On Fri, 2009-11-20 at 11:19 +0100, Sumit Bose wrote: > On Fri, Nov 20, 2009 at 10:40:29AM +0100, Sumit Bose wrote: > > On Wed, Nov 18, 2009 at 12:09:58PM -0500, Simo Sorce wrote: > > > This way we check them once at storage time instead of checking again > > > and again at search time. > > > > > > Applies only on top of the sysdb_enumgrent optimization patch. > > > > > > Simo. > > > > > > > Works well, especially with Optimize-sysdb_enumgrent.patch, but > > please create a utility function or a macro for the range check and fix > > > > responder/nss/nsssrv_cmd.c: In function 'fill_grent': > > responder/nss/nsssrv_cmd.c:1476: warning: unused variable 'uid' > > > > bye, > > Sumit > > > > Can you add a task that removes entries outside of the range at startup?
Maybe later, if you are enlarging the range it is not needed, and if you are *really* changing the range I suggest you stop sssd and remove the cache file. It's extremely rare you are going to change the ID range to exclude existing cached IDs, and if that happens the admin has to go around and probably change ownership and remove files altogether. So he can as well go in total maintenance mode and remove the cache file. Simo. -- Simo Sorce * Red Hat, Inc * New York _______________________________________________ sssd-devel mailing list sssd-devel@lists.fedorahosted.org https://fedorahosted.org/mailman/listinfo/sssd-devel