Open phoppe93 opened 5 years ago
Dear Logstash Team.
i have an issue on Kubernetes with logstash and graylog.
The logstash only connected via tcp one times to graylog.
Here a snipped how i connect to different graylogs inside a cluster
output { if ([rand_int]) { if( [rand_int] == 0 ) { gelf { id => "output.gelf.0" codec => "plain" protocol => "TCP" host => "graylog.logging.svc.cluster.local" port => 12201 sender => "%{source_host}" ignore_metadata => ["@timestamp", "@version", "severity", "host", "source_host", "short_message"] } } else if( [rand_int] == 1 ) { gelf { id => "output.gelf.1" codec => "plain" protocol => "TCP" host => "graylog.logging.svc.cluster.local" port => 12201 sender => "%{source_host}" ignore_metadata => ["@timestamp", "@version", "severity", "host", "source_host", "short_message"] }
Second Issue:
The graylog container are scalable on cpu usage.
After the scale up and one more graylog container is available the logstash not try to connect to the new container.
I have to restart the logstash container.
Is it possible to to have the option to reconnect at an interval and look for new graylog containers ?
Thanks
Could you not just put a load balancer in front of your graylog containers?
Dear Logstash Team.
i have an issue on Kubernetes with logstash and graylog.
The logstash only connected via tcp one times to graylog.
Here a snipped how i connect to different graylogs inside a cluster
output { if ([rand_int]) {
if( [rand_int] == 0 ) { gelf { id => "output.gelf.0" codec => "plain" protocol => "TCP" host => "graylog.logging.svc.cluster.local" port => 12201 sender => "%{source_host}" ignore_metadata => ["@timestamp", "@version", "severity", "host", "source_host", "short_message"] } } else if( [rand_int] == 1 ) { gelf { id => "output.gelf.1" codec => "plain" protocol => "TCP" host => "graylog.logging.svc.cluster.local" port => 12201 sender => "%{source_host}" ignore_metadata => ["@timestamp", "@version", "severity", "host", "source_host", "short_message"] }
Second Issue:
The graylog container are scalable on cpu usage.
After the scale up and one more graylog container is available the logstash not try to connect to the new container.
I have to restart the logstash container.
Is it possible to to have the option to reconnect at an interval and look for new graylog containers ?
Thanks