import pyhdfs
from cStringIO import StringIO
import binascii

-snip-

#Set hdfs connection info
hdfsaddress = “namenode”
hdfsport = 12345
hdfsfn = “filename”

#gzip compression level
clevel = 1

-snip-

      logger.info(“Writing compressed data into ” + hdfsfn + “.gz”)
        #open hdfs file
        fout = pyhdfs.open(hdfs, hdfsfn + “.gz”, “w”)
        #compress the data and store it in compressed_data
        buf = StringIO()
        f = gzip.GzipFile(mode=’wb’, compresslevel=clevel,fileobj=buf)
        try:
                f.write(concatlog)
        finally:
                f.close()
        compressed_data = buf.getvalue()
        #write compressed data into hdfs
        pyhdfs.write(hdfs,fout,compressed_data)
        #close hdfs file
        logger.info(“Writing task finished”)
        pyhdfs.close(hdfs,fout)
-snip-

!/usr/bin/perl
#Check flume DFO directory size

sub trim($);
use strict;
use warnings;

my $exit=0;
my $backlog_size = `du -s flume | awk {‘print \$1’}`;
$backlog_size = trim($backlog_size);

if ( !$ARGV[0] || !$ARGV[1])
{
########################### Usage of the plugin
print “check_flume_backlog critical_size warning_size   \n”;
exit 0;
}

######################### Case 1 if State is Critical
if ($backlog_size > $ARGV[0])
{
print “Critical: “.$backlog_size.”b\n”;
exit 2;
}

######################## Case 2 if State is Warning
if($backlog_size > $ARGV[1] || $backlog_size == 0)
{
print “Warning: “.$backlog_size.”b\n”;
exit 1;
}

######################## Case 3 if State is OK
if($backlog_size < $ARGV[0] && $backlog_size < $ARGV[1])
{
print “OK: “.$backlog_size.”b\n”;
exit 0;
}

sub trim($)
{
        my $string = shift;
        $string =~ s/^\s+//;
        $string =~ s/\s+$//;
        return $string;
}