-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 06/08/13 22:50, Xiao Peng Wang wrote:
> From your previous information, the correct ssh configuration files
> have been copied to /opt/intel/mic/filesystem/mic0 on the host node.
> Please check the /etc/sysconfig/mic/mic0.conf has foll
Also I meant to ask is otherpkgs not required for the genimage
process? Does genimage run otherpkgs without requiring the script be
listed as a postscript?
On 8/6/2013 7:21 PM, Russell Jones
wrote:
Okay, I will get that to you on Friday
Okay, I will get that to you on Friday - the strange thing is I was
unable to replicate the issue a second time If I can replicate
it, I will provide you with that file.
This is 2.8.2.
On 8/6/2013 7:02 PM, Lissa Valletta
wrote:
If you could give us a copy of /xcatpost/mypostscript file on the node,
I would appreciate it. There is a check in otherpkgs that should stop it
from running and I want to see why it is failing.
what level of xCAT are you running
if [ -z "$UPDATENODE" ] || [ $UPDATENODE -ne 1 ]; then
if [
We have noted this as an issue. What you need to do is take otherpkgs out
of the xcatdefaults line in the postscripts table.
"xcatdefaults","syslog,remoteshell,syncfiles","otherpkgs",,
If you have nodes or nodegroups that do need otherpkgs, then explictly
define those groups in the postscripts
Hi all,
I've converted a non-osimage stateless image over to using osimage. I've
noticed now that when booting stateless nodes, the otherpkgs script
executes and attempts to install the extra RPMs. In the past, otherpkgs
didn't run as this was performed during the genimage process.
Does the ot
Glad you are fixed. You can ignore my second response.
Lissa K. Valletta
8-3/B10
Poughkeepsie, NY 12601
(tie 293) 433-3102
From: Jack Singh
To: "xcat-user@lists.sourceforge.net"
,
Date: 08/06/2013 02:57 PM
Subject:Re: [xcat-user] ssh keys corruption
Lissa
Xdsh
Did you check that the password you input it correct.
ssh hpcnode02date
It should prompt for the password, enter it and see if the ssh completes
If that works, then just try the updatenode hpcnode02 -k and see if that
does the trick.
You said one of the users ran ssh-keygen . They were ru
Lissa
Xdsh command told me permission denied, so I modified permissions on the node
and ran updatenode for the noderange
That fixed the problem.
Now password less logins are working fine and xdsh commands are running fine.
So I did do the 'updatenode' command as per the man page.
Thanks
Jack
-
Lissa,
here is required data, please see:
lsxcatd -a
Version 2.8.2 (svn r17010, built Mon Jul 22 23:28:24 EDT 2013)
This is a Management Node
dbengine=SQLite
tried different profiles, always the same result. HW looks ok, because i
can install rhel from DVD.
lsdef n9
Object name: n9
arch=x86_
I followed your procedure but it didn't work, here is the output
root@hpcmgmt:~# xcatconfig -k
Removed id_rsa files from /root/.ssh directory.
Generated /root/.ssh/id_rsa.pub.
Copied /root/.ssh/id_rsa.pub to /install/postscripts/_ssh/authorized_keys.
root@hpcmgmt:~# xcatconfig -s
Generating new n
Lissa
I figured it out, it was ownership and permission issues on .ssh folder.
Thanks for your help.
Jack
--
Get 100% visibility into Java/.NET code with AppDynamics Lite!
It's a free troubleshooting tool designed for produ
I failed to mention the distribution of the hostkeys. So after xdsh -K and
you can xdsh with out any prompting for a password. Then run the following
to distribute the hostkeys. These are sort of optional, but keep you
from being warned about automatically adding entiries to the known_hosts
f
What you want to do is on the Management node run xcatconfig -k and then
xcatconfig -s. This will regenerate the root ssh keys and the ssh common
hostkeys for the nodes and put it in the right places for xCAT.
Then run xdsh -K to your nodes.
It will prompt you for the root password, but af
I have xcat 2.8.2 server running on Ubuntu 12.04 server.
My nodes are Ubuntu12.02 also
One of the users regenerated ssh keys on the management server and on the nodes
using the two following commands
ssh-keygen -t dsa
ssh-keygen -t rsa
Now the xdsh , pscp commands wont work.
So I copied the keys
What level of xCAT is your Management node. (lscatd -a ) What is the OS
on your Management Node. Give us an lsdef of the node you are trying to
install.
Lissa K. Valletta
8-3/B10
Poughkeepsie, NY 12601
(tie 293) 433-3102
From: indi
To: xcat-user@lists.sourceforge.net,
Date: 08/0
Hi ALL,
can you please advice how to solve problem.
trying to install RHEL 6.4 (or 5.9) on IBM HS21 blades. Getting error:
Trying to load: pxelinux.cfg/AC146509 ok
APE I: Can not request [mem 0x0009d308-0x0009d312] for APEI ERST registers
init[1]: segfault at 0 ip 00401f9f sp 7M044c8
Hi there,
On 06/08/13 22:44, Xiao Peng Wang wrote:
> Two things you need to check:
It's worth keeping in mind that this configuration worked under xCAT
2.6.9 and RHEL5 (the switches and wiring hasn't changed), but I will
double check tomorrow.
Thanks!
Chris
--
Christopher SamuelSenio
Chris,
From your previous information, the correct ssh configuration files have
been copied to /opt/intel/mic/filesystem/mic0 on the host node. Please
check the /etc/sysconfig/mic/mic0.conf has following lines, if it has,
it should be an issue of system. Could you check that you have e
Chris,
Two things you need to check:
1. make sure the nodes which switch and port that calculated from
regular expression have not duplicated ones. Run the lsdef node to
check.
2. for the switch+port which has multiple nodes discovered, you need to
make sure whether there are mu
Hi Lissa,
On 06/08/13 20:56, Lissa Valletta wrote:
> I am assuming the below is from the lsdef command.
That's correct.
> If you notice the regular expression for the switchport is not expanding.
That's to be expected, that's a group that the nodes belong to to
define the mappings. A node l
I am assuming the below is from the lsdef command. If you notice the
regular expression for the switchport is not expanding. Under lsdef it
should expand to the correct value. Check your regular expression.
sw02a:
objtype=group
grouptype=static
members=merri001,merri002,merri003,mer
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Hi all,
We've just reinstalled a cluster that was running RHEL5 with RHEL6 and
gone from xCAT 2.6.9 to 2.8.2.
Reusing the same switch definitions that worked for discovery in xCAT
2.6.9 (and adding the nodes from scratch) we find that we are getting
23 matches
Mail list logo
