[Mageia-sysadm] [1827] add small maintdb script

Sat Jul 30 02:36:34 CEST 2011

Op zaterdag 30 juli 2011 01:52:23 schreef nicolas vigier:
+> On Sat, 30 Jul 2011, Maarten Vanraes wrote:
+> > Op vrijdag 29 juli 2011 17:11:00 schreef Michael Scherer:
+> > > Le vendredi 29 juillet 2011 à 16:55 +0200, nicolas vigier a écrit :
+> > > > On Sat, 16 Jul 2011, Michael Scherer wrote:
+> > > > > > +function maintgetall()
+> > > > > > +{
+> > > > > > +    cd "$MAINTDBDIR"
+> > > > > > +    for file in *
+> > > > > 
+> > > > > Wouldn't this one have problem the day we have too much maintainers
+> > > > > ?
+> > > > 
+> > > > too much packages ?
+> > > 
+> > > Mhh, yes :)
+> > > 
+> > > > > ( even if "too much" is, according to my test, more than 60000 so
+> > > > > this is just a retorical question )
+> > > > 
+> > > > According to tune2fs we have 1148096 free inodes, so that should be
+> > > > enough. I don't know the limit for the "*" expansion.
+> > > 
+> > > There used to be one, but maybe it was just in a time when kernel 2.4
+> > > was hype :)
+> > > ( or maybe that was on a bsd or os x ).
+> > 
+> > iirc * expansion is shell dependant and was 255 in the early days and may
+> > have been 16383, but i think it's 65535 now for a while... i know
+> > certain 2.6 kernels do have a limit, because i ran into it once (don't
+> > ask)
+> 
+> You're probably talking about the limit in the size of arguments for a
+> new process. But "for" is not a process, so I'm not sure there is a
+> limit.
+
+afaik it's internal handling; it was exactly a for that gave me my problem 
+before.
+