>Number: 172841 >Category: kern >Synopsis: ZFS NFS exports => 1024 byte NIS-limits for netgroups >Confidential: no >Severity: non-critical >Priority: low >Responsible: freebsd-bugs >State: open >Quarter: >Keywords: >Date-Required: >Class: sw-bug >Submitter-Id: current-users >Arrival-Date: Thu Oct 18 05:50:00 UTC 2012 >Closed-Date: >Last-Modified: >Originator: Johannes Reinhard >Release: FreeBSD 9.1-RC2 >Organization: FAU Erlangen-Nürnberg >Environment: FreeBSD fileserv 9.1-RC2 FreeBSD 9.1-RC2 #0 r241106: Mon Oct 1 18:26:44 UTC 2012 [email protected]:/usr/obj/usr/src/sys/GENERIC amd64 >Description: Using NFS exports from ZFS to a netgroup, netgroups are truncated to the traditional NIS entry limit of 1024 bytes, while they shouldn't. This happens regardless of the netgroup source (both settings files and ldap have been tried in nsswitch.conf). Also, if nscd is running or not does not seem to have any influence.
Here is example output how hostname for blarz.domain.de got truncated. # cat /var/log/messages .. Oct 18 07:17:22 fileserv mountd[7684]: can't get address info for host blarz.domai Oct 18 07:17:22 fileserv mountd[7684]: bad host blarz.domai in netgroup mynetgroup, skipping .. Another way to see the effects is to call # showmount -e >How-To-Repeat: 1. Create a zfs volume 2. Create a large (>1024 byte) netgroup entry in /etc/netgroup - it must contain valid hosts 3. Configure /etc/nsswitch.conf to use /etc/netgroup as a netgroup source netgroup: files 4. Configure zfs NFS export using a netgroup # zfs set sharenfs="-alldirs mynetgroup" pool 5. read /var/log/messages and look for the message above # less /var/log/messages 6. use showmount -e to see where the netgroup got truncated >Fix: Create multiple smaller netgroups and export to all of them. >Release-Note: >Audit-Trail: >Unformatted:
_______________________________________________ [email protected] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-bugs To unsubscribe, send any mail to "[email protected]"
