ELK stops accepting logs
-
Ran a plugin update. Will see if that has any effect.
Updated logstash-codec-json 2.1.0 to 2.1.1 Updated logstash-filter-dns 2.0.2 to 2.1.0 Updated logstash-filter-json 2.0.3 to 2.0.4 Updated logstash-input-beats 2.1.3 to 2.2.0 Updated logstash-input-rabbitmq 3.1.4 to 3.3.0 Updated logstash-input-redis 2.0.2 to 2.0.3 Updated logstash-output-elasticsearch 2.5.1 to 2.5.3
-
During the command
bin/plugin update
, received the errorError Bundler::InstallError, retrying 1/10 An error occurred while installing manticore (0.5.5), and Bundler cannot continue. Make sure that `gem install manticore -v '0.5.5'` succeeds before bundling.
-
Lots of reports on this, although I have a feeling that it might not be a leak but might be just "how much memory Logstash needs."
Try setting it to something high and see if it stops expanding.
-
@scottalanmiller said:
Lots of reports on this, although I have a feeling that it might not be a leak but might be just "how much memory Logstash needs."
Try setting it to something high and see if it stops expanding.
I already went from 1g to 2g. I could be off base, but I suspect a memory leak b/c it runs for days before the OOM condition occurs. The only active feed is from my XO server, and that activity should be light.
-
Found this current Github discussion that describes the plugin update issue I ran into.
-
@Danp said:
@scottalanmiller said:
Lots of reports on this, although I have a feeling that it might not be a leak but might be just "how much memory Logstash needs."
Try setting it to something high and see if it stops expanding.
I already went from 1g to 2g. I could be off base, but I suspect a memory leak b/c it runs for days before the OOM condition occurs. The only active feed is from my XO server, and that activity should be light.
That would be the expected behaviour whether a leak or just hasn't fully grown yet.
-
Finally back up and running...
-
-
@scottalanmiller Reinstalled ruby-maven and manticore.
-
No issues in almost a week. I will probably drop the heap size back to 1g and see if it remains stable.