ADSM-L

Re: Filespaces Table Question

2002-08-08 11:21:58
Subject: Re: Filespaces Table Question
From: Thomas Denier <Thomas.Denier AT MAIL.TJU DOT EDU>
To: ADSM-L AT VM.MARIST DOT EDU
Date: Thu, 8 Aug 2002 11:25:58 -0400
> I was wondering if anyone has a good definition of the CAPACITY value in the
> FILESPACES table. On the export from the DB all it has for a definition is
> "Capacity (MB)". I have been experimenting with this value in hopes of
> finding a way to measure how much disk space is in use, for example select
> SUM(CAPACITY * PCT_UTIL/100) form FILESPACES. However, this gives unexpected
> results on some systems. On one TSM server this reports 9000 TB. When I put
> a time range on the query, the results change. On the server that reported
> 9000TB unbounded, it reports 4GB with a 24 hour time range.

As far as I know, the client software asks the operating system what the
capacity is and sends the information to the TSM server, which stores it
in the database for possible future use in responding to select and query
commands. Hence, the definition of the capacity is whatever definition the
client operating system uses. For our Unix clients the capacity figures
reported by the TSM server exactly match the results of running a 'df -t'
command on the clients and dividing the numbers of allocated blocks by
2048. I don't know exactly how other types of clients determine nominal
capacities. The only times in the filespace tables are the starting and
ending times for the most recent incremental backup of each file space.
This suggests that you have one or more off the wall entries for file
spaces that are not being backed up anymore. I would suggest running
some select commands to search for the off the wall value, such as a
select for all file spaces with claimed capacities over 100 gigabytes.
If you find such an entry, it may well be due to strange behavior by
a client operating system rather than by TSM.

> My overall goal is to computer a Tape Occupancy / Used Disk Space ratio for
> reporting. Perhaps there is a better way to do this?

<Prev in Thread] Current Thread [Next in Thread>