This separate observation would have been good to make in a new thread in the CEP2 forum. :D

If I understand correctly, you had CEP2 *de-selected* in that profile, but down below, the override set to *unlimited*. Could speculate, but speculation it's all it is, maybe the script makes it go back to default 1, when there's an active project reshuffle AND CEP2 is *not* selected?

Certainly, I've de- and re-selected CEP2 so many times to help fetch of Beta and DDDT2, saved the profile and on re-selection of CEP2, the override number staying unwavering at the preferred number. Just tried it with *unlimited* and selecting, de-selecting, re-selecting, with saves in between, and the *unlimited* stuck. Really could not say if there is a bug here. Now that HFCC has been moved to active, CEP2 staying selected on my profiles and the preferred number of CEP2 to be cached not changing, it's waiting on members to report on allocated work dropping to 1. The same observation may then maybe also occur when next Project X launches. There's the time recording issue left to solve, so hopefully it wont be too long to happen, helping stretch the remaining run days on HFCC, for those that have little crunching resources and a higher badge goal. :))))

Thanks for warning, coolstream... we'll be homed in on this in the event. Production was as stable as the first law on thermodynamics [conservation of energy]... no red face smiley posts yet :D

--//--

edit: spell 'too', with double o

I didn't want to start a 'panic' for something that I couldn't be 100% uncertain about, so I went with uplinger's request to post if we 'noticed anything'.

NO. I actually continue to have CEP selected to process 'when available'.

That was my thought too, although I would have thought: 'the script makes it go back to default 1, when there's an active project reshuffle AND CEP2 *is* selected' would be the question.

Yes, but it certainly isn't worth making the project Inactive again for long enough to make changes to a profile just to test the question.

I have changed the CEP settings back to unlimited and will observe if this happens again if another project switches back to being Active.

I haven't checked the CEP forum to see if you have posted about this yet, but if not, I will make a post there too.

EDIT: I have made a post in CEP forum about this.

In between I had on the same host around 30 errored WUs with the same failure and later successfully computed HFCC WUs.
Strange !
Yves

Hi SekeRob,
It is widely a WCG only wired host (no wifi).
I am not sure if it could be a temperature problem (although the host is properly cooled) since we experience currently really hot weather.
During the SIGILL problem with HFCC, other projects (HCMD2, CEP2) have been successfully crunched on the same host.
Finally I rebooted the host.
Cheers,
Yves

Probably your computer is going to be tested which is why a second copy will be sent.

It should go back to normal once the inconclusives are validated