[GE users] qrsh/qlogin woes

brs brs at usf.edu
Wed Jan 6 23:19:16 GMT 2010


Wow.  Thanks for helping get the cogs working again!  My exec nodes were
resolving the external interface of my login nodes and not the internal
interface.  For some reason, I had removed those entries from the NIS
maps.  Problem solved!

Thanks,
-Brian 

On Wed, 2010-01-06 at 22:23 +0100, reuti wrote:
> Am 06.01.2010 um 21:56 schrieb brs:
> 
> > Firewall on qmaster host has this rule (also, I've tested by disabling
> > altogether)...
> >
> > -A INPUT -s 10.250.0.0/21 -d 0/0 -j ACCEPT
> 
> Is the qmaster machine the one you also login in? The qrsh will make  
> a direct connection between the login machine and the exechost. Are  
> both in the same net, or was there any IP forwarding defined before?
> 
> 
> > All compute nodes are on 10.250.0.0/21 and have no iptables rules.
> > Nodes have no local configuration.
> >
> > qconf -sconf shows
> > ...
> > qlogin_command               builtin
> > qlogin_daemon                builtin
> > rlogin_daemon                builtin
> > rsh_daemon                   builtin
> > rsh_command                  builtin
> > rlogin_command               builtin
> > ...
> >
> > Also, we're not using NFS for SGE at all.  Binaries/scripts for submit
> > hosts and exec hosts live on an NFS mount, but spool directories are
> > local (/var/spool/sge) and QMaster is a VM with its own copy of
> > binaries/scripts, no NFS exports or mounts (same version, FYI).
> >
> >> The most likely reason you didn't get a response is that it's the
> >> holiday season and most folks are on vacation through this week.
> >> Check back next week. :)
> >
> > I think my confusion comes from the fact that I no longer seem to get
> > copies of my own sent messages to any of my mailing lists including  
> > this
> > one (so it naturally arouses my fear that the message never got
> > out)! :-O  And besides, far be it from me to try and rain on anyone's
> > post-holiday carousing with my petty issues!  We should all be  
> > curled up
> > by a fire drinking Samichlaus,
> 
> Isn't Samichlaus the name for December the 6th? But at least in  
> certain regions of Germany even today is a public holiday to  
> celebrate the Twelfth Day. But not for me in the county of Hesse.
> 
> -- Reuti
> 
> 
> > not looking at cryptic trace outputs for
> > users who don't know how to take time off :)
> >
> > Thanks for the efforts!
> >
> > -Brian
> >
> > -- 
> > Brian Smith
> > Senior Systems Administrator
> > IT Research Computing, University of South Florida
> > 4202 E. Fowler Ave. ENB308
> > Office Phone: +1 813 974-1467
> > Organization URL: http://rc.usf.edu
> >
> >
> > On Wed, 2010-01-06 at 21:41 +0100, reuti wrote:
> >> Am 06.01.2010 um 21:08 schrieb templedf:
> >>
> >>> The most likely reason you didn't get a response is that it's the
> >>> holiday season and most folks are on vacation through this week.
> >>> Check
> >>> back next week. :)
> >>
> >> Well, I'm more confused by this amount of information :-O
> >>
> >> First the basic stuff to be checked. Which machine was changed: the
> >> source machine or the target machine? Any local configuration for the
> >> target machine? Any firewall in place on the target machine?
> >>
> >> -- Reuti
> >>
> >>
> >>> Daniel
> >>>
> >>> brs wrote:
> >>>> My apologies if this is a duplicate... I didn't see it show up in
> >>>> any of the threads.  Should I send to ge_dev instead?
> >>>>
> >>>> ---
> >>>>
> >>>> I've tried to debug this issue, but before I go any further, I just
> >>>> wanted another set of eyes to take a look, in case its something
> >>>> obvious.  I imported my configuration from a backup of a 6.2u3
> >>>> installation.  I'm now on 6.2u4.  This worked perfectly on  
> >>>> 6.2u3.  I
> >>>> moved the installation from a Solaris box to a Linux box.   
> >>>> That's the
> >>>> only real difference I can see.  Let me know what you think!
> >>>>
> >>>> Attached is the output of various qrsh attempts with various
> >>>> levels of
> >>>> verbosity and trace output.
> >>>>
> >>>> Thanks in advance,
> >>>> -Brian
> >>>>
> >>>>
> >>>>
> >>>
> >>> ------------------------------------------------------
> >>> http://gridengine.sunsource.net/ds/viewMessage.do?
> >>> dsForumId=38&dsMessageId=236905
> >>>
> >>> To unsubscribe from this discussion, e-mail: [users-
> >>> unsubscribe at gridengine.sunsource.net].
> >>
> >> ------------------------------------------------------
> >> http://gridengine.sunsource.net/ds/viewMessage.do? 
> >> dsForumId=38&dsMessageId=236911
> >>
> >> To unsubscribe from this discussion, e-mail: [users- 
> >> unsubscribe at gridengine.sunsource.net].
> >
> > ------------------------------------------------------
> > http://gridengine.sunsource.net/ds/viewMessage.do? 
> > dsForumId=38&dsMessageId=236913
> >
> > To unsubscribe from this discussion, e-mail: [users- 
> > unsubscribe at gridengine.sunsource.net].
> 
> ------------------------------------------------------
> http://gridengine.sunsource.net/ds/viewMessage.do?dsForumId=38&dsMessageId=236922
> 
> To unsubscribe from this discussion, e-mail: [users-unsubscribe at gridengine.sunsource.net].

------------------------------------------------------
http://gridengine.sunsource.net/ds/viewMessage.do?dsForumId=38&dsMessageId=236947

To unsubscribe from this discussion, e-mail: [users-unsubscribe at gridengine.sunsource.net].



More information about the gridengine-users mailing list