[syslog-ng] duplicate lines

Peter Straka thebeno at zoznam.sk
Sun Jun 5 19:29:46 CEST 2005


Hello,
I would ask you if it is posible to write filter which remove every
duplicate lines from one source

usage as: log ( source; filter(duplicate); destination)
parameters:
       number of lines which can be betwen the same line (this is becouse
some messages are repeated in a block of lines)
       time period which will defined  when the counter must be reset
(before that counter will be send)  and next line will be again the first
one
       counter - to count number of droped lines and show this value every
time period - like solaris syslog do

duplicate line means :
1 line: aaaaa bbbbbb
2 line: cccccc
3 line: aaaaa bbbbbb
4 line: cccccc
... line: aaaaa bbbbbb
... line: cccccc
9 line: ddddd

lines 1,2,9 will pass to destination
lines 3-8 will be removed (droped)

in the past it was a filter - DEFAULT which did this job in a "some" way
and I thing now it is nothing which can be usable

any idea?
with regards

Peter STRAKA


More information about the syslog-ng mailing list