I'm hunting for a method to automate the monitoring of orphaned snapshots left behind after VMware VADP backups.
There is a reference in this NBU 7.1 Emergency Engineering Binary Guide
On page 17 it says "A tool has been added to report orphaned snapshots that remainon the client."
Anybody have any ideas what this tool is and is it documented anywhere?
If its not for VMware anybody know of a reliable method?
have no idea where that tool or where it is integrated into NBU. This might not relate to 'VMware' snapshots.
I use a great tool every now and again (use the Health tab)
Shows orphaned snapshots - zombie VMDK's. Zombie VM's.
Plus much more.
There is also this KB from VMware to setup alerts:
Configuring VMware vCenter Server to send alarms when virtual machines are running from snapshots
The only thing i can think is that it is related to bpfis
As the VMWare backup hiost runs the bpfis process then perhaps it can see the orphaned ones in VM?
If so then you can do the following from a command line:
This displays any acive snapshots created by bpfis
You can then use:
bpfis delete -id snapshot_id_found_above
Dont have a failed VM available to try this out but maybe this is what they are referring to?
It's a new way to invoke bppficorr - bppficorr -report -clientlist <snapshot client list file> - and no, it doesn't appear to be documented anywhere.
This command will not only identify orphaned snapshots, but will also attempt to delete them. MAKE SURE YOU KNOW WHAT YOU'RE DOING IF YOU DECIDE YOU'D LIKE TO TRY THIS.
Here are the instructions from the original EEB (which I have edited a bit):
1) Create a client list file on master server (e.g. /tmp/cllist.txt) which contains <client name>:<policy name> pairs for the policies on that client.
- The each pair sould be disjunct set i.e. there should be only one entry per client per policy.
e.g. Entries in client list file could be like -
hosta(client name):IR_flsnap_hosta(policy name on this master server)
- Only IR policies can be used
- The policies being used here must be the policies from this master server where we are executing bppficorr on.
2) Create admin log directory on the master server if not present
3) Execute the bppficorr command as follows -
# cd <NBU install path>/openv/netbackup/bin/admincmd/
# bppficorr -report -clientlist /tmp/cllist.txt
report file path: openv/netbackup/logs/admin/bppficorr_report_file.txt
Num images 0 for client hosta policy IR_flsnap_hosta
Num images 0 for client hostb policy IR_flsnap_hostb
4) Check the report file contents for the identified IR orphan snapshots. The entries will be in format =
client_name:policy name:snapshot id
5) Also verify if deletion of those snapshot ids on the client was successful or failed.
Other people could probably tell you if this is for VMware or not - I'm honestly not sure. I also feel like I should know what "IR" means above - "Instant Recovery," right? My educated guess is that what the instructions are saying above is that this applies only to policies for which Retain snapshots for Instant Recovery has been enabled.
If you need to know more, I'm sure we could find somebody who actually knows the answer. ;) Let me know.
I discussed this with support this morning. They checked the stuff in the technote and said it was for NetApp snapshots not VMware and relates to specific problem/client.
I tried BPFIS QUERY It was weird :
It appears to only give snapshot information on the media server and only about the media server where the command was run (only tested a couple of media servers)
Also the results don't make sense. It gave me a list of snapshot IDs and when the ESX host was checked for deltas there weren't any.
I checked on of the IDs and it returned information on a snapshot that was months old.
bpfis query -id FQDN_client_132765483
INF - BACKUP START 10588
INF - Snapshot method: VMware
INF - Snapshot image : LOCALHOST
INF - Snapshot image owner : NBU
INF - Time created : 1327654839 -> Fri Jan 27 20:00:39 2012
INF - REMAP FILE BACKUP NEW_STREAM
INF - REMAP FILE BACKUP ALL_LOCAL_DRIVES
INF - EXIT STATUS 0: the requested operation was successfully completed
with NetBackup 184.108.40.206 and 220.127.116.11, Netbackup tries to consolidate all orphaned snapshot by itself during the next backup if it could figure out if there is any need to consolidate orphaned snapshots.
you can create an alarm in VC if your VM is running on snapshots.
refer this if you have not seen this before: