Message boards : Number crunching : Completed WU stuck on upload
Message board moderation

To post messages, you must log in.

AuthorMessage
lobo

Send message
Joined: 14 Jan 18
Posts: 7
Credit: 212,946,240
RAC: 0
Message 3255 - Posted: 3 Feb 2019, 4:48:25 UTC

Hi,
I have several single board computers crunching on this project. One of them is having dificoulty uploading any of the completed WU (~50 at this point). do you have any suggestions on how to fix this issue. below the event log when I ask t are the messages I ask it to upload the files:

Sat 02 Feb 2019 10:39:05 PM CST | | BOINC can't access Internet - check network connection or proxy configuration.
Sat 02 Feb 2019 10:39:34 PM CST | Universe@Home | Started upload of universe_bh2_180328_253_1752693248_20000_1-999999_695400_1_r511991118_0
Sat 02 Feb 2019 10:39:34 PM CST | Universe@Home | Started upload of universe_bh2_180328_253_1759428241_20000_1-999999_430400_0_r1436960556_2
Sat 02 Feb 2019 10:39:36 PM CST | Universe@Home | Temporarily failed upload of universe_bh2_180328_253_1759428241_20000_1-999999_430400_0_r1436960556_2: transient HTTP error
Sat 02 Feb 2019 10:39:36 PM CST | Universe@Home | Backing off 00:03:25 on upload of universe_bh2_180328_253_1759428241_20000_1-999999_430400_0_r1436960556_2

I realize, the it seems a problem with my specific computer, specially since it is the only one with a problem.
BTW, the computer is connected to the internet and from what I can tell the Asteroids and SETI projects can connect without issue

Any ideas on what the problem is, and how to fix it would be greatly appreciated
Thank you
Rodrigo

thank you .
ID: 3255 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Nick Name

Send message
Joined: 22 Feb 15
Posts: 12
Credit: 46,880,289
RAC: 0
Message 3258 - Posted: 3 Feb 2019, 21:58:02 UTC - in response to Message 3255.  
Last modified: 3 Feb 2019, 22:11:01 UTC

I have some similar boards and when I've had this error I had to restart the board. The problem seemed to be caused by an internet outage and BOINC not realizing the connection was back.

If a project manual update or board restart doesn't work try modifying your cc_config file.

https://boinc.berkeley.edu/wiki/Client_configuration

Set this to 1. It's supposed to be for a proxy setup but I've seen some users report success even though they weren't running a proxy.
<http_1_0>1</http_1_0>
ID: 3258 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
lobo

Send message
Joined: 14 Jan 18
Posts: 7
Credit: 212,946,240
RAC: 0
Message 3262 - Posted: 4 Feb 2019, 2:58:13 UTC - in response to Message 3258.  

Hi Nick,
after rebooting several times I tried your suggestion, but to no avail.
After some more digging It seems like boinc can connect fine when I "update" the project. and also when other projects upload. it is only when I try to upload to Universe the I get the errors.
Also, when I enable the connection debugging after any communicative to other projects I get the message :

Sun 03 Feb 2019 06:55:45 PM CST | | [network_status] status: online

but after trying to upload It get after many messages of failed uploads I get

Sun 03 Feb 2019 06:56:26 PM CST | | [network_status] status: need connection

and when I update any project again the status is online again

Somehow I think that some configuration file might have gotten corrupted. but have no idea which one or how to fix it.

again any suggestions are welcome
ID: 3262 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Nick Name

Send message
Joined: 22 Feb 15
Posts: 12
Credit: 46,880,289
RAC: 0
Message 3265 - Posted: 4 Feb 2019, 7:04:04 UTC - in response to Message 3262.  

I don't have any "good" ideas. I haven't seen a problem like this after restart except when the project was down or having network problems.

Try enabling <file_xfer_debug>1</file_xfer_debug> in cc_config, I think that might be more helpful although I have never used it.

Have you recently seen a message like "This project is using an old URL, at your earliest convenience detach from the project"? Just in the last week I brought a board back online that had been off for several weeks. It worked fine before, but wouldn't connect to Universe and I had to detach and rejoin. I don't recall what the old URL was, but if your board was using it and it's no longer reachable maybe that's the problem.

You might try to compare project files between boards if you can, things like the job schedule / reply, if you can. Maybe look thru client_state too at the uploads on a working board, see if anything is noticeably different.

You might try PMing Krzysztof and see if he can check server logs for some of those jobs, maybe he can see what's going on. Transient errors don't usually go on for days. Maybe something weird is going on with the server.

Other than that I think you'll have to do some network troubleshooting on that board, basically making sure those requests are actually going where they're supposed to.
Team USA forum | Team USA page
Always crunching / Always recruiting
ID: 3265 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Jim1348

Send message
Joined: 28 Feb 15
Posts: 253
Credit: 200,562,581
RAC: 0
Message 3267 - Posted: 4 Feb 2019, 7:49:56 UTC - in response to Message 3262.  

Somehow I think that some configuration file might have gotten corrupted. but have no idea which one or how to fix it.

The usual way is to detach from the project, and then re-attach. I would re-boot the machine between operations as an added precaution. Good luck; it is a bit unusual.
ID: 3267 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote

Message boards : Number crunching : Completed WU stuck on upload




Copyright © 2024 Copernicus Astronomical Centre of the Polish Academy of Sciences
Project server and website managed by Krzysztof 'krzyszp' Piszczek