by Duncan Epping
It took about about a month to get this published, but here it finally is: vSphere 5.1 Clustering Deepdive on iBooks.
Yeah yeah, we know… you also want Nook and lulu.com says it is pending so that means it probably takes a couple of days before it is up on Barnes and Nobles as well.
Someone asked me a question a while back and I figured it was time to write it down… Or in this case to record a video. The vSphere Web Client is a powerful tool when it comes to finding events and problems. This video shows how you can use the vSphere Web Client to figure out which virtual machines were impacted by a host failure and restarted by HA. On top of that I also show you how you can use PowerCLI to list all virtual machines that were restarted recently by HA. No I didn’t write that PowerCLI blurb myself, I elegantly stole it from the infamous PowerCLI guru Jonathan Medd. So if you need the blurb, hit his article and check the “update 2” section as it contains the code for vSphere 5.0 and up. (I tested it on 5.1 and it works as you can see in the video.)
I received two questions on the same topic last week. The question was around using the PDL enhancements in a non-stretched environment… does it make sense? The question was linked to a scenario where for instance a storage admin makes a mistake and removes access for a specific host to a LUN. For those who don’t know what a PDL is read this article, but in short it is a SCSI sense code issued by an array when it believes storage will be permanently unavailable.
First of all, the vSphere HA advanced option “das.maskCleanShutdownEnabled” is enabled by default as of vSphere 5.1. In other words, HA is going to assume a virtual machine needs to be restarted when it is powered and isn’t able to update the config files. (Config files contain the details about the shutdown state normally, was it an admin initiated shutdown?)
Now, one thing to note is that “disk.terminateVMOnPDLDefault” is not on by default. If this setting is not explicitly enabled then the virtual machine will not be killed and HA won’t be able to take action. In other words, if your storage admin changes the presentation of your LUNs and removes a host accidentally the virtual machine will just sit there without access to disk. The OS might fail at some point, your application will definitely not be happy, but this is it.
To answer the question, yes even in a non-stretched environment it makes sense to enable both disk.terminateVMOnPDLDefault and das.maskCleanShutdownEnabled. Virtual machines will be automatically restarted by HA if they are killed by the VMkernel when a PDL has been detected.
As vSphere High Availability was completely revamped in 5.0 not a lot of changes have been introduced in 5.1. There are some noteworthy changes though that I figured I would share with you. So what’s cool?
Although many of these speak for itself, I will elaborate on why these enhancements are useful and when to use them.
The ability to set slot size for “Host failures tolerated” allows you to manually dictate how many virtual machines you can power-on in your cluster. Many have used advanced settings to achieve more or less the same, but through the UI things are a lot easier I guess.
Now if you do this, it could happen that a virtual machine needs multiple slots in order to successfully power-on. That is where the second bullet point comes in to play. In the vSphere Web Client you can now see a list of all the virtual machines that currently span multiple slots.
Support for Guest OS “Sleep Mode” in environments where VM Monitoring is used was added. This was reported by Sudharsan a while back and I addressed it with the HA engineering team. As a result they added in the logic that recognizes the “state” of the virtual machine to avoid unneeded restarts. Thanks Sudharsan for reporting! (I can’t find this in the release notes however)
With 5.0 the Application Monitoring SDK was opened up to the broader audience. It was still a separate installer though. As of vSphere 5.1 the App Monitoring SDK is part of the VMware Tools SDK. This will make your life easier when you use Application Monitoring.
Those running stateless will be happy about the fact that the FDM VIB is now part of the Auto-Deploy image profile. This will avoid the need to manually add it every time you create a new image.
Last but not least, in 5.1 we re-introduce “das.failuredetectiontime”… well not exactly but a similar concept with a different name. This new advanced setting named “das.config.fdm.isolationPolicyDelaySec” will allow you to extend the time it takes before the isolation response is triggered. By default the isolation response is triggered after ~30 seconds with vSphere 5.x. If you have a requirement to increase this then this new advanced setting can be used.
I just sent out this tweet. Last time we did this promo many people said they somehow missed it and asked if we could do it again. We started the promo this morning, make sure to download it now as it will not be free (yes that is $ 0,-!) for long. Promo ends Thursday.
Special offer, the vSphere 4.1 HA/DRS Deepdive Kindle edition for FREE! vmwa.re/free41
Enjoy,