Sean Glover
Sean Glover
It might be worthwhile to log information as it's retrieved, instead of only once it's all been collected.
> I have it working with Confluent Cloud, for me the main reason behind the timeouts was the large number of consumer-groups (50K due to some other circumstances). > After...
@pkleindl Got it. I can see how that would be confusing if you're expecting to define all the client properties yourself. I can look into making the `kafka-client-timeout` app config...
Redis support as been released. For anyone experience JVM memory constraints you might want to give that a try.
When groups, topics, partitions are not available then in the subsequent scrape they are explicitly removed from the prometheus end point. This is done because there's nothing like a TTL...
> The consumer is logstash which is constantly scraping data from the MSK and sending it to ES. Are you sure that logstash is keeping the consumer group active indefinitely?...
I see. For some reason the request to get partition metadata for the `logstash` group is returning nothing (or it's filtered out by allow/deny lists in kafka lag exporter). You...
@sverrehu Thanks for reviewing some of these open issues. Yes. Many of these corner cases spurred ideas for new features, but unfortunately it's not something I can allocate a lot...
I just released `0.6.8`. LMK if that addresses it.
Ok. I created a ticket for Java 11. Maybe that will help. https://github.com/lightbend/kafka-lag-exporter/issues/286