Werk #927: windows agent: now able to evaluate logfiles written in unicode (2 bytes per character)
|Component||Checks & Agents|
|Title||windows agent: now able to evaluate logfiles written in unicode (2 bytes per character)|
|Date||May 28, 2014|
|Checkmk Editon||Checkmk Raw (CRE)|
|Compatibility||Compatible - no manual interaction needed|
The windows agent was unable to process logfiles which were written as unicode. Those files had binary zeros every other byte, rendering the normal "readline" logfile processing useless.
The agent can now read unicode files correctly, convert each line to a multibyte representation (most of the time it is only a single byte) and apply the configured logfile patterns accordingly.