On 12/28/2007 08:52 PM, Nick Kew wrote: > On Fri, 28 Dec 2007 19:31:31 -0000 > [EMAIL PROTECTED] wrote: > >> @@ -140,11 +143,15 @@ >> http://svn.apache.org/viewcvs.cgi?rev=607245&view=rev >> Backport version for 2.2.x of patch: >> Trunk version of patch works >> - +1: rpluem, niq >> + +1: rpluem, niq, wrowe >> niq: Provisional +1, but the error logging should be at a >> consistent level (maybe WARNING?) >> rpluem: Set it to ERROR in all cases as IMHO this should not >> happen. If this level is too high we can reduce it later. >> + wrowe: disagree with rpluem - it's incredibly disruptive to >> admins >> + to have their logs filled with noise - warning would be >> ok, >> + provided there's no more than one entry per failed >> request. >> + If their request would die outright, only then is >> rpluem right. > > +1 to wrowe's comment. I was thinking the same, just not loudly > enough to be the first to say so.
Ok, I am also fine with the WARNING level here. OTOH ERROR is consistent with the other behaviour of mod_disk_cache where all other non fatal (= do no cause the request to die), non DEBUG message also result in messages at level ERROR. So maybe we should rework the error levels of all messages in mod_disk_cache. Furthermore I do not really expect these messages to appear often. Bill, is your vote conditional on changing this to the WARNING level or can this be back ported as is? Regards Rüdiger
