Questions and Answers : Windows : Unable to upload finished results
Message board moderation
Author | Message |
---|---|
Send message Joined: 23 Jun 06 Posts: 1 Credit: 2,218,762 RAC: 0 |
4/26/2007 1:23:21 PM|climateprediction.net|Sending scheduler request to http://climateapps2.oucs.ox.ac.uk/cpdnboinc_cgi/cgi 4/26/2007 1:23:21 PM|climateprediction.net|Reason: To send trickle-up message 4/26/2007 1:23:21 PM|climateprediction.net|(not requesting new work or reporting completed tasks) 4/26/2007 1:23:26 PM|climateprediction.net|Scheduler request succeeded 4/26/2007 1:24:55 PM|climateprediction.net|Started upload of file hadcm3ohe_0x3g_05691230_1_2.zip 4/26/2007 1:26:53 PM|climateprediction.net|Error on file upload: can\'t write file /home/boinc/data/hadcm3ohe_0x3g_05691230_1_2.zip: No space left on device 4/26/2007 1:26:53 PM|climateprediction.net|Temporarily failed upload of hadcm3ohe_0x3g_05691230_1_2.zip: transient upload error 4/26/2007 1:26:53 PM|climateprediction.net|Backing off 1 minutes and 0 seconds on upload of file hadcm3ohe_0x3g_05691230_1_2.zip 4/26/2007 1:27:53 PM|climateprediction.net|Started upload of file hadcm3ohe_0x3g_05691230_1_2.zip 4/26/2007 1:29:50 PM|climateprediction.net|Error on file upload: can\'t write file /home/boinc/data/hadcm3ohe_0x3g_05691230_1_2.zip: No space left on device 4/26/2007 1:29:50 PM|climateprediction.net|Temporarily failed upload of hadcm3ohe_0x3g_05691230_1_2.zip: transient upload error 4/26/2007 1:29:50 PM|climateprediction.net|Backing off 1 minutes and 0 seconds on upload of file hadcm3ohe_0x3g_05691230_1_2.zip 4/26/2007 1:30:50 PM|climateprediction.net|Started upload of file hadcm3ohe_0x3g_05691230_1_2.zip 4/26/2007 1:32:48 PM|climateprediction.net|Error on file upload: can\'t write file /home/boinc/data/hadcm3ohe_0x3g_05691230_1_2.zip: No space left on device 4/26/2007 1:32:48 PM|climateprediction.net|Temporarily failed upload of hadcm3ohe_0x3g_05691230_1_2.zip: transient upload error 4/26/2007 1:32:48 PM|climateprediction.net|Backing off 1 minutes and 0 seconds on upload of file hadcm3ohe_0x3g_05691230_1_2.zip 4/26/2007 1:33:48 PM|climateprediction.net|Started upload of file hadcm3ohe_0x3g_05691230_1_2.zip 4/26/2007 1:35:47 PM|climateprediction.net|Error on file upload: can\'t write file /home/boinc/data/hadcm3ohe_0x3g_05691230_1_2.zip: No space left on device 4/26/2007 1:35:47 PM|climateprediction.net|Temporarily failed upload of hadcm3ohe_0x3g_05691230_1_2.zip: transient upload error 4/26/2007 1:35:47 PM|climateprediction.net|Backing off 1 minutes and 0 seconds on upload of file hadcm3ohe_0x3g_05691230_1_2.zip 4/26/2007 1:36:47 PM|climateprediction.net|Started upload of file hadcm3ohe_0x3g_05691230_1_2.zip 4/26/2007 1:38:47 PM|climateprediction.net|Error on file upload: can\'t write file /home/boinc/data/hadcm3ohe_0x3g_05691230_1_2.zip: No space left on device 4/26/2007 1:38:47 PM|climateprediction.net|Temporarily failed upload of hadcm3ohe_0x3g_05691230_1_2.zip: transient upload error 4/26/2007 1:38:47 PM|climateprediction.net|Backing off 1 minutes and 37 seconds on upload of file hadcm3ohe_0x3g_05691230_1_2.zip 4/26/2007 1:40:25 PM|climateprediction.net|Started upload of file hadcm3ohe_0x3g_05691230_1_2.zip 4/26/2007 1:42:22 PM|climateprediction.net|Error on file upload: can\'t write file /home/boinc/data/hadcm3ohe_0x3g_05691230_1_2.zip: No space left on device 4/26/2007 1:42:22 PM|climateprediction.net|Temporarily failed upload of hadcm3ohe_0x3g_05691230_1_2.zip: transient upload error 4/26/2007 1:42:22 PM|climateprediction.net|Backing off 1 minutes and 59 seconds on upload of file hadcm3ohe_0x3g_05691230_1_2.zip 4/26/2007 1:44:22 PM|climateprediction.net|Started upload of file hadcm3ohe_0x3g_05691230_1_2.zip 4/26/2007 1:46:22 PM|climateprediction.net|Error on file upload: can\'t write file /home/boinc/data/hadcm3ohe_0x3g_05691230_1_2.zip: No space left on device 4/26/2007 1:46:22 PM|climateprediction.net|Temporarily failed upload of hadcm3ohe_0x3g_05691230_1_2.zip: transient upload error 4/26/2007 1:46:22 PM|climateprediction.net|Backing off 14 minutes and 51 seconds on upload