commvault.com commvault.com
Solved

VMware Transport Mode

  • 24 October 2023
  • 6 replies
  • 179 views
C
Userlevel 3
Badge +12

Hi Vaulters,

 

We have a new CV platform that will be deployed soon, the Infra that will be backed up is all virtual based on VMware 6.7.

 

Regarding design, we are checking which transport mode is the most suitable in our case, and as of now, since the VMs are hosted on datastores residing on a Storage Disk Array, it looks like SAN mode is the most efficient way to back up our VMs.

 

Can anyone give us the right steps to implement this ? A step-by-step procedure, since we couldn’t retrieve this on the documentation.

 

We are also looking for HotAdd transport mode, which also seems feasible in our case.

Overall, what are the pros and cons of each tranport mode (SAN and HotAdd) ? For example, we read that restore performance is poor on SAN mode when VM disks are thin.

 

Kind regards.

 

icon

Best answer by Sebastien Merluzzi 25 October 2023, 09:41

View original

6 replies

S
Rank
Userlevel 6
Badge +14

Hi @Commvault Engineer ,

Can you check this:

https://documentation.commvault.com/2023e/expert/36340_transport_modes_for_vmware.html

Configuration:

https://documentation.commvault.com/2023e/expert/32202_configuring_transport_modes_for_vmware.html

We have different scenarios here: https://documentation.commvault.com/2023e/expert/36343_scenarios.html

Let us know if you have any questions.

Best Regards,

Sebastien

Onno van den Berg
Rank
Userlevel 7
Badge +19

If the storage solution that is used to host the data related to backup data can be exposed as object than I would for sure go for hot-add combined with storage accelerator. The nice thing about hot-add is also that you can leverage the FREL appliances and deploy them on the clusters directly from Command Center. 

C
Userlevel 3
Badge +12

Hi @Commvault Engineer ,

Can you check this:

https://documentation.commvault.com/2023e/expert/36340_transport_modes_for_vmware.html

Configuration:

https://documentation.commvault.com/2023e/expert/32202_configuring_transport_modes_for_vmware.html

We have different scenarios here: https://documentation.commvault.com/2023e/expert/36343_scenarios.html

Let us know if you have any questions.

Best Regards,

Sebastien

Hi @Sebastien Merluzzi Thanks a lot for your prompt feedback. Actually, we already came across all of those documentation references and that's why we're now thinking about opting for SAN mode.

 

Our concerns are more about implementation process, not from CommVault side, since it’s clear that from CV we just need to precise that we are using SAN mode as per your second shared link. But, it’s more about Infrastructure side, we’d love to have an overview on how does that work, I guess it comes down to these steps:

  1. Map datastores LUNs from storage side to the physical MediaAgent which has VSA installed (Proxy).
  2. Configure the needed zoning from SAN switch.
  3. Scan the new LUNs on the MediaAgent.

And from here, comes our concerns :

  • Does the scanning of the LUNs from MediaAgent side causes any data corruption ?
  • Are the above steps sufficient from Infra perspective ? Or, is there any other missing step ?
  • During a backup, how does exactly the ESXi know that a backup is performed on its datastores ?
  • Regarding applications/DBs, is per our understanding, it's always better to use its specific agent, when doing so, it means that some VMs will be needed to be integrating as standalone client on CV in order to install the required agent (for example, Oracle DB iDA), does the backup of the DB hosted on the VM must pass by the IP network ? Or does it also passes by the SAN configured for VM backup ?

It's a lot to digest, but any guidance is welcomed, it would help us understand how the implementation of the process is done.

 

Regards.

C
Userlevel 3
Badge +12

If the storage solution that is used to host the data related to backup data can be exposed as object than I would for sure go for hot-add combined with storage accelerator. The nice thing about hot-add is also that you can leverage the FREL appliances and deploy them on the clusters directly from Command Center. 

Unfortunately, it’s not, it’s only used for SAN in order to host ESXi datastores in which VMs reside.

S
Rank
Userlevel 6
Badge +14

@Commvault Engineer ,

Scanning LUNs do not corrupt the Data, the same way when you scan the Disks using Diskpart.

You can use VMFSSANINFO Tool on the VSA Proxy to confirm that the VMFS LUNs are visible:

https://documentation.commvault.com/2023e/expert/113660_vmfssaninfo_tool.html 

This is what you should see:

 

We send VDDK API commands to communicate with vCenter/ESXi Hosts.

You can also check the Data Flow for VMware Backup Process:

https://documentation.commvault.com/2023e/expert/135133_data_flow_for_vmware_backup_process.html

C
Userlevel 3
Badge +12

@Commvault Engineer ,

Scanning LUNs do not corrupt the Data, the same way when you scan the Disks using Diskpart.

You can use VMFSSANINFO Tool on the VSA Proxy to confirm that the VMFS LUNs are visible:

https://documentation.commvault.com/2023e/expert/113660_vmfssaninfo_tool.html 

This is what you should see:

 

We send VDDK API commands to communicate with vCenter/ESXi Hosts.

You can also check the Data Flow for VMware Backup Process:

https://documentation.commvault.com/2023e/expert/135133_data_flow_for_vmware_backup_process.html

Thanks a lot @Sebastien Merluzzi.

Reply


Terms & ConditionsCookie settings