Re: Xcat image capture issues

2012-03-16 Thread Dmitri Chebotarov
Evelio

What type of hardware d6100a is? If it doesn't have IPMI enabled motherboard or 
IPMI card, r* commands may not work. 
By the name I would guess it's Latitude d610...

Try rebooting manually just to test xCAT/PXE.

Thanks.

On Mar 15, 2012, at 12:03 , Evelio Quiros wrote:

 Hello,
 
 Yes, both listener.py and rootimg.gz are both located at
 /install/image/x86/installer_files/
 
 
 But NOWHERE on the documentation page does it mention a symlink between
 x86_64 and x86.
 
 Ok, I added the symlinks in both locations, and it did seem to accept it:
 
 nodeset d6100a image
 d6100a: image image-x86_64-rh5image-rh66-v0
 
 Which created this file in /tftpboot/pxelinux.cfg:
 
 
 lrwxrwxrwx 1 root root   6 Mar 15 11:36 0A19 - d6100a
 
 cat d6100a 
 #image image-x86_64-rh5image-rh66-v0
 DEFAULT xCAT
 LABEL xCAT
 KERNEL xcat/image/x86_64/vmlinuz
 APPEND initrd=xcat/image/x86_64/initrd.img
 imgurl=http://10.0.0.3//install/image/x86_64/installer_files/rootimg.gz
 image=/install/image/x86_64/rh5image-rh66-v0.img blocks=512 action=save
 installnic=eth0 reboot  noipv6
  IPAPPEND 2
 
 I did a tabedit to add the nodetype.supportedarchs.
 
 
 tabdump nodetype
 #node,os,arch,profile,provmethod,supportedarchs,nodetype,comments,disable
 compute,rh6.0,x86_64,compute,,,osi,,
 d6100a,image,x86,rh5image-rh66-v0,image,x86,x86_64,,,
 
 
 
 However, when I do this:
 rpower d6100a reset
 
 I get this:
 d6100a: Error: timeout
 
 So, I guess I must be missing something that tells xCAT where the machine
 is, and how to reach it.
 
 
 Al Q
 
 
 On 3/15/12 11:05 AM, Dmitri Chebotarov dcheb...@gmu.edu wrote:
 
 Evelio
 
 Could you please check that you have these files/folders as well?
 
 /install/image/x86/installer_files/listener.py
 /install/image/x86/installer_files/rootimg.gz
 
 Also, can you check that there are symlinks for x86 or x86_64 in
 /install/image/ and /tftpboot/xcat/image/, ie:
 
 $ ls -l
 x86
 x86_64 - x86
 
 Thanks.
 
 On Mar 15, 2012, at 9:56 , Evelio Quiros wrote:
 
 Thanks Dmitri,
 
 In following this article, I get to the bottom of the page, where I put
 in
 the 3 files needed for a stateless image.
 
 I place the 3 files in the located indicated, and when I run
 
 nodeset d6100a image
 
 I get this:
 
 Error: vmlinuz and initrd.img not found at /tftpboot/xcat/image/x86_64
 Error: Some nodes failed to set up image resources, aborting
 
 But vmlinuz and initrd.img are at /tftpboot/xcat/image/x86_64
 
 
 Is there something else that needs to happen first ?
 
 
 Also, the last step talks about NFS servers for the image store.
 Are these required ?
 
 Thanks for all your help,
 Al Quiros
 
 
 
 On 3/14/12 4:05 PM, Dmitri Chebotarov 4dim...@gmail.com wrote:
 
 Evelio 
 
 Please follow this article to add support for 'image' and 'install' :
 
 
 https://cwiki.apache.org/VCL/adding-support-for-partimage-and-partimage-
 ng
 -to-xcat-2x-unofficial.html
 
 Thanks. 
 
 -- 
 Dmitri Chebotarov
 
 
 On Wednesday, March 14, 2012 at 15:37 , Evelio Quiros wrote:
 
 Hello again,
 
 Dimitri, thanks for the advice.
 
