 |
|
|
1)
Questions and Answers :
Windows :
Visual Fortran Run-Time Error
(Message 50762)
Posted 9 Nov 2014 by  miw
Post: OK. Thanks. I've disabled this application for now. I had read the other threads. I don't actually mind if the odd unit fails because the envelope is being pushed, and Was aware that stopping and starting these models was somewhat risky so could also accept the odd bomb at PC restart which happens pretty rarely anyhow. But this behaviour is 100% mortality, which I had not seen before. I had assumed it was because there was a group of units that were bad, and that they were all that was left in the pool. |
|
2)
Questions and Answers :
Windows :
Visual Fortran Run-Time Error
(Message 50760)
Posted 9 Nov 2014 by miw
Post: I'm getting exactly the same error on every single Coupled Model Full resolution job I get. Happens within a few seconds of start. It looks like climate.cpdc is an initial data file for the job, and a batch of bad WUs got out and are being recycled indefinitely. Hopefully they will be tracked down and killed soon. (is there a function that pulls a job out of circulation when it returns n times with error status?) It's a pain because it occupies a slot doing nothing useful until I click away all 6 visual Fortran messages. |
|
3)
Message boards :
Number crunching :
manticore.oerc.ox.ac.uk offline ?
(Message 45547)
Posted 12 Feb 2013 by miw
Post: Yes. I have files that are supposed to be being downloaded from Manticore, but I cannot ping or otherwise contact manticore.oerc.ox.ac.uk Traceroute gets to cengb.backbone.ox.ac.uk |
|
4)
Message boards :
Number crunching :
Project File Upload handler is missing
(Message 40968)
Posted 3 Nov 2010 by miw
Post: Ha. I am no vi power user, but I am just proficient enough to never have bothered to learn emacs. For my money they are the only general text editors that have enough power. OK. I sometimes use ed when I'm on a very slow link. dunno if it gives any clue, to convert "a" to "n" you invert the least significant nybble... --mark. |
|
5)
Message boards :
Number crunching :
Project File Upload handler is missing
(Message 40966)
Posted 3 Nov 2010 by miw
Post: /etc/init.d/boinc_client stop cp client_state.xml client_state.bak vi client_state.xml :.,$s/hnndler/handler/ :wq /etc/init.d/boinc_client start all is well by the looks. The third and fourth of the files are uploading bravely as I type. if it gives any clue, to convert "a" to "n" you invert the lower-order nybble... Thanks a bunch!! cheers, Mark |
|
6)
Message boards :
Number crunching :
Project File Upload handler is missing
(Message 40964)
Posted 3 Nov 2010 by miw
Post: I think this is it: --mark -------------- <name>hadam3p_pnw_v4d2_1998_1_006723123_0_1.zip</name> <nbytes>6017245.000000</nbytes> <max_nbytes>150000000.000000</max_nbytes> <md5_cksum>f71d66610ab993a15f2d818467f49b3b</md5_cksum> <generated_locally/> <status>1</status> <upload_when_present/> <url>http://boinc1.coas.oregonstate.edu/cpdn_cgi_main/file_upload_hnndler</url> <persistent_file_xfer> <num_retries>134</num_retries> <first_request_time>1286795614.498710</first_request_time> <next_request_time>1288783623.952069</next_request_time> <time_so_far>483.509431</time_so_far> <last_bytes_xferred>323.000000</last_bytes_xferred> </persistent_file_xfer> <signed_xml> <name>hadam3p_pnw_v4d2_1998_1_006723123_0_1.zip</name> <generated_locally/> <upload_when_present/> <max_nbytes>150000000</max_nbytes> <url> http://boinc1.coas.oregonstate.edu/cpdn_cgi_main/file_upload_handler </url> -------------------------------------------- |
|
7)
Message boards :
Number crunching :
Project File Upload handler is missing
(Message 40961)
Posted 3 Nov 2010 by miw
Post: Spoke too soon. Hadam3p_pnw files are all failing again with the original error message. Looks like that upload server is down. At least the FAMOUS files all got through. I'll wait a day or so and see what happens.... --mark |
|
8)
Message boards :
Number crunching :
Project File Upload handler is missing
(Message 40960)
Posted 3 Nov 2010 by miw
Post: OK. More information: It is a hadam3p_pnw workunit, and it seems that posting here has solved the issue because now the files have started to upload, as have the Famous files that were blocked behind it... --mark |
|
9)
Message boards :
Number crunching :
Project File Upload handler is missing
(Message 40959)
Posted 3 Nov 2010 by miw
Post: Host: 1055668 Ubuntu 10.10 (64 bit) , Boinc 6.10.58 Have had a unit trying to upload for a couple of days, but file uploads always fail with: project file upload handler is missing Is this server side or client side? Problem persistant across the recent CPDN database server shutdown and also a reboot of the machine. --Mark |
|
10)
Message boards :
Number crunching :
RAC for HADAM3P Fixed!
(Message 38489)
Posted 9 Dec 2009 by miw
Post: But is it why I lost over 2 million credits last night? Has anyone else lost a bunch of credit recently? Mark The Recent Average Credit (RAC) for hadam3p has been unrealistically inflated since this model type was released in March 2009. While the total credits for each hadam3p model was and is correct, the RAC was way too high. Milo fixed this today (Dec 8) and the RAC for those running any hadam3p models should now be what one would expect for the average credit per day. |
|
11)
Message boards :
climateprediction.net Science :
HADCM3 5.15?
(Message 24860)
Posted 23 Oct 2006 by miw
Post: Cool. Thanks! |
|
12)
Message boards :
climateprediction.net Science :
HADCM3 5.15?
(Message 24851)
Posted 23 Oct 2006 by miw
Post: I recently got a model with the app name hadcm3 5.15 (instead of hadcm3lb 5.15). The name and the result page on the web are very reminiscent of the models that had to be killed with the killer trickle back in April (except the version is different I guess.) Is all well with this WU, or is it a throwback? --Mark |
|
13)
Message boards :
Number crunching :
Trickles, credits, etc.?
(Message 23308)
Posted 22 Jun 2006 by miw
Post: Well, a bunch of trickles have been posted to my WUs. So, the next credit update I will jump a lot! Well, for me about 20 trickles got posted, but at this stage exactly zero credits. I think \"some\" is the operative term here. :-) Mark |
|
14)
Message boards :
Number crunching :
Boinc Client Host Total Stats not Updating for CPDN
(Message 12650)
Posted 18 May 2005 by miw
Post: Using any of the BOINC clients 4.32 or greater that have the stats tab, I note that under the "Host Total" category, the CPDN host total has not gone up since about 9th May. All the other projects I have are adding to the Host total, just CPDN is not, and CPDN was fine up to 9th May. Same problem on all my hosts, and it exists under all versions I have tried from 4.32 up to 4.42. Any clues? Is CPDN just not sending host total information out anymore? --Mark |
©2024 climateprediction.net