of file hadcm3ohe_0x3g_05691230_1_2.zip 4/26/2007 2:01:14 PM|climateprediction.net|Started upload of file hadcm3ohe_0x3g_05691230_1_2.zip 4/26/2007 2:03:13 PM|climateprediction.net|Error on file upload: can\'t write file /home/boinc/data/hadcm3ohe_0x3g_05691230_1_2.zip: No space left on device 4/26/2007 2:03:13 PM|climateprediction.net|Temporarily failed upload of hadcm3ohe_0x3g_05691230_1_2.zip: transient upload error 4/26/2007 2:03:13 PM|climateprediction.net|Backing off 36 minutes and 18 seconds on upload of file hadcm3ohe_0x3g_05691230_1_2.zip 4/26/2007 2:27:54 PM|climateprediction.net|Sending scheduler request to http://climateapps2.oucs.ox.ac.uk/cpdnboinc_cgi/cgi 4/26/2007 2:27:54 PM|climateprediction.net|Reason: Requested by user 4/26/2007 2:27:54 PM|climateprediction.net|(not requesting new work or reporting completed tasks) 4/26/2007 2:27:59 PM|climateprediction.net|Scheduler request succeeded 4/26/2007 2:28:10 PM|climateprediction.net|Started upload of file hadcm3ohe_0x3g_05691230_1_2.zip 4/26/2007 2:30:10 PM|climateprediction.net|Error on file upload: can\'t write file /home/boinc/data/hadcm3ohe_0x3g_05691230_1_2.zip: No space left on device 4/26/2007 2:30:10 PM|climateprediction.net|Temporarily failed upload of hadcm3ohe_0x3g_05691230_1_2.zip: transient upload error 4/26/2007 2:30:10 PM|climateprediction.net|Backing off 1 hours, 20 minutes and 5 seconds on upload of file hadcm3ohe_0x3g_05691230_1_2.zip |
Send message Joined: 5 Feb 05 Posts: 465 Credit: 1,914,189 RAC: 0 |
Normal occurrence around here. The project needs to open up more space. May take a couple of days. |
Send message Joined: 20 Mar 05 Posts: 3 Credit: 161,424 RAC: 0 |
Same problem here. I won\'t bore you with the error messages. It uploads an approx 8mb file, but at the end cannot write the file. I\'m sure they\'ll have it fixed soon enough. Cheers, Landog |
Send message Joined: 29 Sep 04 Posts: 2363 Credit: 14,611,758 RAC: 0 |
There\'s an item about this problem in the News thread (in the Number crunching section): http://climateapps2.oucs.ox.ac.uk/cpdnboinc/forum_thread.php?id=5447 Cpdn news |
Send message Joined: 20 Sep 04 Posts: 45 Credit: 1,652,249 RAC: 0 |
Hi all, I\'ve read that \"Milo\" has made some space on the server (11th May 2007), but as of today and for the last few days, I keep getting these error messages: 14/05/2007 17:17:48|BBC Climate Change Experiment|[file_xfer] Started upload of file hadcm3ln_8gvu_00634304_0_14.zip 14/05/2007 17:17:53||Project communication failed: attempting access to reference site 14/05/2007 17:17:53|BBC Climate Change Experiment|[file_xfer] Temporarily failed upload of hadcm3ln_8gvu_00634304_0_14.zip: http error 14/05/2007 17:17:54||Access to reference site succeeded - project servers may be temporarily down. 14/05/2007 17:17:54|BBC Climate Change Experiment|[file_xfer] Started upload of file hadcm3ln_8gvu_00634304_0_14.zip 14/05/2007 17:17:59||Project communication failed: attempting access to reference site 14/05/2007 17:17:59|BBC Climate Change Experiment|[file_xfer] Temporarily failed upload of hadcm3ln_8gvu_00634304_0_14.zip: http error 14/05/2007 17:17:59|BBC Climate Change Experiment|Backing off 1 hr 30 min 7 sec on upload of file hadcm3ln_8gvu_00634304_0_14.zip 14/05/2007 17:18:00||Access to reference site succeeded - project servers may be temporarily down. Edit: I now find out that there is a 14 day retry limit of uploading these ZIP files. So, I now think that the previous 0_13.zip file is now lost - it was \"stuck\" and disappeared from the \"transfer\" window recently. So I thought \"great\" - it\'s been uploaded. But chances are that\'s it\'s been deleted by BOINC Manager.... So, this now really sucks as a LONG BBC WU will now be incomplete due to some external issue.... /Edit Uploading to other projects is not a problem. Trickles to (and credits from) BBC project site are fine. BOINC version is 5.8.16 So, is this another CPDN project server related issue or what? regards Tim (PS Sorry if this is the wrong thread, but the posting shown here: http://climateapps2.oucs.ox.ac.uk/cpdnboinc/forum_thread.php?id=5447 cannot be replied to !! (Mods only??) regards, Tim UK BOINC Team Founder Join the UK BOINC Team: http://www.ukboincteam.org.uk/newforum |
Send message Joined: 29 Sep 04 Posts: 2363 Credit: 14,611,758 RAC: 0 |
Hi Tim I think the most probable scenario is that your _13 file uploaded together with a trickle when you allowed network access again after the first server outage. You can check all your boinc manager messages to see whether the upload of this file was reported. Exit from boinc/CCE manager Go to C Program files Double-click on Climate Change Experiment to open the folder Double-click on stdoutdae.text to look at the messages contents Tell us what you find The BBC server status now looks OK: http://bbc.cpdn.org/server_status.php Have you tried to upload your _14 file again? You couldn\'t post in the News thread because it\'s locked and for mods only. We try to keep discussion out of the News thread to make it easier for infrequent forum visitors to digest. I tried to look at the server pages for you model to check on the dates and trickles. Your computer/s are hidden. It would help us get to the bottom of what happened to that file if you please unhide your computer, at least for the time being. Cpdn news |
Send message Joined: 20 Sep 04 Posts: 45 Credit: 1,652,249 RAC: 0 |
\"mo.v\" wrote: Double-click on stdoutdae.text to look at the messages contents Ok, here\'s the relevant section:
The _13.zip file is now nowhere to be found on my PC, so must have been deleted. \"mo.v\" wrote: The BBC server status now looks OK: Glad to hear it\'s working....but I still get (as of now):
.... \"mo.v\" wrote: I tried to look at the server pages for you model to check on the dates and trickles. Your computer/s are hidden. It would help us get to the bottom of what happened to that file if you please unhide your computer, at least for the time being. Er, nope - my (single) PC crunching on BBC project is NOT hidden - but my CPDN project PC was hidden...but that\'s not the project for which I have an issue !! It\'s BBC WU \"ZIP\" file that won\'t upload (but trickles do !!, as I\'m getting credits for the work returned). (This is the link to the name of the WU, for which the ZIP file that won\'t upload: http://bbc.cpdn.org/workunit.php?wuid=644462 Hope all the above helps...! regards Tim regards, Tim UK BOINC Team Founder Join the UK BOINC Team: http://www.ukboincteam.org.uk/newforum |
Send message Joined: 29 Sep 04 Posts: 2363 Credit: 14,611,758 RAC: 0 |
Yes thank you Tim, that\'s very helpful. I didn\'t notice I had to look for your model on the BBC site, sorry. Regarding the _13 upload, I\'ve checked on the model\'s trickles and going by the timesteps recorded it looks as if it reached 2050 and trickled for that year on 30 April. The server must have been accepting trickles but not file uploads. The two-week timeout period starts when the file first attempts to upload. The upload of this _13 zip file was abandoned on 14 May which is 2 weeks later. This is really unfortunate because as you say, the file is gone and can\'t be retrieved. Even if you restored a backup from before the creation date of the _13 file, the file wouldn\'t be created again second time round. The trickle would, but not the zip file. I\'m going to ask the programmers whether a longer deadline for the upload of these zip files can be written into the code for future models. Because your model\'s so near the end, I don\'t think you should abandon it because of this. I\'m pretty sure that some models that missed their 2040 zip upload have been included in the overall results. So you have to get the _14 zip file to Oxford. I\'ve just looked at the BBC server status and it\'s all GO. The http error in the messages looks like a firewall problem. \'Access to reference site succeeded\' means that your internet connection can contact Google, ie it\'s working, but can\'t connect to the BBC server. Look please at this http://www.climateprediction.net/board/viewtopic.php?p=53711&highlight=http#53711 Hope that helps. Cpdn news |
