File System Learning Approach

[Valdis.Kletnieks at vt.edu]

On Mon, 15 Apr 2013 16:24:17 +0200, Pietro Paolini said:
> I have to compare different filesystems options for an embedded platform

Step 0: Get an accurate list of what requirements the embedded platform
has.  Is the storage flash based? Spinning oxide?  What are the relative
read/write ratios (some systems will have almost no writes, but a Tivo
will have almost as many writes as reads (possibly more, even).  What
sustained rates of file I/O and metadata activity will you expect?
Do you have a fairly high CPU budget and can afford more cycles, or are
you using a slower ARM and have to do it with as few cycles as possible?
How big will the filesystem be?  How many meg/gig of data, and how many
files?  Small numbers of honking-huge files, or zillions of 139 byte files?
Does the filesystem have to support TRIM or whiteouts?  Plus any application
specific issues I may have overlooked...

Step 1: for i in $filesystems; do compare_specs($needed_specs, $i); done.

You get the idea.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 865 bytes
Desc: not available
Url : http://lists.kernelnewbies.org/pipermail/kernelnewbies/attachments/20130415/f0b2893d/attachment.bin