I recently redeployed a FAS3040C from one facility to another. This process involved completely wiping out the filer and rebuilding it from scratch, including reinstalling DOT from scratch.
I deleted the appliances form Operations Manager, but when I rediscovered them (using the new names), Operations Manager displays them with the OLD names. I know the old names only exist in Operations Manager, nowhere on the filers themselves. For the life of me, I can't figure out how to truly delete the old filers in Operations Manager and start from scratch.
I had a very similar issue where I redeployed an old FAS2020 and even after I deleted the old object out of DFM, it kept all the existing info when rediscovered..
The steps from the KB article, https://kb.netapp.com/support/index?page=content&id=1011876 , worked perfectly, first I changed the old object's IP to the current newly deployed filer and then renamed it it:
1 - dfm host set [hostname] [hostPrimaryAddress]=new-address
2 - dfm host rename old-hostname new-hostname
OpsManagers tracks an appliance through three items - host name, IP address, and its System ID. If only one of these changes, it automatically detects the change (except for primary address) and updates its database accordingly. So all the historic data pertaining to a particular filer will continue to remain valid even afte those changes.
However, there are situations (like this thread) in which the user wants to forgo without the historic data. For such situations, a change of atleast two items from the above list of three is needed. As the filer name is already changed here, you can try changing the IP address too, then add the filer with this new name+IPAddress. This will dis-associate all the historic data of that particular filer (though it is not removed from database).
Please refer the FAQ: http://now.netapp.com/NOW/knowledge/docs/DFM_win/rel371/html/faq/index.shtml#_9.10
If needed you can also purge all (historic) data for a filer from the database, but make sure you have a (recent) backup of your database in case something goes wrong.
* write down the host Id's to be removed.
dfm host list -a
* stop dfm and start with only the sql service
dfm service stop
dfm service start sql
dfm service list - only sql should be started
* delete the host, this can take a while
dfm host delete -f <database_ID>
* start dfm again
dfm service start
dfm service list
OK, so I tried this. I'm trying to retire old filers from Ops Manager.
I removed, using this method, a filer with over 100TB capacity from operations manager. I'm still waiting for the change to be reflected in the Group Status Summary page. The graph hasn't dipped down, all the numbers are the same as before. It's been about an hour since I deleted the filer.
Thoughts? Not patient enough?