[Samba] Winbind, cached logons and 'user persistency'...
rpenny at samba.org
Fri Jan 18 09:59:30 UTC 2019
On Fri, 18 Jan 2019 10:41:10 +0100
"L.P.H. van Belle via samba" <samba at lists.samba.org> wrote:
> Hai Marco,
> > -----Oorspronkelijk bericht-----
> > Van: samba [mailto:samba-bounces at lists.samba.org] Namens
> > Marco Gaiarin via samba
> > Verzonden: vrijdag 18 januari 2019 10:03
> > Aan: samba at lists.samba.org
> > Onderwerp: Re: [Samba] Winbind, cached logons and 'user
> > persistency'...
> > Mandi! L.P.H. van Belle via samba
> > In chel di` si favelave...
> > > Maybe the winbind cache time is set to low for this.
> > OK. But this look still strange/dangerous to me. Two 'open point':
> > 1) seems to me that there's many 'cache time' parameters:
> > + idmap cache time, default 604800 (one week); seems related only
> > to SID<->GID/UID query, so unrelated here.
> > + winbind cache time, default 300 (5 minutes); this seems the
> > parameter i need to tackle with.
> > but... HOW work that cache? There's a 'negative' timeout also? Or
> > simply cache data and use cached data if all DC are not available?
> Poe, this i dont know, i dont know all code...
> Rowland, you know this?
No, I have never had to mess with this, but 'man smb.conf' says this:
This parameter specifies the number of seconds the winbindd(8)
daemon will cache user and group information before querying a
Windows NT server again.
It looks like you reduce the time to make the cache refresh more often
and increase it to make the cache last longer, I would presume setting
it to '0' would make winbind query the server without using the cache,
but this is just a guess.
More information about the samba