Saturday, September 27, 2008

Political war between Marathon and VMware on Fault Tolerance Feature of VMware.

Political war between Marathon and VMware on Fault Tolerance Feature of VMware.

A great political debate has been started on VMWare fault tolerance feature during
VMWare world meet at LA 2008. A four reason listed by Marathon has draw sharp criticism from VMware lover. Please read the complete story here. One of the four reason listed for short fall of VMWare FT is listed as ========================Snip==================================


2. Complexity on top of complexity. In order to use VMware Fault Tolerance, you’ll first have to install both VMware HA and DRS. No small feat in and of themselves. Then, because VMware FT requires NIC teaming, you’ll also have to manually install paired NICs. Then you’ll need to manually setup dual storage controllers (with the software to manage them) because it requires multi-pathing. And to top it all off, you’re required to use an expensive, and often complicated, SAN.

========================Snip====================================


To this Mike has given a great response on his blog and responded how easy HA and DRS to configure

=========================snip==================================


See those two check boxes? Click them and you've just enabled HA and DRS. If that's
too hard then please comment and let me know how it could possibly be easier. Even my dog has figure out how to do this now. Granted, it's a pretty smart dog.

=======================snip=====================================

What a response… Well Q is if you want to sell you product sell it ,but please do not do some cheap marketing. Where VMwaer is right now MS or any other will take ages to reach and I am ready to bet on my word

DAMN you Microsoft







Friday, September 26, 2008

How To Rename The Host and Change IP address

How To Rename The Host and Change IP address




Sometimes host has to be renamed or IP address has to be changed. Please follow these steps.

  1. Change the name under

[root@xxxxxx root]# more /etc/sysconfig/network-scripts/ifcfg-vswif0

DEVICE=vswif0


MACADDR=00:50:56:4f:65:19


PORTGROUP="Service Console"


BOOTPROTO=static


BROADCAST=xx.xx.xx.255


IPADDR=xx.xx.xx.xx


NETMASK=255.255.255.0


ONBOOT=yes

  1. [root@xxxx root]# more /etc/sysconfig/network

NETWORKING=yes


HOSTNAME=<Full DNS Name>


GATEWAY=XX.XX.XX.XX


GATEWAYDEV=vswif0

3. [root@xxxxxx root]# more /etc/hosts


# Do not remove the following line, or various programs


# that require network functionality will fail.


127.0.0.1 localhost.localdomain localhost


xx.xx.xx.xx %FQDN Name of the host% %NetBIOS name%


Best would be reboot the host.

How to Rename Service Console

I was given task to build the system and then I started building it. It is not that I don’t know how to perform ESX build and believe me each hardware and every environment is different. Person has to go through these experiences as they explore the environment.

I build a new ESX box (DL 380) and then gave everything default. I choose a default VLAN for service console. When machine came back online it is not able to ping the default gateway. I started trouble shooting and starting to figure out how to make sure that if NIC are getting trunked on correct VLAN. To find out this one I used

“esxcfg-nics -l” . This will list us which NIC is up. Now the challenge was if the correct NIC is listening to the “Service Console”. To find out this one I used “esxcfg-vswitch -l” This will show the VLAN ID as well as which NIC is it is configured for.

clip_image002

Here you can see that “vmnic1” is “UP ” but “Service Console ” is configured for “vmnic0” . So here I have to basically swap the cable and bring “vmnic0” or change service console to listen to “vmnic1”. Easier would be swap the cable and then label the “service console” to listen to appropriate VLAN. Right now “Service Console ” is set to VLAN 0. To change the VLAN label use following “ esxcfg-vswitch –v <vlan id> -p “Service Conole” vSwitch0 ” (Command in the fist quotes should be run).

I did that and that fixed the network.

How to fix “Unable to mount root fs ” Error

Mount point has been created and given to me by the NetAPP engineer. I mounted the space on host. After a performed the mount for some reason it started showing up some wired error message like when I ran “vmware-cmd –l ” I get a message vmware auth failed to login. I thought something really goofed up when I created the “VMKernal” to access NFS mount. I decided to reboot the host and see if that clear the error messages but no it has created more problems for me. When I rebooted the host it gave me message

clip_image002

Something really bad happened ☹ . I started hunting on google to got some clue

http://www.linuxforums.org/forum/linux-kernel/7283-kernal-panic-vfs-unable-mount-root-fs-00-00-a.html

I decided to change boot parameters. In order to do so let the machine boots up and then when you get this screen

clip_image004

Here you press “E or e ” on Key board and it will take you the screen

clip_image006

Again press “E or e” and then it will give you option where you can edit the field and type the value as suggested in the earlier link.

it says above that i gotta change "LABEL=/" to "/dev/hda0"

clip_image008

Once you are done with editing then reboot the host. But in my case situation still persist. So I decided to do more google and here is the command which work like a butter.

esxcfg-boot ”

Follow the steps.

1. Boot the system and wait till you see following screen .Select here “Service Console only”

clip_image009

2. It will let you inter through into service console (Everything is done through physical connection like ILO).

3. Once you logged in then run the command in following order

1. esxcfg-boot –p

2. esxcfg-boot -b

3. esxcfg-boot –r

Reboot the host and select “VMware ESX server ” and host should bring of the esx.

