17 Posts
0
950
Networker vProxy to DD policy running successfully but nothing happens.
I have set up Networker in our VMware demo environment using the Networker OVF and the vProxy OVF. It connects to a Data Domain using a boost account in an snmp community. The vProxy is running fine and accessing the individual VMs. I've configured a vProxy policy and workflow with a backup action to send one of the environments VMs to the DD system and all seems fine.
When I run the workflow and check the only messages it reads is "starting workflow" and "workflow succeeded". In the policy_notifications.log file it says its duration was 0 hours/mins/seconds and when I try and run the same workflow from the vcenter plug-in it says "successfully started policy" and then nothing shows up in the tasks list.
The policy_notifications.log file also reveals that the NVE backup which comes preconfigured on the system has been failing with 0 successful backups.
Any ideas why this might be happening?
barry_beckers
393 Posts
0
October 30th, 2021 02:00
Ideally yes.
By default you'd see that the vproxy you have configured would have been set to use hotadd (13 sessions) and have NBD set to 0. So it will only be able to backup vm's that it has direct access to, so to its datastore(s) that the vm vmdk's are located on.
We always deal with esxi clusters in which all nodes have access to the same datastores, so that one vproxy would be able to do the backups.
However other esxi clusters would have access to other datastores and hence would require at least one vproxy to be located on that esxi cluster.
We now changed our design so that we'd have even two vproxies on each multinode esxi cluster, to have a failover in case one vproxy is not working and also to be able to perform maintenance (for example an upgrade (or rather it is a redeploy using a newer ova nowadays completely integrated from nw19.3 or so onwards)) on a vproxy without affecting backup/restore.
So you might wanna deploy more vproxies, one within each esxi cluster, to prevent too much backup data flowing between esxi nodes in one cluster to the vproxy on another when you'd use NBD. Would require 10Gb connections also according to Dell recommenations. But even regardless we don't use NBD and only use hotadd.
I would have expected backups to fail in case it wouldn't be able to access the remotely located vm's as NBD would be set to 0?
Possibly that vproxies might not even be required in the future as for example Dell PPDM already supports vproxyless vm backups by hooking directly a plugin into esxi hosts themselves (aka transparent snapshots).
bingo.1
2.4K Posts
0
October 23rd, 2021 08:00
Hard to say because we do not know anything else but "it works - but there is no output". It seems as if the processes all run fine but there is an obvious configuration error which tells NW not to start any backup.
May I suggest that you carefully check each separate log file to see where the obvious 'zero-action' is happening. If you do not see an obvious issue, then it would make sense to start the process with increased verbosity or a higher debug level.
barry_beckers
393 Posts
0
October 26th, 2021 02:00
As it is a demo environment, whoch nw nve did you deploy?
Also it should definitely show more than that. The thing is to set up a policy a lot of things have to be correctly setup, so defining the VC you connect to, with a user id and require large list of permissions, the policy, worjflow, backup action, schedule and so on... like for example is the workflow enabled (not just autostart enabled)?
So nmc log should show some info of the run workflow, as it simply shows the contents from the logs underneath /nsr/logs/policy/ / or the underlying directory .
RusscsCarter
17 Posts
0
October 29th, 2021 12:00
Upon further work I've found that its only failing to backup virtual machines that are on other hosts than itself and the proxy node. Is this normal? Do I just need a networker proxy on every esxi host that I intend to backup vms from?