> > On 05.04 17:49, Joost de Heer wrote:
> >> Problem was indeed (as Marc Elsen pointed out) a filesystem problem with
> >> Solaris. I've solved it by running newfs (byebye cache...) and after
> >> that
> >> tunefs. I should've checked the FAQ better.... it just was in a place
> >> where I didn't look ('System-dependant weirdness'->Solaris).
> > Pardon me, but: did you make sure that it was Solaris problem BEFORE you
> > removed the cache, or just after it?
On 06.04 21:36, Joost de Heer wrote:
> After Marc pointed me to the FAQ, I read the thing, saw in
> /var/adm/messages the exact errors given in the FAQ, and thus concluded it
> was this problem. I first tried just tunefs, with no result, so after that
> I did a newfs, tunefs, and then recreated the cache dirs with 'squid -z'.
> It's happily serving requests again (and happily serving happy users in
> the US who won't need to use a cache in the Netherlands anymore...)
For now it's OK. But note that it could also be caused by running out of
inodes - if you did not check for this, you really don't know if it was
actually this problem. The FAQ entry explicitly says it happens if you are
NOT out of inodes.
Cleaning whole cache usually helps problem not to appear, but might not to
solve it. For now it's O.K., but next time check free inodes too.
For increasing numbers of inoeds you do need to re-create filesystem, but
insteaad of playing with tunefs you have to specify less bytes-per-inode
as newfs option.
--
Matus UHLAR - fantomas, uhlar@xxxxxxxxxxx ; http://www.fantomas.sk/
Warning: I wish NOT to receive e-mail advertising to this address.
Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu.
(R)etry, (A)bort, (C)ancer
