Analyse MySQL diskusage
20. March 2014

It may happen that your mysql instances face some diskspace problems. Usually this is an application-side problem. Following steps may help to identify the bogus points. Show databases and their sizes SELECT table_schema “DB Name”, Round(Sum(data_length + index_length) / 1024 / 1024, 1) “DB Size in MB” FROM information_schema.tables GROUP BY table_schema; Display table information […]


If you use curl to get some json in your shell and try to read what you got, this is no fun. I discovered jq as a very nice commandline tool to format json output. Fetching graphite time-series data is a good example on how to use jq. curl -s “http://graphite_host/render?*&format=json&from=-5min” | jq ‘.’ [ { “datapoints”: […]

How to create Ansible deb packages
10. March 2014

Ansible is a quite nice tool for ssh based systems management and orchestration. It is written in python and hosted on github. Quite easy to install by using pip or install, but creating a deb package out of the sources is much better in terms of manageability. To do so, just do the following: […]

Python and oracle databases
21. February 2014

As oracle databases are not that easy stuff, it seems normal that connecting to one is also not that easy. Doing the following might help. Install Oracle Instantclient (on OSX) This brought me into some trouble starting with simply trying something with something downloaded from oracle :-). The following howto helped me a lot on […]

compacted output for pdsh
12. February 2014

Using pdsh is some kind of a knife with latches on both sides. Nevertheless it is a great tool to do things really fast on a lot of systems. It is able to execute ssh commands in parallel on many systems but also produces a overblown output, fairly bad to read. To get a consolidated […]

Interesting facts about rpm scriptlets
16. January 2014

While I build rpm’s very often, there are always some facts I have to lookup on different sources. This article is indended to combine some interesting facts about rpm scriptlets to save time and maybe help somebody. Order of execution It is always a good idea to know about the order of execution to prevent […]

Optimizing Debian for SSD storage
20. November 2013

The default I/O scheduler is optimized to schedule requests in an equal manner. Since SSDs are even faster in reading than writing, deadline scheduler favours read requests. This helpts to increase the throughput in some circumstances.

Debian based OpenVPN Router Applicance
16. November 2013

I use PC Engines Alix boards as VPN Appliances intended to connect some branch offices. They are rock solid, work for years without any problem and bring the flexibility one needs to implement all the nice stuff. 3x Lan Interface, 2x USB, passive cooling. Quite nice. Complete sets are available on Amazon or at some […]