Simon Hood <[email protected]> writes: >> I see this when I update a rule and a user already has jobs running that >> exceed the spec. Any chance the slots were consumed before the rule went >> into affect? > > Hi Michael, > > thanks for the suggestion, but this rule has been in place for a long time > and seems to have worked previously. > > My suspicion is that the sge master daemon crashed at some point and this is > connected with the issue above
I can't remember what I said independent, but it might be related to the crashes, which are presumably due to the tight integration regression (articles passim). I can't remember what actually happens, and whether it might corrupt things. There appears to be something wrong anyway, because I sometimes see negative slot counts listed by qrstat. -- Community Grid Engine: http://arc.liv.ac.uk/SGE/ _______________________________________________ users mailing list [email protected] https://gridengine.org/mailman/listinfo/users
