Directory

Dears,

That is almost half year, when this plugin is in production monitoring.
Yesterday, monitoring server was found died... halted because ... inodes finished.

# df -i
Filesystem Inodes IUsed IFree IUse% Mounted on
/dev/mapper/VolGroup-lv_root
13912080 13912080 0 100% /
tmpfs 1005525 1 1005524 1% /dev/shm
/dev/sda1 128016 62 127954 1% /boot

yep, that is 13,912,080 millions of inodes used

The problem found:
plugin invokes libnet-openssh-perl library which created files in uesr's (nagos) home directory, which created file for each performed check:

# ls -la /home/nagios/
drwx------ 2 nagios nagios 853970944 Mar 22 12:26 .libnet-openssh-perl

it took some time to count:

.libnet-openssh-perl]# time find . | wc -l
13741298

real 1m55.035s
user 0m8.142s
sys 0m14.129s
.libnet-openssh-perl]# time /bin/ls -f | wc -l
13741299

real 0m36.880s
user 0m5.900s
sys 0m8.820s
[root@chq1nms1 .libnet-openssh-perl]#


Fixing:

# rm -rf .libnet-openssh-perl

# df -i
Filesystem Inodes IUsed IFree IUse% Mounted on
/dev/mapper/VolGroup-lv_root
13912080 172097 13739983 2% /
tmpfs 1005525 1 1005524 1% /dev/shm
/dev/sda1 128016 62 127954 1% /boot

Keep in mind (after integrating).

Plugin itself is excellent!

BR to developer and to users,

/A


P.S. after 30 min after fix:

.libnet-openssh-perl]# time /bin/ls -f | wc -l
2008

real 1m34.013s
user 0m0.004s
sys 0m3.130s