A few Netapp dedup commands

I found these Netapp commands when I started digging into some dedup performance issues:

• > sis start -s /vol/$volume    
• This fixes very slow normal dedup operations due to this bug http://support.netapp.com/NOW/cgi-bin/bol?Type=Detail&Display=657692
• > sis status   
• Shows the status and a few settings of all of your volumes
• > df -s /vol/$volume   
• Shows some space savings
• > sysstat 1
• Shows you some basic performance stats updated once per second

Compare two files and show the additional lines in one that aren't in the other.

Quote often I have two sorted files and I want to "subtract" one from the other.  I've used a few different tools like perl, grep, and awk with success but they can be slow.

In typical linux fashion there is a tool that does exactly what I need very quickly called join.

Here is a quick example: join -1 3 -2 1 -v 1 < file1  < file2  > output

That is saying use column 3 on file1 ( -1 3 ), and column 1 on file2 ( -2 1 ).  It will then show only the lines that are in file1 but not file2 ( -v 1 ).

The files must already be sorted on those columns before you run join.

Way to create an rpm

I ran across a tool that looks very promising to turn a directory of files into an rpm.

It is called fpm and I'm starting to look into it now.