Scheduler request failed: HTTP internal server error

Message boards : Number crunching : Scheduler request failed: HTTP internal server error

To post messages, you must log in.

Previous · 1 · 2 · 3

AuthorMessage
Gerry Rough
Avatar

Send message
Joined: 1 May 09
Posts: 19
Credit: 91,491
RAC: 0
Message 803 - Posted: 15 Jul 2009, 23:59:48 UTC - in response to Message 802.  

seems this is on a comeback tour with this batch... bringing it back up..

as Steven is having a problem..

also, adding, a good way to prevent this is to set your cache to 1 day or less and have your computer connect every .1 days....


I have my hosts set to 1 days cache and connect every .5 days. This seems to work well for me. Prior to these settings, I had either 2 or 3 days cache and connect every 1 day, and everything got really messed up with very large caches constantly. Your advice is well taken from someone who has been there. :-)

But I also have an idea to get those results reported. What about manually highlighting those uploading files and aborting the transfer to all but, say, one hundred or thereabouts. That would upload only one hundred files at a time, rather than the mass upload that is killing the transfers and the server. Would this help?

(Click for detailed stats)
ID: 803 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile Tim Turner
Avatar

Send message
Joined: 1 May 09
Posts: 570
Credit: 184,322
RAC: 0
Message 807 - Posted: 16 Jul 2009, 0:46:27 UTC - in response to Message 803.  

it's not a upload problem it's just a reporting problem..... he's able to upload from what i see in the up/down error page...
ID: 807 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile [AF>Libristes] Dudumomo

Send message
Joined: 1 May 09
Posts: 63
Credit: 39,940
RAC: 0
Message 897 - Posted: 28 Jul 2009, 7:43:40 UTC - in response to Message 807.  

I would like to thanks the authors to this how to : http://boincfaq.mundayweb.com/index.php?language=1&view=493
I did it and it works !
Thank you !!!
ID: 897 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile Ageless
Avatar

Send message
Joined: 11 Apr 09
Posts: 172
Credit: 7,631
RAC: 0
Message 898 - Posted: 28 Jul 2009, 8:19:21 UTC - in response to Message 897.  

You're welcome. :-)
Jord

'Cause you seem like an orchard of mines, Just take one step at a time.
ID: 898 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile [AF>Libristes] Dudumomo

Send message
Joined: 1 May 09
Posts: 63
Credit: 39,940
RAC: 0
Message 904 - Posted: 28 Jul 2009, 20:12:56 UTC - in response to Message 898.  

I just wonder if it is possible for you, admins, to correct this problem ?
Because for user who don't have H24 connection, it's impossible to keep a sufficient cache.
I would be glad to run at 100% on this project, but I can't, It is too much for me to manage it. (1 or 2 day cache would be perfect, but it means on a Quad on Linux, more than 3500W WOOW)

Thanks for all your great job so far !

(By the way, this batch of WUs seems to get a muh higher rate of error than the previous one)
ID: 904 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile Ageless
Avatar

Send message
Joined: 11 Apr 09
Posts: 172
Credit: 7,631
RAC: 0
Message 905 - Posted: 28 Jul 2009, 21:18:09 UTC - in response to Message 904.  

I just wonder if it is possible for you, admins, to correct this problem ?

No, we can't. It's either a limitation in the database or in the back-end software. David Anderson has looked into it already but after trying a couple of changes, not managed to find what is causing it. So we assume it's a database limitation. We can't change database software as the BOINC server software depends on MySQL.
Jord

'Cause you seem like an orchard of mines, Just take one step at a time.
ID: 905 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile [AF>Libristes] Dudumomo

Send message
Joined: 1 May 09
Posts: 63
Credit: 39,940
RAC: 0
Message 908 - Posted: 29 Jul 2009, 8:21:33 UTC - in response to Message 905.  

Okay I see.
One solution could be to increase the length of WU when the project will be stable ?
ID: 908 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile Ageless
Avatar

Send message
Joined: 11 Apr 09
Posts: 172
Credit: 7,631
RAC: 0
Message 909 - Posted: 29 Jul 2009, 9:47:08 UTC - in response to Message 908.  

One solution could be to increase the length of WU when the project will be stable ?

It's not that easy. The Autodock and Autodock with MGL Tools applications will fly through work. That's how they are programmed, not much we can do about that. The longer running Gromacs application would be a solution, but we can't do everything with Gromacs.

For the moment all we can ask of all of you is not to use a large cache. Or if you must do so, to manually report work frequently.
Jord

'Cause you seem like an orchard of mines, Just take one step at a time.
ID: 909 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile [AF>Libristes] Dudumomo

Send message
Joined: 1 May 09
Posts: 63
Credit: 39,940
RAC: 0
Message 910 - Posted: 29 Jul 2009, 11:16:35 UTC - in response to Message 909.  

Okay !
Anyway I found my solution !
I keep a sufficient cache (like 500Wu) and suspend the connection, then every night or when I can, I start to upload my result, and because I have a slow connection, I have the time to update the project every 50WU uploaded for example.

I've just tried it and it works great !
It is easier for me to do it once a day than several times.
ID: 910 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Previous · 1 · 2 · 3

Message boards : Number crunching : Scheduler request failed: HTTP internal server error


©2018 All rights reserved | Design by Digital BioPharm Ltd