Veeam Backup and Replication.

pawing used Ask the Experts™
what are the best practices to reduce "Stun" when creating a VM backup using veeam backup and replication.  we are using Nutanix solutions using VMware, I have read best practices. however, they have conflicting ideas about there best practices.

would somebody be able to give me some ideas based on their experience on what would be the best practices to reduce "stun" during backup.
Watch Question

Do more with

Expert Office
EXPERT OFFICE® is a registered trademark of EXPERTS EXCHANGE®
VMware and Virtualization Consultant
Fellow 2018
Expert of the Year 2017
If you are referring to the practice of stunning as part of the Snapshot routine, which is used in the Backup API, by most if not ALL Backup applications of VMware vSphere as a method of getting the lock off the parent VMDK, to backup...

1. Schedule Backups on fast storage.
2. Schedule Backups out of hours, when the VMs are quiet, idle, or not under heavy I/O.
3. Don't use snapshots, or a backup product which uses them. e.g. Snapshot on the SAN.
Mr TorturSystem Engineer
with Veeam, VMware and Nutanix I don't see other way to evade the stun state.
A big way to reduce it is San snapshot but at the moment Veeam supports it only for some HP and Netapp. Incoming EMC support is expected.

I agree with Andrew Hancock comments about reducing stun state.

What are the "conflicting ideas about best practices" bothering you, can you explain ?


Thanks a lot for the help guys, we found a solution on how to minimize the stun state. it is more efficient compare to our current setup.

when I was looking for answers to our problem, I came across with one of VMware document in that document it says that use Network Transport instead of appliance Transport (hot-add). At the same time I was also asking technical support from Veeam and there were saying to use Hot-add transport (our current setup).

Our current setup uses appliance transport with one veeam proxy  serving  12 exsi host.

we tested 2 transports  in our environment and it turns out they have their pros and cons. Using Network transport somehow minimizes the stun state, we only loss about 8 ping packets for the whole duration of backup. but the waiting time to finish the backup is too long.

we tested also the appliance transport with veeam proxies installed in each host, and it turns out this is the most efficient configuration that will be applicable with our needs. we only loss about 9-10 ping packets during the entire duration of backup. the only thing is we have to purchase more OS license.

 thanks a lot for the help.
Mr TorturSystem Engineer

yes pros and cons, it depends.
Finally in a transport term the more performant way would be San transport (FC), then hot-add and finally network, but also depending of your ressources (network speed, physical esx ressources for proxies, etc...). And also you can have other less basic solution (e.g. make hot add with proxies but over FC).

In the past with Veeam (or vRanger) I usually found hot add better than LAN but had a few issues with hot add because of its design.
Veeam proxies are made to be scalable, so yes you found it on your own in case of low performance you can test to add more proxies in your VMware environnment, also Veeam proxies must be windows OS so you got to pay the license.

Do more with

Expert Office
Submit tech questions to Ask the Experts™ at any time to receive solutions, advice, and new ideas from leading industry professionals.

Start 7-Day Free Trial