Queue removing WU's

[cakin]

OK, is this normal? My queue has done this to 6 WU's since midnight.

10:05am: Q0001 Removing corrupted Work Unit 02jl01aa.28961.15442.129826.220

I went back and checked my logs, this started on 4-4!

I am pointed at seti.tacube. Is this a Berkeley to tacube issue or a tacube to me issue?

What would happen if I just point a Berkeley for a while?

 

[Poof]

Cakin - it could be that you really do have some corrupted WUs. Berekeley has sent some out before (and they have come in batches). I think the newer versions of Setiqueue look for them and delete them so that your machines won't try to work on them. If your machine does get one, it'll hang trying to send the bad result back to either the queue or SETI, so might as well get rid of those beforehand...

 

[cakin]

maybe that explains it, the laogs show 6-12 deleted every day since the 4th

 

[Poof]

Yeah - what happens is that you might have a set of WUs that are in sequence from some corrupted part of the tape... a part that when the data was split and sorted, the corruption was accidently missed and the WUs went out anyway. And because of the way Setiqueue works, it'll make that connection and suck down a whole pile of contiguous WUs in a row, many times with WUs in split sequence.

If during next week, you still get corrupted WUs, then you might want to check your HD though...