Howdy,

as you can see on the attached screenshot my IO is significantly higher since June 30th (16:00). I have no clue what causes this, so I started monitoring it hoping I could solve it. In Longview I can see that there are several services with increased IO after that date. They are: atd, fail2ban, ntpd, php5-fpm, pickup, qmgr, and rsyslogd. Quite random if you ask me.

Any pro that can help me get a fix for this?

https://www.dropbox.com/s/ajbmze2zt5oml ... 3.png?dl=0

First of all, if that graph tops out at 171 ops per second, it's not a problem at all. Your server did very little IO to begin with, and it still does very little IO.

The difference may very well have been caused by a change in the measurement method. A new kernel could do that. Moving to KVM could do that. A simple update to the measurement script could do that. In any case, there's nothing to worry about.

@hybinet Thank you for taking the time to respond to my question. Luckily I wasn't to worried, but it's good to hear it from some else as well.