Yellow Bricks

by Duncan Epping

ha

Max amount of VMs per Host?

· ·

If I would ask you what the max amount of VMs per Host is for vSphere what would your answer be?

My bet is that your answer would be 320 VMs. This, of course, based on the “virtual machines per host” number that page 5 of the Configurations Maximum for vSphere shows.

But is this actually the correct answer? No it’s not. The correct answer is, it depends. Yes… it depends on the fact if you are using HA or not. The following restrictions apply to an HA cluster(page 7):

  • Max 32 Hosts per HA Cluster.
  • Max 1280 VMs per Cluster.
  • Max 100 VMs per Host.
  • If the number of Hosts exceeds 8 in a cluster, the limit of VMs per host is 40.

These are serious restrictions that will need to be taken into account when making a design for a virtual environment. It touches literally everything. From your Cluster size down to the hardware you’ve selected. I know these configuration maximums get revised with every update but it is most definitely something one would need to consider and discuss with the customer…

Just wondering what your thoughts are,

vSphere HA Isolation Response

· ·

As of VMware vCenter 2.5 Update 2 the HA default isolation response changed from “Power Off” to “Leave powered on”. A lot of people liked this new default setting because it would lower the chances of downtime due to a “false positive”. I’ve never been a fan though, I just don’t like using degraded hardware or a degraded ESX host for that matter.

Those that did like the change should take notice of the fact that with vSphere comes a new default isolation response:

Note that this change is only for new clusters, if you upgrade(d) your vCenter the selected isolation response will remain. For those of you who never looked into the setting “Shut down”, it uses VMware Tools to initiate a guest shut down. If the shut down does not complete within five minutes the VM will be powered off. These five minutes are a configurable setting, if you want to increase or decrease it add the following advanced option das.isolationShutdownTimeout with the new value in seconds.

Virtual Machines appear to be running or registered on multiple ESX Servers

· ·

I was doing my daily round on the VMTN Forums and noticed this topic on VMs flickering between ESX hosts. I’ve personally never witnessed this and didn’t even knew it was a known issue. Luckily Troy Clavell pointed the topic starter out to a KB article related to this exact issue. Apparently it’s being caused by the fact that the VM is registered on two hosts at the same time.

Symptoms:

  • After one of the following, a Virtual Machine appears as being registered on two ESX Servers:
    • A VMotion fails to complete correctly or times out in VirtualCenter
    • A DRS issue where virtual machines are VMotioned automatically in quick succession
    • When a machine is powered on during VMware HA failover.
    • The Service Console on an ESX host is low on memory starving the vpxa process
  • In VirtualCenter, you see the virtual machine as appearing on one ESX Server for a few seconds, then it seems to be on the other.
  • The virtual machine may appear to jump back and forth among different ESX hosts.

I’m not going to copy/paste the solution cause the KB article will probably change over time, but it’s most definitely worth looking into… it does sound like something that can happen to all of us.

Alarm Actions, exploring the new version of ESX/vCenter

· ·

Eric Sloof(1, 2, 3) already explored the new “alarms and actions” capabilities. There’s one that really stands out in my opinion in the next release of ESX/vCenter(vSphere):

Indeed, “Enter maintenance mode“. Now you might wonder when you would want to use this. Wouldn’t it be nice that when your hardware is degraded, for instance memory status changed or hardware power changed, the host enters maintenance mode. I wouldn’t want to run virtual machines on a host that has a memory error or runs on only one power supply. Of course this action only applies to the “host” objects:

HA enhancements, exploring the next version of ESX/vCenter

· ·

Let’s start with a screenshot:

These are the properties of an HA cluster, as you can see there are two sections that changed:

  1. “Enable Host Monitoring” is a brand new feature. Anyone who did network maintenance while HA was enabled knows why this feature will come in handy. Those that didn’t: Isolation response! If ESX is unable to send or receive it’s heartbeat and can’t ping it’s default isolation response address it will shutdown all VM’s. To avoid this behavior you can switch of HA for a specific host with this new feature. In just four words: Maintenance mode for HA.
  2. Besides the amount of host failures a cluster can tolerate you can also specify a percentage. With the “host failures” option VMware uses the highest values of CPU and Memory reservations to calculate the amount of slots. (For more on slot / slot size read the Resource Management Guide for ESX 3.5) With the new option “Percentage of cluster resources” this isn’t the case. This new option uses the actual reservation of the VM and calculates the total percentage of resources used based on these calculations. If no reservation have been made it uses the default 256Mhz / 256MB reservation. In other words, you will be more flexible and will get a higher consolidation ratio. If the default reservation values are to low you can always use the advanced options to increase it. Another new option is “specify a failover host”. This option can be compared to “das.defaultfailoverhost”. The good thing about this option is that the designated host will be used for fail-over only. DRS will not migrate VM’s to this host, and it’s not possible to start VM’s on this host.