Friday, September 19, 2008

Enable ssh on ESX host

By default SSH is disabled on ESX hosts and we have to enable SSH. Setting to be done.
Open sshd_config file:

# vi /etc/ssh/sshd_config
Find out line that read as follows:
PermitRootLogin no
Set it as follows:
PermitRootLogin yes

Save and close the file. Restart the sshd:
# /etc/init.d/sshd restart

Serendipity - A Fortunate Accident


16Sep2005
Uploaded by fun_come

Tuesday, September 9, 2008

No interfaces are available for installing a new cluster

User was trying to setup Microsoft NLB. They added one node and it was successful and when they were trying to add another node it was throwing error

No interfaces are available for installing a new cluster

clip_image002

User pointed out that may the VLAN trunking which was implemented couple of weeks back. I was not sure what VLAN has to do with it but when I looked at the error I made a guess that it has something to do with OS rather than VMware. I went to MS support site and gave a search with the error message and found wonderful KB 828258.

………….SNIP …………………………………..SNIP…………………………………

This problem may occur if the subsequent nodes are on servers that were created by using an image. The problem occurs because the servers that are created by using an image may have the same Network GUIDs in their registry if the network adapter is installed as part of the image. The Network Load Balancing Manager uses this GUID to attach and configure Network Load Balancing on the servers. The Network Load Balancing Manager detects the GUID of the first server, and then does not register subsequent network adapters that have the same GUID.

-------------------SNIP………………………………………SNIP………………………………..

It was true because these VM’s was build with “GHOST image”

To resolve this follow these steps in the same sequence

1. Uninstall the tools from the VM’s using add and remove program.

2. Uninstall the network adapter from device manager.

3. Reboot the machine.

4. Login to the machine locally and then setup the network again reboot the VM’s

5. Install the tools and then reboot the VM’s

These steps have to be repeated for all the nodes which will be involved in NLB.

Once this is completed then NLB will be configured without any error.

How to create new "VMDK" desciptor file


Strange issue with the VMDK descriptor file.


One the virtual machine was registered and powered on , but for some reason “vmx” file was missing. I thought I would create a “vmx” file using wizard. I unregistered the “VM” and then tried to create the “vmx” file using wizard. VM was got created but was not able to attached the disk file because “vmdk” descriptor file was missing. “Peewe” how the heck I should create the descriptor file .

  1. Tried using other descriptor file from other VM’s and removed the line something like this

RW 419430400 VMFS “rescue_me-flat.vmdk”


This did not work out because the other information like may not be correct.


ddb.virtualHWVersion = “4″
ddb.geometry.cylinders = “26108″
ddb.geometry.heads = “255″
ddb.geometry.sectors = “63″
ddb.adapterType = “lsilogic”
ddb.toolsVersion = “7201″



I was not surprised with this. So I thought lets create a exact same size disk on some other host and copy over the descriptor file from that host to current host (Because current was not having enough space to create the new disk file).

This also did not work because I guess information is so encrypted that I was not able to find out what need to change. So how did I fix it?


  1. Power off some other VM’s (In order to free up some space).

  2. Once I have enough space create a new disk with same EXACT SIZE using the wizard at the same location where we are trying to create a descriptor file

(This step is very important I wasted my 2 hour to just figure this out)

  1. Now remove the newly created disk from the VM.

  2. Unregister the VM from the host.

  3. Copy the newly created “vmdk” file with the old name (For which we are trying to generate descriptor file).

  4. Edit the renamed “vmdk ” file and make sure the entry is pointing to correct flat file

E.g:

# Extent description
RW 419430400 VMFS “rescue_me-flat.vmdk”

  1. Once the above steps are done then edit the vmx file and search for the line

E.G

scsi0:0.fileName = " rescue_me-flat.vmdk "

  1. Once all the above steps are done then re-register the file using following


Vmware-cmd –s register <path to vmx file>

I would like to thank Paul Shannon for his wonderful blog which really helped me to achieve what I wanted.







Monday, September 8, 2008

Happy B'Day Sweetheart

Happy Birday Sweetheart. As we call you Vibha, Sonuva, Zonki,Gudiya,Chonputuk,Sonpari
We have started preparing for your arrival by seeking blessing from Almighty. Your mom was bit tensed before your arrival but I was sure that we are going to make it with the help of Almighty. Doctor “Carmela A Barr” and her team prepared the bed for your arrival. You were under surveillance to make sure that you are also prepared to join our family. On September 8th 2007 around 10.05 PM you finally arrived in this world and into our family. Every one visited to see my princes ,Rishi, Madhu, Saumitra. But much awaited was Great Grand Mother .She came on 17th September. You were bit late in getting her blessing but she always had her blessing before you came into this world. Finally everyone wanted to leave me alone and on Nov 18th you left for India. When you were 4 days old you gave a first pose and your mom was able to capture it. After looking at the pose we started thinking that you are going to be a great thinker? You started growing like a beautiful princes and then the first word came out of your mouth was “Dada” a great grandfather (Mr. Vishnu Narayan Roy ) of yours . You are always happy to be with your Grand Mother.





FERRERO COLLECTION

Saturday, September 6, 2008

Memory spikes and Reservation

Here is the host configuration

clip_image002

clip_image004

One of the users complains about memory spikes messages that they are getting from the system.

Available Ram is 170988.0 KB for xxxxxxxx . Please take appropriate action.

===============================================================

I ran spotlight (My Fav. freeware tool from Quest Control) on those machine and found the spikes (Check spike before 11.35AM)

clip_image005

I went to the VC and selected the VM’s properties and set the memory reservation to and the memory utilization on virtual machine changed (Check the graph pattern after 11.35AM).

clip_image007