Disaster recovery orchestration for the Enterprise
Post Reply
Mermaid
Veeam Software
Posts: 3
Liked: 1 time
Joined: Apr 02, 2018 4:16 am
Full Name: Satoko Takahashi
Contact:

Trigger of VAO failover to VBR

Post by Mermaid » Sep 05, 2018 4:41 am

Hi guys

Could you please let me know
What a kind / various fail will be trigger VAO failover to VBR?

SIncerely,

Alec King
Veeam Software
Posts: 870
Liked: 150 times
Joined: Jan 01, 2006 1:01 am
Contact:

Re: Trigger of VAO failover to VBR

Post by Alec King » Sep 05, 2018 11:31 am

Hello,

If I understand correctly - you are asking what can trigger VAO to orchestrate the failover of replica VMs?

To execute a VAO failover plan, there are two options -

1. Use the VAO UI to RUN your failover plan, manually
2. Use the VAO API to remotely trigger a plan.

The VAO REST API can be used to RUN a failover plan remotely without requiring UI. You can trigger the API from any other system, e.g. BCDR management tool, infrastructure monitoring tool, etc.
Alec King
Vice President, Product Management
Veeam Software

Mermaid
Veeam Software
Posts: 3
Liked: 1 time
Joined: Apr 02, 2018 4:16 am
Full Name: Satoko Takahashi
Contact:

Re: Trigger of VAO failover to VBR

Post by Mermaid » Sep 06, 2018 12:10 am

Hello, Mr.King

Thank you for your quick reply.
May I ask you more deep dive question?

What are the points VAO check when productions site down before trigger planned failover?

I surmise these points are similar VBR.
ex. heartbeat, ping check, host discovery process that runs every 4 hours or something...

Sincerely,

Alec King
Veeam Software
Posts: 870
Liked: 150 times
Joined: Jan 01, 2006 1:01 am
Contact:

Re: Trigger of VAO failover to VBR

Post by Alec King » Sep 06, 2018 9:15 am

Hello,

To confirm, VAO does not automatically start, or halt, failover based on monitoring data. There is a large risk in taking monitoring data (e.g. 'ping test') and based on that data, to either start or stop the failover of critical systems. Failover is a drastic and serious action. To hand this process completely to monitoring automation is risky.

When a failover is triggered, VAO performs checks only to confirm that the DR environment is ready for failover. VAO will check availability & status of DR Site vCenter, Veeam servers, hosts, replica VMs etc. VAO does not check whether the Production site systems are offline or not, as this is not relevant. For example -
- This could be a planned migration and all source systems are still working, nonetheless the plan should still be run.
- There could be application corruption or data loss - so again, even though production VMs are still running, nonetheless failover is still required.

So, by default, VAO will obey the command to perform failover, and will not halt this process because of any monitoring data from the source production site.

Having said that, VAO is a very flexible system and can be extended and customized, for example using PowerShell scripting and APIs. Just as failover can be automatically triggered using these capabilities, failover could also be automatically halted. However I'd repeat, there is a real risk when using only monitoring data to control failover. I would still recommend human approval is part of this process.

Post Reply

Who is online

Users browsing this forum: No registered users and 2 guests