Skip to main content
Solved

Vmware AccessNode Selection Order

  • Nov 17, 2026 11:20 AM UTC
  • 3 replies
  • 32 views

Forum|alt.badge.img+13

is there an option available to prefer VSA selection on 

  • san mode capability (san datastores mapped)

over

  • VSA and ESX Host share the same subnet 

dispatching Rule ?

In my setup, most of the VMs are backed up using nbdssl, although access nodes with mapped san datastores are available for most, but not all VMs of the VM Group

I don’t want to disable the IP rule, because if nbdssl is the only option, I’d like to use VSA, that do not have to cross a firewall when accessing the ESX host

(still running on 11.36.130)

rgds
Klaus

Best answer by johanningk

Hi guys,
I found the source of the problem.

Comparing backups started in two different environments (CommCells), I saw that the VM discovery in one environment, discovered san attached datastores for a VSA, while the other environment did not.

Having in mind, that in the past, the default behavior changed for a couple of features with newer SPs, I added the additional setting bEnableSanDatastoreDispatch = true (which should be the default, according to the documentation), enforcing the presence of this parameter on all VSA.

Surprise !
Last night’s backups used san mode for the vast majority of VMs !

BOL lists the dispatching rules in the following order:

  1. Virtual machines are backed up by proxies running as physical machines that have SAN connectivity to the datastore.
  2. Virtual machines are backed up by proxies that are running as virtual machines on the same host.
  3. Virtual machines that are part of a vSAN stretched cluster are backed up by proxies that are in the same fault domain.
  4. Virtual machines are backed up by proxies that are running as virtual machines on hosts that have access to the datastore.
  5. Virtual machines are backed up by proxies that are on the same subnet as the virtual machine's host.
  6. The remaining virtual machines are distributed evenly across the proxies.

 but this was not true for one of my environments

we use nbdssl and not nbd to meet the requirement of end-to-end encryption of data. This also includes the encryption of data in-transit.

rgds
Klaus

View original

3 replies

Forum|alt.badge.img+10
  • Vaulter
  • Nov 18, 2026 4:04 AM UTC

Hi Klaus,

Unfortunately I don’t know of a method to have an access node selected based on SAN capability. The reason for this is that the transport mode is determined just prior to opening a VM disk. By this point the VM has already been assigned to an access node.

To determine a transport mode prior to assigning a VM would require testing each VM and each disk before committing a VM to an access node.

What I would recommend is to configure VM groups based on their transport mode. For example, have a SAN VM group for VMs that can use SAN and only assign access nodes with the appropriate LUNs assigned. You can still configure the transport mode as automatic so if SAN fails it can fall back to NBDSSL.

Hopefully this helps.

Regards,


In my own environment we use just nbd alone with proxies, this is very important for backup speed


Forum|alt.badge.img+13
  • Author
  • Challenger
  • Answer
  • Nov 18, 2026 9:17 AM UTC

Hi guys,
I found the source of the problem.

Comparing backups started in two different environments (CommCells), I saw that the VM discovery in one environment, discovered san attached datastores for a VSA, while the other environment did not.

Having in mind, that in the past, the default behavior changed for a couple of features with newer SPs, I added the additional setting bEnableSanDatastoreDispatch = true (which should be the default, according to the documentation), enforcing the presence of this parameter on all VSA.

Surprise !
Last night’s backups used san mode for the vast majority of VMs !

BOL lists the dispatching rules in the following order:

  1. Virtual machines are backed up by proxies running as physical machines that have SAN connectivity to the datastore.
  2. Virtual machines are backed up by proxies that are running as virtual machines on the same host.
  3. Virtual machines that are part of a vSAN stretched cluster are backed up by proxies that are in the same fault domain.
  4. Virtual machines are backed up by proxies that are running as virtual machines on hosts that have access to the datastore.
  5. Virtual machines are backed up by proxies that are on the same subnet as the virtual machine's host.
  6. The remaining virtual machines are distributed evenly across the proxies.

 but this was not true for one of my environments

we use nbdssl and not nbd to meet the requirement of end-to-end encryption of data. This also includes the encryption of data in-transit.

rgds
Klaus