > limbo because of this. You might also want to consider merging your 66 > > comupters into a single one! > > > > Sorry folks I've been travelling. > 1-I thought I did nothing but leave the program to run based on preferences > without changing the defoult ones. > 2-I can't find those "trickle_up_*xml</b> files you mention. > 3-I only have one computer and I have no idea of what are you taking about > when you say that "You might also want to consider merging your 66 > > comupters into a single one!" > 4-As you have guess I'm a computer user but unfortunately I can't follow your > expert computer jargon. So if you know what is wrong with my computer, please > tell me step by step how I should mend it. > Thanks > Hola, again. Found the trickle files. Took all but 30 of then into a new directory. How can I "do an update"? |
©2024 climateprediction.net
