Hello, Is it possible to have multiple data categories and data points in a "Generic Program with Data" probe similar to the built in "Linux: TCP Connections by State" ? For example: I wrote the following perl script to extract read IO/s from monitored hosts into data points: --SNIP-- #!/usr/bin/perl open(IOSTAT,"-|","/usr/bin/iostat -xtd sda 15 2") or die; @collect = <IOSTAT>; close IOSTAT; @collect = grep {!/^$/} @collect; @data = split(/\s+/,$collect[-1]); $read_ops = $data[3]; print "<perldata> <hash> <item key=\"data\">$read_ops</item> <item key=\"Read Ops\"> </item> </hash> </perldata>"; --SNIP-- Converted here into a "Generic Program with Data" probe (1 liner) --SNIP-- /usr/bin/perl -e 'open(IOSTAT,"-|","/usr/bin/iostat -xtd sda 15 2") or die;@collect = <IOSTAT>;close IOSTAT;@colle ct = grep {!/^$/} @collect;@data = split(/\s+/,$collect[-1]);$read_ops = $data[3];print "<perldata> <hash> <item k ey=\"data\">$read_ops</item> <item key=\"Read Ops\"> </item> </hash> </perldata>";' --SNIP-- What I would like to do is create more data points out of different streams of data from the -1- probe. In my script the iostat output captured in @data has read io/s as element [3], while write io/s are captured in element [4] (etc). Going back to my reference to the Linux: TCP connections by State probe; in that you can specify multiple data point sources: eg, ESTABLISHED, TIME_WAIT, SYN_RECV etc. This would be preferred as then I could process all the data from the iostat probe and create very powerful reports. I just dont know if that is possible as my current experiments have yielded no success.