jsword
Contributor
Contributor

Application Monitoring/Traditional Clustering within VMs with VMotion/DRS/VMWare HA

I am often asked by customers what effect the use of VMotion, DRS, or VMware HA will have on various traditional clustering and data replication tools used inside VMs such as:

  • MSCS

  • CA Wansync

  • Neverfail

  • Veritas Clustering

  • SteelEye LifeKeeper and Data Replication

  • Red Hat Clustering

  • Linux-HA

  • DRBD

Do these traditional solutions work as expected inside the VMs after VMs are moved around using VMotion, DRS, or VMware HA?

PS - This question is not meant to foster a debate between whether one should use VMwareHA versus traditional clustering. That subject is adequately discussed in other threads. I often have customers looking to move clustered legacy applications to VMware ESX environements who do not want to initially change the current application or cluster configurations during the migration process.

0 Kudos
2 Replies
JoJoGabor
Expert
Expert

It shouldn't have any effect. The OS is not aware of any change. If you are really unlucky the cluster heartbeat may happen at the exact fraction of a second that VMotion causes a blip on the network, but the chances of that are tiny.

0 Kudos
Texiwill
Leadership
Leadership

Hello,

Actually it can have an impact. Those VMs that participate in clusters are ones that usually can not use VMware DRS or VMotion due to other constraints. Since many of these 'detect' blips on the network and if that 'blip' is too long due to other issues you could end up with failover when you least like it.

How you setup the cluster (i.e cluster in a box (CiB) or cluster across ESX hosts) will tell you whether or not VMware DRS and VMotion will apply. In addition, if you are using VMware HA, you need to be careful where a cluster node lands during failover, it should not land on the same host as another cluster. And if you use CiB you have to land them on the same node.

So yes, using traditional Clustering does impact HOW you do your implementation and takes quite bit more thought when it comes to failover cases.


Best regards,
Edward L. Haletky
VMware Communities User Moderator
====
Author of the book 'VMWare ESX Server in the Enterprise: Planning and Securing Virtualization Servers', Copyright 2008 Pearson Education.
Blue Gears and SearchVMware Pro Blogs -- Top Virtualization Security Links -- Virtualization Security Round Table Podcast

--
Edward L. Haletky
vExpert XIV: 2009-2022,
VMTN Community Moderator
vSphere Upgrade Saga: https://www.astroarch.com/blogs
GitHub Repo: https://github.com/Texiwill
0 Kudos