In addition to preventing gross fragmentation as the file system nears full, it
also reserves storage for root and other, privileged processes. Reducing it
below 5% is not recommended, sans if the file system is very large and your
files are typically not that big.
It's really a consideration of:
- The size of files typically written to the file system
- How much reservation do some operations need if the users fill it up
I'm sure there are 1-2 other considerations I'm not thinking of.
________________________________
From: "Langley, Morgan (GE Capital)" <[email protected]>
To: Red Hat Enterprise Linux 5 (Tikanga) discussion mailing-list
<[email protected]>
Sent: Thu, November 4, 2010 8:54:12 AM
Subject: Re: [rhelv5-list] df is inaccurate
df is inaccurate
As other people have noted, five percent are reserved on an ext2/3/4 filesystem
by default. It is a good idea to leave at least 5% of reserved blocks on the FS
to prevent fragmentation. 10% would be more efficient for this purpose but if
you know what you are doing, you can reduce this reserved blocks count. (use
tune2fs –m)
Regards,
Morgan
________________________________
From:[email protected] [mailto:[email protected]] On
Behalf Of Nick Lunt
Sent: Thursday, November 04, 2010 1:41 PM
To: Red Hat Enterprise Linux 5 (Tikanga) discussion mailing-list
Subject: Re: [rhelv5-list] df is inaccurate
Possible I suppose, but this happens on every rhel server I've ever worked on !
From:[email protected] [mailto:[email protected]] On
Behalf Of Langley, Morgan (GE Capital)
Sent: 04 November 2010 12:24
To: Red Hat Enterprise Linux 5 (Tikanga) discussion mailing-list
Subject: Re: [rhelv5-list] df is inaccurate
Hmm, maybe you have some reserved blocks which are not showing up in df? I have
not checked this though…
________________________________
From:[email protected] [mailto:[email protected]] On
Behalf Of Nick Lunt
Sent: Thursday, November 04, 2010 12:57 PM
To: Red Hat Enterprise Linux 5 (Tikanga) discussion mailing-list
Subject: [rhelv5-list] df is inaccurate
Hi folks
every now and then I seem to get a flurry ofusers asking me "why is 'df'
inaccurate?".
I've never really understood why though, so anyone know ?
Eg:
$ df -klP /
Filesystem 1024-blocks Used Available Capacity Mounted on
/dev/mapper/VG00-LVol00 4031680 2204200 1622680 58% /
$ bc
bc 1.06
Copyright 1991-1994, 1997, 1998, 2000 Free Software Foundation, Inc.
This is free software with ABSOLUTELY NO WARRANTY.
For details type `warranty'.
4031680 - 2204200
1827480
So either df is wrong, I'm blind, or bc can't do take aways ;)
Anyone know why df is wildly inaccurate ?
Cheers
Nick .
__________ Information from ESET NOD32 Antivirus, version of virus signature
database 5590 (20101104) __________
The message was checked by ESET NOD32 Antivirus.
http://www.eset.com
_______________________________________________
rhelv5-list mailing list
[email protected]
https://www.redhat.com/mailman/listinfo/rhelv5-list
