[xcat-user] IBM x3455 XCAL cluster

[Vallard Benincosa]

Hi Antonis,
1.  Your $XCATROOT/etc/conserver.cf file is missing the last lines to
allow you to connect to it:

The file should look like this:
LOGDIR=/var/log/consoles
#
# list of consoles we serve
#    name : tty[@host] : baud[parity] : logfile : mark-interval[m|h|d]
#    name : !host : port : logfile : mark-interval[m|h|d]
#    name : |command : : logfile : mark-interval[m|h|d]
#

node01:|sol.ipmi2 node01::&:
node02:|sol.ipmi2 node02::&:
node03:|sol.ipmi2 node03::&:
node04:|sol.ipmi2 node04::&:
node05:|sol.ipmi2 node05::&:
node06:|sol.ipmi2 node06::&:
node07:|sol.ipmi2 node07::&:
node08:|sol.ipmi2 node08::&:
node09:|sol.ipmi2 node09::&:
node10:|sol.ipmi2 node10::&:
node11:|sol.ipmi2 node11::&:
node12:|sol.ipmi2 node12::&:
%%
trusted: 127.0.0.1, masternode

Note the last line and the %% must be included.  After you change your
file to be like that run:
service conserver restart
Then check your /var/log/messages to make sure that it started without
any problems.  You should be able to then connect to it with rcons or
wcons.  If it doesn't work, please let us know the message you
receive.

2.  With respect to your eth0/eth1 problem, your noderes.tab file
seems to have some errors:  It should read:
compute masternode,masternode,/install,0,N,N,N,N,N,N,N,eth0,eth0,masternode
all          masternode,masternode,/install,0,N,N,N,N,N,N,N,eth0,eth0,masternode

The way you currently had it set:
compute	masternode,masternode,/install,NA,N,Y,N,N,N,Y,N,eth0,eth1,masternode
Translates to:
1.  (The NA) Nodes don't use serial console redirection
2.  (The first Y) Nodes are install servers so they copy every RPM to
the compute nodes.  You probably don't want to do that.
3.  (The last Y) Nodes are setup so that people can't login to them by
default.  We usually set this to Y if we're using Torque and Maui.
This way users can't log into nodes unless they are running a job
through the queuing system.
4.  (The eth1) sets eth1 as your primary NIC.  That's why eth0 is not
getting configured.

Hope that helps,
Vallard





On Feb 3, 2008 11:07 PM, Antonis A. Constantinou
<a.constantinou at newcytech.com> wrote:
>
>
>
>
> Dear all
>
>
>
> Thanks for all your help regarding the setup of my cluster. Yesterday I was
> able to finally perform the network installation on the 12 nodes. I even
> setup IPMI and tested with some notes using the rpower command to remotely
> turn on the notes. The conserver setup however for does not seem to work. I
> attach my latest configuration files and I would be gratefull if you can
> have a look and point me to the right direction to make SOL work. I have set
> the bios setting as described in the link you provided for me. Furthermore
> one issue I have is that after the installation has finished and the nodes
> boot up eth0 is not active for some reason. Eth1 on each nodded gets its ip
> from the DHCP on the masternode but eth0 which is the one associated with
> the node hostname seems to deactivated. Shall I manually configure it o each
> node or I am missing something here.
>
>
>
> Thanks gain for all your help.
>
> Best Regards
>
> Antonis Constantinou
>
> Systems Engineer
>  NewCytech Business Solutions Ltd
>  Tel: +357 22470888
>  Fax: +357 22494254
>  E-mail:a.constantinou at newcytech.com
>  NewCytech is one of the largest business technology companies in Cyprus,
> offering innovative products and versatile services to clients in a broad
> range of business. We offer an array of solutions that consists of products
> and services. Our products feature both hardware and software.
>
> This email is confidential and intended solely for the use of the individual
> to whom it is addressed. Any views or opinions presented are solely those of
> the author and do not necessarily represent those of the NewCytech Business
> Solutions Ltd. If you are not the intended recipient, be advised that you
> have received this email in error and that any use, dissemination,
> forwarding, printing, or copying of this email is strictly prohibited. If
> you have received this email in error please notify us immediately by
> telephone on +357 22 470600 or reply to this message.
>
>
>
>
>
>
> _______________________________________________
> xcat-user mailing list
> xcat-user at lists.xcat.org
> http://www.xcat.org/mailman/listinfo/xcat-user
>
>