Not sure what causes it, but whenever I set my Network to Enabled (which will cause 24 machines to start connecting to various BOINC Projects), I always find tons of EAH errors (Download errror : Download failed!)
Random machines are affected, effectively and quickly reducing the Daily Quota (some machines downto 1/day managed to fail exactly that one WorkUnit download, rendering EAH with 0 computing time for that day)
Since I'm not seeing the same behaviour with Folding@Home for example, it seems to happen with EAH only.
Copyright © 2024 Einstein@Home. All rights reserved.
Lots of Download errors
)
The de server has been having issues. Change your timezone 3 hours back and you will download from the uk server.
Thanks for the quick reply,
)
Thanks for the quick reply, that explains the download mess :)
Forgot to ask : Are they
)
Forgot to ask :
Are they going to fix it anytime soon ?
(every day causes me alone about 100MB+ (!) wasted Traffic on failed downloads)
Since it doesn't seem feasible to change timezone on my entire Network, I'm considering to Pause EAH until the Problem is fixed.
I think the errors are on the
)
I think the errors are on the application files only, but not sure. It is currently a little after 2PM local UWM time (where Einstein is from) on the first full working day after the holidays. So, I suspect it should get fixed in the next 24 hours or less, if nothing else is majorly broke.
The path is the same on all
)
The path is the same on all download servers, it's always "/download"
If one would modify /etc/hosts and add
130.209.45.153 einstein.aei.mpg.de
maybe it would fix the problem until it's fixed on server side.
Quite some work though if one has a zillion of computers like FalconFly ;-)
RE: Quite some work though
)
If he has routers available, he can just program the routers with that information and they all would forward automatically.
Hm, no Router here
)
Hm, no Router here :(
Anyway, I'm just seeing another 100MB of failed transfers, so I'll just set everything to "Request no more work from Project" since the Traffic is getting out of hands (I'm already looking at about 5GB per month for the various BOINC Projects, if everything runs normal :p )
The Source of the Errors are various, excessive MD5 Checksum errors (both for the Albert Applications and random Data/WorkUnit Files)