You want pfmon to send the data to STDOUT instead of a file, and then
pipe it to awk (as John suggested). If it's a "good" program, then by
not specifying a filename it will probably just default to STDOUT. Or
with some commands you can specify "-" as the file to mean stdout.
So, as John suggested,
pfmon | awk '{ SOME STUFF }' > output_file
Just make sure that pfmon without a filename parameter will spit to
STDOUT, or you'll have 100GB going somewhere that you weren't expecting
:)
On Mon, 2004-02-02 at 10:52, Sreekumar Kodakara wrote:
> Hi,
>
> I have a command (pfmon in IA64 system) which takes filename as a
> parameter. It samples the counters present in the processor at a given
> sampling rate and stores it into the file. This command stores lot more
> data than what I want. Since it stores it in column format, I can write a
> small awk script to parse the file to get the data which I want. The
> problem is that the raw file which I get from pfmon is just too big (
> about 100 GB or so) and hence I run out of disk space before I can process
> it with my awk script.
>
> My question is there a special file in linux ( The system
> is RH with kernel version linux-2.4.18-e.25, I dont know the RH version
> number), which I can specify in the command line of pfmon and which I
> can read simultaneously with a pipe or something so that I can process the
> data as and when it generates?
>
>
> Thanks for the help.
>
> Sreekumar
>
> _______________________________________________
> TCLUG Mailing List - Minneapolis/St. Paul, Minnesota
> http://www.mn-linux.org tclug-list at mn-linux.org
> https://mailman.real-time.com/mailman/listinfo/tclug-list
>
_______________________________________________
TCLUG Mailing List - Minneapolis/St. Paul, Minnesota
http://www.mn-linux.org tclug-list at mn-linux.org
https://mailman.real-time.com/mailman/listinfo/tclug-list