I was busy setup the demo solution for the Cisco summit yesterday. The Demo we had were displaying the VMware, Cisco UCS, Nexus 5000, MDS 9124 & Netapp Storage Solution. 1 of the Surprise thing happened during the setup, which the power source for our MDS 9124 had been tripped during the installation yesterday. In this scenerio, all our connection to ESX host and VM were disconnected. It took us for 25 mins to recovered the power failure and the MDS Switch was back on line after that. I was thought to reboot all the ESX host as we are performing BOOT FROM SAN for all the ESX hosts that we setup. Surprise happened here, which I found all the ESX host were still continue running. I did the command uptime and check the system uptime from vcenter, it showed that the ESX host were not rebooted during the SAN connection drop from UCS to our Netapp FAS storage.
I further checked the virtual machines been power on in the ESX servers, which show all the VM were continue running without system crash or rebooted. Now I realize that the failure on SAN switch may not necessary result system crash or hung, in fact it may allow you resume the system state once the SAN switch are back online, of course, this is no guarantee assumption, just some surprise finding experienced yesterday would like to share here. Read more »
VMware had done the best job since they introduced vmotion and cold migration which allow us to move the state of the virtual machine from 1 Host to another with or without down time. In virtualization, it encapsulated the entire Server into the virtual machine format and it is hardware independent. Therefore you can always move from 1 host to another without having re-install or reconfigure your operating system or application.
Well, since we had achieved this in virtual environment, what about physical environment?
Today, 10 G Ethernet consider as common technology at least to the enterprise customer. Many of us are looking into 10G Ethernet from server to switch via L2 or L3 networking perspective. The main reason behind this are due to the Virtualization and Cloud Computing technology driven. Few years ago when we came out the design with multiple gigabits NICs on the ESX host to support NIC teaming, load balancing and etc, which may end up deliver around 14 Gbps of network bandwidth and 8Gbps of FC bandwidth per host. Each ESX host are consolidate 20 to 30 virtual machines. At that time, this may be the best option we could had done due to technology limitation. Is this design GOOD enough? I will say yes for previously, and NO for today technology availability.
Photo shot of the real environment which have 14 gigabits NIC Ports and 4 x 2 Gbps FC.
If you are currently deploy the nexus 1000V with the vSphere 4 update 1 and plan to run windows 2008 R2, here are some major issue we had faced for the pass 2 weeks. We experienced all the windows 2008 R2 with packets drop right after the vmotion completed. the Windows 2008 R2 virtual machine will not be reachable from external, but you can ping the gateway from the virtual machine console as usual. After our research and troubleshoot, this was due to the multicast function which seems to be conflict with windows 2008 R2 and become 1 of the major bugs in the Nexus 1000 version 1.2. There are some work around which suggest to turn off the igmp snooping, but our scenario, this didn’t work as it needed. If you are running on windows 2003, you will not experience the situation as we experienced here. Read more »