Send message Joined: 13 Jan 06 Posts: 1498 Credit: 15,613,038 RAC: 0 |
If there is a backup which contains the _13 zip, then it should be possible to get it to be sent to the project using Thyme\'s technique? I'm a volunteer and my views are my own. News and Announcements and FAQ |
Send message Joined: 20 Sep 04 Posts: 45 Credit: 1,652,249 RAC: 0 |
\"mo.v\" wrote: Yes thank you Tim, that\'s very helpful. I didn\'t notice I had to look for your model on the BBC site, sorry. Understood - I\'ve made a backup of the _14.zip file so, if there\'s another mechanism to upload it (email?), then of course that route can be taken. Of course, if this is now an on-going issue, then once the WU is finshed (468 hours time), then I may have a few ZIP files to send !! \"mo.v\" wrote: I\'m going to ask the programmers whether a longer deadline for the upload of these zip files can be written into the code for future models. Don\'ty worry - I know the importance of completing the entire WU...and after 3966 hours of crunching, I\'m not going to stop now. I doubt there\'s much point to making a programming change, as the BBC project is nearing it\'s end UNLESS of course this is a common problem affecting other crunchers....although why it\'s taken until now to surface is a bit of a concern...how many other \"ZIP\" files have \"vanished\" because people assumed that once it had left the \"transfers\" tab, it had been uploaded (although it had been deleted !!). And we all know about teh lack of resources of BOINC projects (in most cases) but this highlights a big concern, that all this work being done is for \"nought\" if the projects severs are \"down\" for any length of time. Maybe there should be a central \"back-up\" server, operated jointly by all BOINC projects so, if a specific server failed, at least another upload route is available....? This degree of \"redundancy\" can then prevent lost WUs - and they can be re-sent from the central server to the specific projects once their own servers are back up. \"mo.v\" wrote: So you have to get the _14 zip file to Oxford. I\'ve just looked at the BBC server status and it\'s all GO. The http error in the messages looks like a firewall problem. \'Access to reference site succeeded\' means that your internet connection can contact Google, ie it\'s working, but can\'t connect to the BBC server. Still no joy - just tried it....! \"mo.v\" wrote: Look please at this Well, I switched the XP firewall off, shut down BOINC and restarted it. Still stuck and I\'m getting the same http error. What\'s strange is that other projects are uploading files OK and nothing has changed on my PC or router for ages....OK, so maybe a Windows Update might have changed something, so I\'ll take a look for that. My router firewall only blocks incoming stuff. Of course, it might be my ISP who is blocking things....however, where on earth do I find out the upload server URL\'s in order to check things? If I can find out the correct upload URL, then I can do a simple \"ping\" and \"traceroute\" to establish that they can be seen external to BOINC Manager...that at least might help to confirm where the problem might lie. regards Tim regards, Tim UK BOINC Team Founder Join the UK BOINC Team: http://www.ukboincteam.org.uk/newforum |
Send message Joined: 20 Sep 04 Posts: 45 Credit: 1,652,249 RAC: 0 |
\"MikeMarsUK\" wrote:
Hi Mike, Thanks for the idea. It had occurred to me, but my last backup on this PC was quite a while before the _13.zip file was created.... BOINC is the only program that resides on my C drive, and I rarely back this up as the PC has been very reliable up to now (that\'s it - the \"kiss of death\" has been invoked !!). As such, backing up the drive has never been a major priority.... :-( as there\'s been very little to \"lose\", as most projects I crunch on are much shorter duration. I\'ve unsuccessfully completed BBC, CPDN and SAP WUs before now...usually the WU errors....but this is my 1st BBC WU (of 3) that might actually complete....468 hours to go.....and now I get this file to upload that won\'t.... regards Tim regards, Tim UK BOINC Team Founder Join the UK BOINC Team: http://www.ukboincteam.org.uk/newforum |
