Requirements include . . .
- Aggregate log data from files, applications, locally, over the network
- Functionality can be enhanced through plug-ins or other mechanism
- Can extract errors from log data
- Extracted errors can be handled and resolved as in an issue tracker
- Provide visualizations of log data
I'm preparing to work on standing up a log aggregation and error monitoring system for multiple servers and applications. I previously investigated the combination of Logstash and Sentry for this purpose and was planning on implementing it. However, for a couple reasons, I'd prefer using a Go-centric solution if there were a comparable one available.
The first reason is the environment. I'd like to keep things as simple as possible. We already have Java apps all over the place, so running Logstash wouldn't be a problem. Sentry, however, would be the only Python application I'd have to support.
Some people suggested Packetbeat, InfluxDB, and Grafana. They look very full-featured and impressive, but it looks like none of those does anything like Sentry does in the way of error management. In fact, I don't know of any other open-source apps beside Sentry that provide this functionality. If there were an alternative to Sentry, even if it weren't written in Go, I'd be interested in trying it out.