Hi Reuti, The entire cluster is managed by Rocks and /etc/hosts is handled by Rocks entirely as well. No one has modified /etc/hosts on the qmaster manually. I can try remove omega-0-12 from the system (so Rocks will remove the entry in /etc/hosts) and see what happens.
Cheers, D On Tue, Sep 8, 2015 at 7:19 PM, Reuti <[email protected]> wrote: > Hi, > > > Am 08.09.2015 um 09:23 schrieb Derrick Lin <[email protected]>: > > > > Hi guys, > > > > Thanks for the helps. I ran the SGE tools on the qmaster, and found the > issue: > > > > [root@alpha01 lx26-amd64]# ./gethostname > > Hostname: omega-0-12.local > > Aliases: omega-0-12 > > Host Address(es): 192.168.11.12 > > in /etc/hosts or any additional hostname resolution like NIS. There > shouldn't be any additional entry for a loopback interface except for the > usual 127.0.0.1. Some Linux distributions add there an additional interface > to allow to contact the localhost even under his external name when the > machine is not connected to a network. > > -- Reuti > > > > > > > Somehow the qmaster "thinks" itself as omega-0-12. I couldn't recall I > have made changes in the qmaster recently. > > > > Where I should be looking at to fix this issue? > > > > Regards, > > Derrick > > > > On Mon, Sep 7, 2015 at 3:04 PM, Reuti <[email protected]> > wrote: > > > > Am 07.09.2015 um 00:36 schrieb Derrick Lin: > > > > > Hi Simon, > > > > > > It looks normal: > > > > > > [root@alpha01 ~]# nslookup alpha01.local > > > Server: 127.0.0.1 > > > Address: 127.0.0.1#53 > > > > > > Name: alpha01.local > > > Address: 192.168.11.200 > > > > There are some tools `gethostbyname` resp. `gethostbyaddr` in > $SGE_ROOT/utilbin/$ARC to check what SGE sees. > > > > -- Reuti > > > > > > > > > > All nodes are configured based on the same image via cluster > management tool. > > > > > > Cheers, > > > D > > > > > > On Fri, Sep 4, 2015 at 12:16 PM, Simon Matthews < > [email protected]> wrote: > > > What does the rDNS show for the IP address of alpha01.local? > > > > > > Simon > > > > > > On Thu, Sep 3, 2015 at 6:44 PM, Derrick Lin <[email protected]> wrote: > > > > Dear all, > > > > > > > > I have been having issue on executing all SGE commands on the > qmaster, > > > > typically, it gives such error: > > > > > > > > [root@alpha01 ~]# qconf -sc > > > > error: commlib error: access denied (client IP resolved to host name > > > > "alpha01.local". This is not identical to clients host name > > > > "omega-0-12.local") > > > > > > > > DNS is working fine, as alpha01 and omega-0-12 both can be resolved > > > > correctly. > > > > > > > > The issue happens on the qmaster ONLY, the rest of the cluster nodes > can > > > > execute the same command fine. > > > > > > > > Any idea will be much appreciated. > > > > > > > > Cheers, > > > > Derrick > > > > > > > > _______________________________________________ > > > > users mailing list > > > > [email protected] > > > > https://gridengine.org/mailman/listinfo/users > > > > > > > > > > _______________________________________________ > > > users mailing list > > > [email protected] > > > https://gridengine.org/mailman/listinfo/users > > > > > >
_______________________________________________ users mailing list [email protected] https://gridengine.org/mailman/listinfo/users
