SUMMARY of 'Sizes of filesystems compared to the size specified i n format....'

From: Coombes, Lorraine <>
Date: Wed Aug 29 2001 - 04:16:30 EDT
> I have a quick question that has baffled me for
> ages... 
> I work on a test environment so I am often building
> replicas of our live
> systems the one thing that confuses me is when I try
> to size a filesystem.  
>  I use format to specify the size I require in Mb,
> but I find that when I
> mount the filesystem I find it is smaller than the
> size I have specified.  I
> used to think the rule was that it kept 10% of the
> filesystem size, but this
> doesn't seem consistent.
> Is there any rule of what space it keeps that I can
> use to make sure I get
> my filesystems the right size the first time.
> Many Thanks
> Lorraine

I posted the above question and had many replies, so thankyou to you guys.
There are too many to list.  What a kind bunch you are.
I'll try to summarise ...

It seems there is no easy rule of working out the size but here are some of
the things that effect it..

Most responses mentioned that the amount held back was in relation to the
amount of inodes and these depend on the filesystem size.  You can adjust
how much space is held back using tunefs.

Cylinder Boundaries may also be affecting the size.  Below is some of the
responses I had.

Karl Vogel :

*	the bigger the filesystem, the more inodes are needed

*	man tunefs. You can specify how much space to hold back, plus how
   inodes to set up on a given drive.  If you have lots of small files,
   keep the default inode count.  If you have a few big files, drop the
   inode count.

Jay Lessert :

There are two sources of file system overhead:

1)  Number of inodes (newfs -i).  This is probably where your unexpected
    variation is coming from.  Unfortunately, mkfs uses a non-obvious
    and non-documented algorithm for actually deciding how many inodes
    to create.  As I recall, it's a function of cylinder group size,
    and I've run into situations where I had to lie about that (newfs -c)
    in order to get the number of inodes I wanted.

2)  Percentage free space (newfs -m).  This is probably the 10% you're
    thinking about, 10% is the default.

Koos van den Hout:
man newfs

          -m free        The minimum percentage of free space  to
                         maintain in the file system.  This space
                         is off-limits to normal users.  Once the
                         file system is filled to this threshold,
                         only the super-user can continue writing
                         to  the file system.  This parameter can
                         be  subsequently   changed   using   the
                         tunefs(1M) command.  The default is 10%.

man tunefs

     -m minfree     Specify the percentage  of  space  held  back
                    from  normal  users;  the  minimum free space
                    threshold.  The default value  used  is  10%.
                    This  value  can be set to 0, however up to a
                    factor of three in throughput  will  be  lost
                    over the performance obtained at a 10% thres-
                    hold.  Note:  If the value  is  raised  above
                    the current usage level, users will be unable
                    to allocate files  until  enough  files  have
                    been  deleted  to get under the higher thres-

Michael Horton:

The Solaris format command creates partitions/slices on cylinder boundaries.
This is determined by the hard drive's disk geometry.  It appears that you
are creating partitions/slices that do not fall on the cyclinder boundary
and the format command is making the appropriate changes (as it is suppposed
to do).


" This message contains information that may be privileged or confidential and 
is the property of the Cap Gemini Ernst & Young Group. It is intended only for 
the person to whom it is addressed. If you are not the intended recipient, you 
are not authorized to read, print, retain, copy, disseminate, distribute, or use 
this message or any part thereof. If you receive this message in error, please 
notify the sender immediately and delete all copies of this message ".
Received on Wed Aug 29 09:16:30 2001

This archive was generated by hypermail 2.1.8 : Wed Mar 23 2016 - 16:25:02 EDT