 I have a question about these commands.
 Below, you wrote nodeset d6100a image,
 
 But there is no image argument for nodeset.
 Here is the man page:
 
 nodeset [-h|--help|-v|--version]
 nodeset noderange
 
 
 [install|shell|boot|runcmd=bmcsetup|netboot|iscsiboot|osimage=imagenam
 e
 |s
 tatelite]
 
 When I try the command as-is, I get this:
 
 
 nodeset d6100a image
 d6100a: Error: Unable to identify plugin for this command, check
 relevant
 tables: nodetype.os
 Error: Some nodes failed to set up image resources, aborting
 
 I have done tabdump on the various tables, and done tabedit to modify
 them
 to my environment.
 
 
 I know this is not an xCAT list, but your guidance would be
 appreciated.
 
 Thanks,
 Al Quiros
 
 
 
 On 3/13/12 4:45 PM, Dmitri Chebotarov dcheb...@gmu.edu
 (mailto:dcheb...@gmu.edu) wrote:
 
 Evelio,
 
 You should be able to capture image via xCAT before you can use
 VCL/xCAT.
 
 To capture image via xCAT:
 
 nodech d6100a nodetype.os=image nodetype.profile=rh5image-rh66-v0
 nodeset d6100a image - this will create PXEboot config file for
 d6100a:
 
 cat /tftpboot/pxelinux.cfg/d6100a-PXE-if-mac
 
 #image rh5image-rh66-v0
 DEFAULT xCAT
 LABEL xCAT
 KERNEL xcat/image/x86_64/vmlinuz
 APPEND initrd=xcat/image/x86_64/initrd.img
 
 
 imgurl=http://xcat-master/install/image/x86_64/installer_files/rootimg.
 gz
 image=/install/image/x86_64/rh5image-rh66-v0.img blocks=512
 action=save installnic=eth0 reboot noipv6
 IPAPPEND 2
 
 'tabdump nodetype' should look similar to this - note that arch is
 x86:
 
 
 
 #node,os,arch,profile,provmethod,supportedarchs,nodetype,comments,disab
 le
 ...
 d6100a,image,x86,rh5image-rh66-v0,image,x86,x86_64,,,
 ...
 
 Then do 'rpower d6100a reset'. This will restart d6100a and when node
 boots over PXE it will boot into
 http://xcat-master/install/image/x86_64/installer_files/rootimg.gz
 and,
 given that image works fine and maps NFS store OK, it dump hd image
 to
 

Re: Creating a base image

2012-03-16 Thread Mike Haudenschild
Hi Aaron,

The first thing I notice in the log output is this:

 13:32:34|10951|20:20|image|VMware.pm:initialize(273)|no methods are 
 available to control VM host vsphere, the
 
 vSphere SDK cannot be used to control the VM host and the host OS cannot be 
 controlled via SSH

Are you able to SSH from the management node to the VMWare host?

ssh -i /etc/vcl/vcl.key VM host

Regards,
Mike

Sent via iPhone



Re: Creating a base image

2012-03-16 Thread Aaron Coburn
Given that your VMhost is called vsphere, is it safe to assume 
that you plan to use the vSphere API to access the host?

If that is the case, you need to make sure that the vSphere SDK 
for perl is installed. According to your logfile, the perl libraries are 
either not installed or not in perl's path (i.e. @INC).

If you intend to use the perl SDK, you can install it from here: 

http://www.vmware.com/support/developer/viperltoolkit/

(be sure to select the release corresponding to your VMware version)

If you are not going to use the vSphere API, then make sure to 
read through this page as you configure your VMhost:

https://cwiki.apache.org/confluence/display/VCL/VMware+Configuration


Aaron

--
Aaron Coburn
Systems Administrator and Programmer
Academic Technology Services, Amherst College
(413) 542-5451 acob...@amherst.edu


On Mar 16, 2012, at 1:13 PM, Aaron Bryant wrote:

 After seeing your suggestion, I have done the following changes to no avail:
 
 - Disabled Selinux
 - Changed the syntax of the hosts file from IP Address - Alias - Hostname, to 
 IP Address - Hostname - Alias (Just to make sure this wasn't causing a 
 problem)
 
 I can't seem to find any problems with permissions.  I also could not find 
 any valid reason for the cat to fail, although directly after it fails, it 
 prints the output of the cat command in the log file.
 
