Re: file system overhead
Les Mikesell <lesmikesell(a)gmail.com> writes:
First make sure you are accounting for the disk space being
sold in 1000^3 GB instead of 1024^3.
I did, but it certainly is worth repeating.
Then add back the swap space that FC4 puts in a logical volume
instead of a physical partition that you'd notice with fdisk.
Bingo! I was naively expecting the swap space to be a normal file and
show up in df(1).
$ swapon -s
Filename Type Size Used Priority
/dev/mapper/VolGroup00-LogVol01 partition 102399992 0 -1
Crunching the numbers again I see that >97% is now accounted for.
Good enough.
Now to figure out how to lower the swap allocation...
Elliot Lee <sopwith(a)redhat.com> writes:
I know the maximum size of LVM volumes can depend on their chunk
size.
Maybe you are using the wrong chunk size in this particular LVM volume or
something.
Chunk size (and perhaps lowering the number of inodes) is something
else I guess I need to look into. From "df -i" I notice that only 1%
of the inodes are used.
-wolfgang
--
Wolfgang S. Rupprecht http://www.wsrcc.com/wolfgang/
Direct SIP URL Dialing: http://www.wsrcc.com/wolfgang/phonedirectory.html