Send message Joined: 20 Sep 04 Posts: 45 Credit: 1,652,249 RAC: 0 |
\"mo.v\" wrote: ..... OK, panic over. Seems there *was* a connection issue via the ISP, which has now been resolved. _14.zip has now uploaded !! That\'s a weight off my mind.....! regards, Tim regards, Tim UK BOINC Team Founder Join the UK BOINC Team: http://www.ukboincteam.org.uk/newforum |
Send message Joined: 7 Nov 06 Posts: 21 Credit: 43,546 RAC: 0 |
Even if you restored a backup from before the creation date of the _13 file, the file wouldn\'t be created again second time round. The trickle would, but not the zip file. I really don’t understand this. I asked you about it in my thread, if you recall and you said that the program does not ask the server; it depends on commands from within its own code. I had assumed that all data is stored within the Boinc directory. Therefore, if the Boinc directory was to be restored, the data is restored with it. When the restored program starts to run again it can’t know that a copy of itself ran past the _13 point unless that fact is stored somewhere else, either on the server or somewhere else on the host computer. It’s not on the server – and anyway, the server would not record it as uploaded unless it had been. I would be surprised if it is outside the Boinc directory – say in the Registry – as I can’t see the point in it and it would seem designed to make recovery difficult. I would also be disappointed if the project hid its data outside the Boinc directory where it could get left behind and clutter the computer up and I do not wish to accuse your programmers of doing it. This leads me to the conclusion that the data is within the Boinc directory and will be restored along with the other files. Therefore, if this is so, the restore will follow exactly the same logic, examine exactly the same data and will arrive at the same conclusion: the ten year file is to be produced at the appropriate time. Or have I missed something? Andy |
Send message Joined: 29 Sep 04 Posts: 2363 Credit: 14,611,758 RAC: 0 |
Tim, I\'m glad the http problem is now sorted. If I were you I\'d make another backup now of the complete contents of the boinc folder. In a sense, the closer a model is to finishing, the more valuable a backup is. Get to Les\'s easy manual backup method through my sig. It\'s item #1 in the README about avoiding model crashes. Tim, if the programmers are able to write a longer file upload deadline into models, I\'m sure this would only be for future cpdn models. Current BBC and cpdn models wouldn\'t benefit. Even with the new server in action, future server problems can never be entirely ruled out and they can take ages to solve. I\'m thinking of crunchers who run models with network activity opened every day and never visit the forums even to look at the news and announcements. BTW, if you have a projects section on your team forum, could you please if necessary add a link to the cpdn News thread? It would also be worth providing a link to the 4 READMEs which for both BBC and cpdn crunchers are the most useful trouble-shooting resource. It\'s all in my sig. Andy, I\'ll have to ask somebody who knows more than me to explain about how the 10-year zip files are created. I\'m basing what I say on my own rerun of an restored old backup (blush!) of a BBC model that had to again get past two 10-year zip file creation points. I sat and watched to see what was created during the restored second run. The trickles were recreated and sent (but ignored by the server which knew it had already received them). The decadal zip file was not created again during the rerun. Cpdn news |
Send message Joined: 7 Nov 06 Posts: 21 Credit: 43,546 RAC: 0 |
Your evidence looks convincing. I’m well puzzled. I would certainly be interested in any explanation that you are offered. Andy |
Send message Joined: 29 Sep 04 Posts: 2363 Credit: 14,611,758 RAC: 0 |
Tim, please also see Tolu\'s post here: http://climateapps2.oucs.ox.ac.uk/cpdnboinc/forum_thread.php?id=5528 Cpdn news |
©2025 cpdn.org