 The problem is still occurring even after the simple changes I have made.  
 Any additional suggestions?
 
 
 
 On 2/28/2012 4:21 PM, Dmitri Chebotarov wrote:
 Aaron
 
 The problem seems to be with your /etc/hosts file.
 'cat /etc/hosts' failed with exit status 255, when exit status 0 is 
 expected. B/c of this, vcl couldn't get private ip address and didn't 
 attempt to connect to the host. I cannot think of any valid reasons why 'cat 
 /etc/hosts' would fail... Can you double-check syntax, permission, selinux, 
 ... ?
 
 ...
 |10951|20:20|image|  WARNING 
 |10951|20:20|image| 2012-02-28 
 13:32:34|10951|20:20|image|DataStructure.pm:get_computer_private_ip_address(1589)|failed
  to cat /etc/hosts on this management node, exit status: 255, output:
 |10951|20:20|image| 127.0.0.1 localhost localhost.localdomain localhost4 
 localhost4.localdomain4
 |10951|20:20|image| 192.168.0.101 1vsphere 1vsphere.upb.pitt.edu
 ...
 On Feb 28, 2012, at 13:28 , Aaron Bryant wrote:
 
 Greetings,
 
 I completed installing Cygwin SSHD and I'm attempting to run the vcld 
 -setup command.  My problem occurs in this step while creating a base 
 image.  I am able to get through all of the prompts and start the creation 
 before the script crashes.  I have checked the '/var/log/vcld.log' log file 
 and this is a portion of the output:
 
 ---
 VCL::Module::Provisioning::VMware::vSphere_SDK
 
 |10951|20:20|image|  WARNING 
 |10951|20:20|image| 2012-02-28
 
 13:32:34|10951|20:20|image|VMware.pm:initialize(273)|no methods are 
 available to control VM host vsphere, the
 
 vSphere SDK cannot be used to control the VM host and the host OS cannot be 
 controlled via SSH
 |10951|20:20|image| (
 
 0) VMware.pm, initialize (line: 273)
 |10951|20:20|image| (-1) Module.pm, new (line: 207)
 |10951|20:20|image| (-2)
 
 Module.pm, create_provisioning_object (line: 423)
 |10951|20:20|image| (-3) State.pm, initialize (line: 117)
 |10951|
 
 20:20|image| (-4) Module.pm, new (line: 207)
 |10951|20:20|image| (-5) vcld, make_new_child (line: 564)
 
 
 |10951|20:20|
 
 image|  WARNING 
 |10951|20:20|image| 2012-02-28 13:32:34|10951|20:20|image|
 
 Module.pm:create_provisioning_object(431)|provisioning object could not be 
 created, returning 0
 |10951|20:20|image|
 
 ( 0) Module.pm, create_provisioning_object (line: 431)
 |10951|20:20|image| (-1) State.pm, initialize (line: 117)
 |
 
 10951|20:20|image| (-2) Module.pm, new (line: 207)
 |10951|20:20|image| (-3) vcld, make_new_child (line: 564)
 |10951|
 -
 
 
 I have checked and am able to ssh between the management node and the 
 virtual machines using key authentication.
 
 Please see the log file attached for more information.
 
 Thanks,
 Aaron Bryant
 Logfile.txt
 
 
 
 Thank you.
 
 -
 Dmitri Chebotarov
 dcheb...@gmu.edu
 703-993-6175
 -
 
 
 
 Logfile.txt



smime.p7s
Description: S/MIME cryptographic signature