Automatic starting a replica

[jcorbin121]

Automatically starting up (powering on) a replica - Is this feature in a 5.x version? 6.x version? How will it work, is it rule based? heartbeat based? We are getting ready to deploy 5.2 and have 10 or so servers that run apps that need HA, we would like to test using a replication scenario. Any info we can get our eyes on would be much appreciated.

thanks!

John

[Vitaliy S.]

Hello John,

Automatic failover is not a part of v5 and v6. If you want to failover automatically to a replicated VM in case of a disaster, you may want to use these instructions.

Thanks.

[Gostev]

John, would you use automatic failover in production? We found that many customer strongly oppose automation around failovers, as even some network glitch can cause the failover with any sort of automation in place. Customers do like push-button failover, which is what we are adding in v6 with automatic network reconfiguration and re-IP. Thanks.

[jcorbin121]

True I may not necessarily use automated fail-over in production, the push-button solution is probably going to cover most scenarios for my production environment.

[jhaines]

Hi,

I have a VMware 5.5 ESX host replicating to a offsite ESX host, I have replication working but can't see how to automate failover if the master VM fails?

Thanks
Julian

[foggy]

Julian, Veeam B&R does not provide built-in automatic failover capabilities due to considerations given above and in this thread. However, nothing stops you from implementing it using third-party solutions.

[Vitaliy S.]

During VeeamON session Vladimir and I had a session showing how to achieve this with Veeam ONE and PowerShell/PowerCLI scripts. All these instructions will be posted after Veeam Availability Suite v8 will go live on our website.

As regards, previous versions, then you can use post alarm actions in Veeam ONE to trigger VM failover through a PowerShell script.

[veremin]

Just be aware to avoid the biggest issue with automatic failover - split brain situation, where based on one failure, the dr vm is powered up, and in reality the production vm is still up and running. Thus, multiple comprehensive verifications are mandatory.

Thanks.

[virtualz]

Hi Vitaliy,

Can you confirm the links that you referred to in your previous post?

Thanks!

[Vitaliy S.]

Hi Adam,

You can adapt these links to your needs or use examples from this thread > ONEbelievable tricks: Veeam ONE and Automation Services

Thanks!