Capacity license auditing using nbdeployutil ...possible and potentially significant errors ?
Can anybody else confirm this behaviour in Front End license spreadsheet generated by nbdeployutil?
From what I can see capacity license auditing using nbdeployutil can give possible and potentially significant errors if you have been doing some maintenance on policies.
I run nbdeployutil at the start of every month to track front end license usage. The report is based on one month's history. I allow for about 10% accuracy as it seems jump around a bit. The trend line is the important part.
From what I can see if you:
- move a server to a different policy
- split a large policy into smaller ones
- rename a policy
These have the potential to give you a misleading result because nbdeployutil doesn't have enough smarts to realise that it is actually counting the data a least twice and possibly multiple times. This probably can be seen in the Itemized tab of the capacity report.
nbdeployutil should never be used 'as is'. (See the 'Disclaimer' tab.)
Especially if Oracle and/or other databases are backed up via multiple policies and multiple log backups per day.
Or any situation where a client appears in more than one policy.I always recommend that Backup Admins use the Itemization column to look for duplicates/overlaps.
Duplicates can be removed from "Charged Size" column (with a note/reason added in the next column). Doing so will result in more accurate "Charged" size on the Summary tab.
