mainesysadmin.com

While working with vmware support on the phone they required me to run a job from Virtual center to gather diagnostic log data from our VI3 cluster.  You can try this by opening your, “Virtual Infrastructure Client” and selecting the “Administration” menu and then “Export Diagnostic Data”.  When I run this command for as little or much information required I ALWAYS get the following error, “Failed to create diagnostic bundle.”

If you ssh to your host directly you can run the following command from each host and sftp the logs from the host afterwards.

vm-support

In the last year we had a physical server’s O/S have a severe blue screen error which required the server to be rebuilt.  This server was a critical server for our operation and the downtime was not easy.  In order to prevent that from happening in the future we decided that server would be a better fit in the virtual infrastructure because of the High Availability offered by Vmotion.  The only problem is this server requires a USB license dongle.  Everyone knows that you don’t have a USB port to plug into when running VI3, so what do you do?

Keyspan USB server (USB over IP)

This specific USB server allows you to choose which USB port(s) you want to dedicate to each server.  Some USB servers do not allow you to dedicate each port to a different server.

What a breeze to setup as well.  After you mount the device, give it power and ethernet…Just install the Keyspan software.  The software client will find the USB server automatically.  Than you need to edit the config to dedicate the USB device to that VM and you’re ready for action.

Now you can reboot and vmotion to your heart’s desire and not lose connectivity to your USB license dongle.  Great little trick if you’re as dedicated to virtualization as we are!

After a year of service the 3 685 c class blades that I work with have served us well.  They started as Dual Proc AMD 8216 (dual core 2.4ghz) with 16gb of ram.  In order to expand the virtualization in our environment we decided to double the resources in each of the blades by adding 2 more procs and 16gb of ram.  I added the ram last week, but because of limitations in the system bus you need to have all 4 procs installed before the other 8 dimm banks are activiated.

Steps involved to upgrade procs & memory:

1. Add new licenses from vmware (esx is licensed by every 2 physical procs/sockets)
2. Login to Virtual Center and put host in maintenance mode.
3. This should vmotion all virtual machines from that host to other running hosts.
4. Shutdown host
5. Carefully install cpus and ram
6. Start server (will reboot once after starting)
7. Upgrade complete with ZERO downtime in the middle of the day.
8. 685c blades are now 4 cpu, 8 core and 32gb ram and ready for some more action.

I landed in San Francisco yesterday at around 1:3o pm. Another excellent experience with United; 6 hours and on time from Boston. I rented a car and found my way to the hotel pretty easily. Most of the roads feel pretty difficult to maneuver. The Marriot I’m staying at is around the corner from the Moscone center which is great.

Today I was able to register for the event which included a free laptop bag and t-shirt. I started my day with an 8am lab dealing with DR and multiple sites. The lab felt slightly unorganized but it was the first one of the day and also first of the conference. There were multiple typos in the lab manual. If someone had never used Virtual Center before I can see it useful. I followed this lab with one on Security and Hardening ESX. I actually took this lab last year as well and I’ve found that this lab might not have changed at all and if it did is was probably slightly. I’m not sure if that was good or bad. I thought the lab would have been slightly more updated.

I’ve found a neat KVM utility for vmware vm’s.  This utility is interesting to me because it utilizes virtual center permissions but you don’t have to deploy the whole boat for the user requiring simple console access.  You can take a peak at it here:

http://www.ntpro.nl/blog/categories/7-Virtual-Machine-MKS-Client

As promised I remembered to dig into my virtual center event logs to pull out the error message that I received after upgrading my 3 esx servers.  “An error occurred during configuration of the HA Agent on the host.” “hostmisconfigured : failed to resolve hostname/ip using short hostname yourdomain.”Simply add your hostnames and ip’s to /etc/hosts

I thought the upgrade went completely smooth but not so much.  After the upgrade and reboot each esx’s host file needed to be edited to include it’s own name and ip address.  127.0.0.1 will not work.  I added my other hosts to each host’s /etc/hosts file for good measure.  Once that file was saved I enabled HA and all was happy.  I’ll add the error I saw in VC later…