Howto prevent ntopng causing out of disk space and inodes?

Question

I've been trying the ntopng software from http://www.ntop.org/get-started/download on a CentOS box at work to monitor what is happening on our network. The problem I'm having is that I can't find any part of the software that gives an option to delete old data.

The disk usage by this software was so uncontrolled it used 79% of the space and 87% of the inodes on the disk in that server after a few months.

Has anyone seen a solution for having only the most recent xx days or GB of data when using ntopng? Also ensuring it doesn't run out of inodes as well. Can/should I just delete any old files and empty directories in its rrd data directory?

Solution

I've encountered a similar problem. One day df -ih shows that there is no free inodes on root partition. OS is Ubuntu 12.04.

After some research I found that there is a lot of small .json files in /var/tmp/ntopng/\*/top_talkers/ which contain information about the most active consumers of traffic.

We don't need that much history (I've seen some files which are 1 year old) so I decided to delete all files older than 2 months:

find /var/tmp/ntopng/*/top_talkers/* -mtime +60 -delete

Also I’ve added task to the /etc/crontab:

@monthly    ubuntu  /usr/bin/find /var/tmp/ntopng/*/top_talkers/* -mtime +60 -delete

Now server uses only 55% of inodes.

Comments

[user9517]

How quickly did it use up the space ?

[BeowulfNode42]

So you don't think there's any related meta data that needs updating. Have you noticed any adverse effects from just deleting these files?

[dima.butyrin]

So far, so good. I'm not sure is there any related metadata or not, looks like everything is work as usually.

[tigerjack]

For me, the most used directories are the */rdd ones. Do you think it's safe to use the same command (maybe with a longer mtime)?