Got me on the ruby part. Meant to say JVM. My point was more to the fact that their site's incredibly short on examples. FWIW, I inherited a thrice-removed in

Got me on the ruby part. Meant to say JVM. My point was more to the fact that their site's incredibly short on examples. FWIW, I inherited a thrice-removed infrastructure in a latency-sensitive environment. It was fun ;) Logstash was installed on a couple machines the night before I was going to load some configs from my machine. I didn't load anything yet, except that logstash-web was running at 2s uptime. For some reason, (with upstart's help), the logstash installation included a broken webservice with a bad config. It got stuck in a JVM-birthing restart loop that was enough to cause significant impact. It happens. (Funny enough, same place used mongodb. We ran into an issue once where database cleanups failed once mongodb reached around 50% of the disk. Local copy cleanups are fun.)