A weird problem
15 Feb 2007 2:45:29 UTC
Topic 192431
(moderation:
I have two computers, both of whiich do work on Einstein@home and occasionally on Seti@home. This bug only appears on one of the two, and only for Einstein@home -- even though the computers frequently use the same wi-fi network, so it can't be network related.
Scheduler request failed: HTTP internal server error.
Anyone got any ideas? Thanks.
A weird problem
)
The E@H servers have been having problems lately. Just let your hosts crunch the other project, they will get back to einstein when it is ready.
BOINC WIKI
BOINCing since 2002/12/8
RE: I have two computers,
)
I started having this problem too since 12 Feb 2007 on just 1 of 2 machines that work on Einstein. As the maching keeps trying to connect, 1 out of 3 of the error messages report the project is down, but the other machine connects fine. They are both Macs running 10.4.8 (the latest version), on the same network, and the only thing I can think of that I changed or did was I did disconnect it from the network & move it briefly, but then I reconnected it to the network & the network & Internet work fine.
I've tried restarting BOINC & rebooting the machine that can't connect, but it still can't connect to Einstein.
Odd how it is just 1 machine & not the other, which leads me to think the problem is at my end & Einstein's problems are just contributing to my "Fog of War"! ;-)
RE: Odd how it is just 1
)
You may be right, you may be seeing symptoms from 2 problems: 1 problem yours, the 2nd problem Einstein's. Is the machine connecting to other projects? If it is then probably it's just the problem at Einstein.
Read this thread carefully because it has some good advice.
BOINC FAQ Service
Official BOINC wiki
Installing BOINC on Linux
RE: Odd how it is just 1
)
One observation that has been reported in relation to the problems is that if a new data file is needed there is an increased chance of problems. That may not have any bearing on your case, or it may explain it. If the host has worked properly in the past, and no changes to the network have been made recently, I would just let it run until things clear up.
It has also been suggested that setting the client to no new work until all tasks have been reported may give you a better chance of downloading when you allow new work.
BOINC WIKI
BOINCing since 2002/12/8
RE: Read this thread
)
Thanks Dagorath & Keck Komputers for the help & good advice. Unfortunately, when I went to try out the suggested 'stop allowing work' fix, the problem had, overnight, gone away by itself, if a bit mysteriously (to my non-DC expert eye):
Note this hard working 800 MHz machine only does 2 WUs per day!
When I went to the "Results for computer" web page for that computer, it seems they had been being reported all along. No harm, no foul I guess! There does seem to be some sort of Einstein system disconnect though, e.g., left hand not knowing what the right hand is doing.
One of the 'problems' with Einstein & BOINC is that there are usually no problems, so when there are those of us only paying half attention to their doings tend to panic when there is a problem. Like the "know enough to be dangerous" types everywhere, we start pushing buttons we have not a clue as to what they do & fire off emails to "THEM" who caused our problem & distured out happy world.
Thanks again for your help!
Bruce (AKA, Molten Salt guy)