is there anything such as "resizing file for optimal disk usage"

if so, whats it about?

Hi,

I'm going to try to explain with an example...
If you have a file of 2049 bytes and if each block has a size of 2048 bytes, that file will occupy 2 blocks, wasting 4096-2049 bytes... got it ? With this in mind, if you have thousands of small files with a little more of 2048 bytes, you can try to reduce them and get some free disk space...

good example Jsilva.


This is really about determining what type of data you have on a particular filesystem... On systems with large amounts of data you will see the real impact of proper sizing of blocks.

However, In HPUX we dont use block size, we use extent-based sizing. All of our extents are divisible by 4mb... 4/8/12/16 etc... upto 256mb extents...

Many Filesystems in HPUX with mounted databases, use 64/128 for the extent size. My DBAs had me make my stripesize to 128mb b/c of the characteristics of their data.

The other thing about choosing the correct extent or block size is so that your number of writes are fewer...

If you have chunks of data... a small block size can be equally detrimental... causing you to have many more writes to those smaller blocks.

Choosing the right size is very important... Many vendors have recommendations that you can use to help determine what is right for your system.

__________________
My brain is your brain

thank you

but how do you go about reducing them. what command?

does this involve editign the file with vi or cutting it with wc???

Compress them with zcat, gzip, compress, zip...

If you have text files, you can edit them, just delete the lines you don't want... but it's a tedious job ! Compressing the files will be easier, cleaner and faster !