[GE users] SunGrid Engine & BLAST

Hrishikesh Deshmukh hdeshmuk at gmail.com
Mon Jan 17 16:09:12 GMT 2005


I think i have now figured out few more things, the problem was of
permissions though /usr/local/sge was owned by sgeadmin, the dirs
inside were owned by root!(since remote dektop was used by root to
copy them to exec. hosts).

So now sgeadmin owns /usr/local/sge and sub dirs in it.
So when i run the ./install_execd again,i get the following error:
Checking cell directory
-----------------------

Cannot find required files. The following error was returned:

>common directory not found: default/common<


Please make sure that you have installed the qmaster host before
installing an execution host.

The installation procedure will only work if you use a shared
directory (e.g. shared with NFS) for your installation.

The installation of the execution daemon will abort now.

So how does one make NFS mount/some shared dir on the i guess master host!

Thanks,
Hrishi


On Fri, 14 Jan 2005 20:05:33 -0500, Hrishikesh Deshmukh
<hdeshmuk at gmail.com> wrote:
> Could this be a problem,sgeadmin is sudo.
> I downloaded the installation files on a machine and made the machine
> the master by running the ./install_qmaster, then asked the SYS ADMIN
> of the dept. to copy the installation dirs and files to all the
> machines,its easy to do this using remote desktop.
> Could this be causing some kind of permissions!! conflict, i think
> sgeadmin owns the install dirs and files.
> 
> Just guessing about the critical error:resolving problems origin!!!
> 
> Thanks,
> Hrishi
> 
> On Fri, 14 Jan 2005 14:56:18 -0500, Hrishikesh Deshmukh
> <hdeshmuk at gmail.com> wrote:
> > I kind of guess that! I really appreciate this help.
> > I am installing sge as sgeadmind (sudo).
> > I did define port 28 in /etc/services on master and execution host.
> > I did sudo ./install_execd
> >
> > Hrishi
> >
> > On Fri, 14 Jan 2005 11:48:34 PST, Rayson Ho <raysonho at eseenet.com> wrote:
> > > Just I came back from lunch :)
> > >
> > > Thanks for comparing the output!!
> > >
> > > So, commd on the exec hosts are not running, and since they need to bind to
> > > ports below 1024, did you start execd as root??
> > >
> > > Also, "printenv COMMD_PORT" did not show anything, did you define the port
> > > number for it as 28 in /etc/services??
> > >
> > > Rayson
> > >
> > >
> > > >Here are the DIFFERENCES in the outputs for the cmd: netstat -an |
> > > >grep 28
> > > >on master host:
> > > > ...
> > > >tcp4       0      0  *.28                   *.*      LISTEN
> > > >
> > > >on (potential) executions host:
> > > >udp4       0      0  *.3283                 *.*
> > > > 1104930 dgram       0      0        0  1104fc0        0  1104b28
> > > > 1104b28 dgram       0      0        0  1104fc0        0  1104b60
> > >
> > >
> > > ---------------------------------------------------------
> > > Get your FREE E-mail account at http://www.eseenet.com !
> > >
> > > ---------------------------------------------------------------------
> > > To unsubscribe, e-mail: users-unsubscribe at gridengine.sunsource.net
> > > For additional commands, e-mail: users-help at gridengine.sunsource.net
> > >
> > >
> >
>

---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe at gridengine.sunsource.net
For additional commands, e-mail: users-help at gridengine.sunsource.net




More information about the gridengine-users mailing list