#logstash

after a few days of running perfectly, my logstash instance went up to using 900% CPU and then shortly thereafter died.

lots of messages in the log about "message not fully read"

and then finally, an out-of-memory error. I suppose I'll just give it a bigger heap, but i'm not sure why it would have been running fine up until now.

it was processing about 70kB of messages a second, but it had been up around 220 no problem.

anyone seen this before? ES was going pretty crazy in terms of CPU as well.

whack: poke

is this example still valid? can't get it to work; https://github.com/logstash/logstash/blob/maste... in my case it only matches the first line of "# User@Host: ..."

<http://goo.gl/byPlBo>; (at github.com)

Hey there, a lot of people us LS to output to mongo but has anyone got a filter for mongod logs to go into ES?

gme: why don't you directly output to ES at the same time?

andylee: go on...am i doing something really silly.. I purely want the mongo logs to be analysed / recorded in ES. But not interested in using mongo for input or output of ES.. does that make sense ???

gme: I have an use case maybe like what you are saying. Since both of them support jason type, I think you can mongoexport your data first, then forward them to logstash, and output to ES.

json, i mean :)

who maintain the debian packages for logstash?