Hello
I am trying to input data of free -tm
in splunk. The raw data would look like
total used free shared buffers cached
Mem: 3820 3570 250 0 649 1695
-/+ buffers/cache: 1226 2594
Swap: 2047 0 2047
Total: 5868 3570 2297
And the data is being transformed into (its not properly aligned here, but it splunk it's in perfect alignment)
Memtype total used free shared buffers cached
Mem 3820 3615 205 0 629 2083
bufferscache 0 901 2919
But its not extracting the data properly. Didn't work even with forceheader=1.
I don't know if this makes any difference or not, but here is the code
#! usr/bin/perl
use warnings;
use strict;
my (@version,$version,@file,$file);
$version = `uname -a`;
@version=split(' ',$version);
unless ($version[0] eq "Linux")
{
#print "Version Not Linux\n";
exit;
}
@file = `free -tm`;
$file[0]=~ s/\s+/Memtype\t /;
$file[1]=~ s/://;
$file[2] =~ s/-\/\+\s+//;
$file[2] =~ s/\///;
$file[2] =~ s/:\s+/\t0\t/;
print $file[0];
print $file[1];
print $file[2];
Any help or suggestions??
I would transform your script to create key=value pairs results as follow:
Memtype=Mem total=3820 used=3615 free=205 shared=0 buffers=629 cached=2083
I would include a time stamp:
12/14/2012 00:00:00 +0000 Memtype=Mem total=3820 used=3615 free=205 shared=0 buffers=629 cached=2083
Then, Splunk will automatically extract each field.
Lp
I would transform your script to create key=value pairs results as follow:
Memtype=Mem total=3820 used=3615 free=205 shared=0 buffers=629 cached=2083
I would include a time stamp:
12/14/2012 00:00:00 +0000 Memtype=Mem total=3820 used=3615 free=205 shared=0 buffers=629 cached=2083
Then, Splunk will automatically extract each field.
Lp
It will be a good practice to include the time stamp in your script. In this way, you will have the time when the event was created.
Splunk was able to extract it even without the timestamp. Thanks for help. 🙂
How to I place the timestamp in it? Should this be done in the script itself? As of now the props.conf Date_config has been set to current. Would that be fine?