Converting Agent-based Servers to Agentless

Hi aivanov_35976:

First the warnings! Although the SQL/SharePoint/Exchange Servers are protected properly in an agentless manner, you loose some features such as mountability/attachability, the capability of having multiple protection groups per agent etc.

As these may (or may not) be of importance for your environment, you need to do a thorough assessment of what agents should be protected agentlessly and which to be agent based. Moreover, as you have already noticed, changing the protection type from agent based to agentless triggers a base image. The reason is that the agent based protection keeps track of the changed blocks inside the protected machine (specifically in the aa_log files in the 'System Volume Information' folder on each volume) while the agentless protection does it -- in the VMWare case -- via CBT, in files located on the ESXi host.  

Now, these being said there may be a way of combining the recovery points. Please note that this does NOT exclude the new base images.

You need the agent id of the machines that you have already removed from protection. Assuming that you did not keep a copy of the RR registry either on the agent or on the core, the easiest way to find it out is via Powersshell using the command line below:

PS C:\> (Get-RecoveryPoints -ProtectedServer <agentname>)[0].agentid

once the agent ID is obtained, all you have to do is to identify the agentlessly protected agent in the registry  (HKEY_LOCAL_MACHINE\SOFTWARE\AppRecovery\Core\Agents) and replace the current agent ID with the one you got above and restart the core service.

Please check the picture below. In my case the agent name is "Oracle 7". (I used an agent name containing a space to cover all possibilities).

Hope that this helps.

Please let us know how it goes.

Hi aivanov_35976:

First the warnings! Although the SQL/SharePoint/Exchange Servers are protected properly in an agentless manner, you loose some features such as mountability/attachability, the capability of having multiple protection groups per agent etc.

As these may (or may not) be of importance for your environment, you need to do a thorough assessment of what agents should be protected agentlessly and which to be agent based. Moreover, as you have already noticed, changing the protection type from agent based to agentless triggers a base image. The reason is that the agent based protection keeps track of the changed blocks inside the protected machine (specifically in the aa_log files in the 'System Volume Information' folder on each volume) while the agentless protection does it -- in the VMWare case -- via CBT, in files located on the ESXi host.  

Now, these being said there may be a way of combining the recovery points. Please note that this does NOT exclude the new base images.

You need the agent id of the machines that you have already removed from protection. Assuming that you did not keep a copy of the RR registry either on the agent or on the core, the easiest way to find it out is via Powersshell using the command line below:

PS C:\> (Get-RecoveryPoints -ProtectedServer <agentname>)[0].agentid

once the agent ID is obtained, all you have to do is to identify the agentlessly protected agent in the registry  (HKEY_LOCAL_MACHINE\SOFTWARE\AppRecovery\Core\Agents) and replace the current agent ID with the one you got above and restart the core service.

Please check the picture below. In my case the agent name is "Oracle 7". (I used an agent name containing a space to cover all possibilities).

Hope that this helps.

Please let us know how it goes.