Absolutely on Linux mint. I think I have it pinned down to times when my range extender(wireless) has lost its connection and attempting to re-establish communication. During this time cpu usage will drop to zero until either communication is re-established or it gives up trying. I have seen this with numerous MCM1 units, but they seem to recover without a problem and become valid, even if this occurs multiple times.

Cheers

Yes, had to completely abandon wireless on any Xnix machine. Any instability in the connection at all crashes out the work. It's bad enough when my wired internet goes down. They don't error immediately because the machines can see the router, but it does cause all sorts of havoc.

Now wired the linux ubuntu with wrapped ethernet cable to the windows machine which never has this issue with wifi, then set windows to share the internet connection. After that it's completely configuration free, the linux on a different ip range connects to windows, where windows acts as the router with the 192.xxx.yyy.1 ip. What I'm suggesting is, if you have the linux devices nearby and a single windows up close, connect the linuxes to a router by cable and the router to the windows machine and let that act as internet sharing device. Somewhere there was also mention of running a power off command with iwconfig to stop energy saving kicking in, i.e. your wifi not going to sleep. Then it transmits on maximum all the time.

But here's an entirely different question to technicians or close beta observers. Is this project same as mcm using symbol linking to the project folder? Iow, there's only one copy of the application software and various other permanent/semi-permanent data files in the project folder and each task slot simply has an symbolic link to those files, so no copying is required, something we really really would like to see for cep2.

Just noticed that a beta repair job had slipped in, all the 0's in batch 25 .

BETA_ ugm1_ ugm1_ 00025_ 0000_ 3-- - In Progress 10/10/14 14:56:43 12/10/14 00:32:42 0.00 0.0 / 0.0
BETA_ ugm1_ ugm1_ 00025_ 0000_ 2-- 721 Error 10/10/14 14:54:36 10/10/14 14:56:41 0.00 145.1 / 0.0
BETA_ ugm1_ ugm1_ 00025_ 0000_ 1-- 721 Pending Validation 06/10/14 14:54:35 07/10/14 00:18:19 5.06 104.7 / 0.0
BETA_ ugm1_ ugm1_ 00025_ 0000_ 0-- - No Reply 06/10/14 14:54:34 10/10/14 14:54:34 0.00 0.0 / 0.0

The Result Log for the errored unit is:

<core_client_version>6.10.58</core_client_version>
<![CDATA[
<message>
Can't get shared memory segment name: shmget() failed
</message>

_1 log looks normal.

_3 is progressing ok, checkpoints every 2 minutes. Edit: now Valid.

More work units are now being released. The application version is 7.22. This beta is a potential fix for some of the sizing issues we saw in the last round of beta with some work units.

Seippel