SMP client terminating

[delvryboy]

I have had the client terminate about 5-6 times yesterday and had to manually restart it. Causing a WU to finish in about 20 hours, compared to the 12 it was on par for. Today, I had it happen again, taking a 91% complete WU with it. I'm really unsure if its something on my end, or a common problem.

I'm running the SMP on a quad w/XP.

Thanks

 

[delvryboy]

It did it again. I reconfigured it to run as a service, so at least it starts on its own again. My worry is that I will lose another WU.

 

[Xilikon]

This is a bad idea... You must find out why it's terminating. About that, what do you mean by terminating : stalling, window closing or what ?

 

[delvryboy]

Window closes. I look in the event log and it says DeinoMPI has terminated unexpectedly.

 

[delvryboy]

Some events say the same ^ but FAH. I also get an error that says Fah has hung on startup.

 

[Xilikon]

OK, I suggest you delete this version and install the MPIEXEC one : http://www.stanford.edu/~kasson/folding/Folding@Home Windows SMP Client.EXE which is the 5.91 and not the 5.92 you are running (it's considered a rough beta, more complicated to setup).

 

[delvryboy]

OK, let me try and get through the current WU and I'll give that a whirl.

The new quad is good for 2600+PPD though..... nice change

 

[delvryboy]

It's been running smooth for 24 hours now. I'm not going to change anything until it acts up again.

 

[Tigerbiten]

OK, let me try and get through the current WU and I'll give that a whirl.

The new quad is good for 2600+PPD though..... nice change

You can swop the clients in the middle of a protien with just the normal shutdown-startup risk.

What speed are you running and which protien are you crunching ??

I got my 6 new Quads at 3.1 Ghz, running twin VM, crunching p2605 = ~4,170 PpD per Quad. 25,000 PpD
But they are pure folding boxen.

Luck ...........