New Logging Infrastructure

Howdy folks. We just finished rolling out our new logging infrastructure. @steve will share the technical details soon, but tldr we replaced our central Graylog cluster and now stream logs directly to Elasticsearch via Vector.

At first you won’t notice much different aside from logs being faster and more reliable – specifically less delays, dropped logs, and duplicates. Later on this new system will let us offer some neat features, such as direct query support, Kibana access, customizable retention, and configurable log drains to Papertrail/Elasticsearch/etc.

Give it a try with flyctl logs and let us know below if you have any issues or feedback. Cheers.

7 Likes

awesome! Thanks @michael and @steve

1 Like

Yay! This is great!

@michael How do we set up log drains now? I’d love to configure log drains on our setup to DataDog.

1 Like

Hey, @mike! This feature isn’t actually ready yet but it’s something that we’ll be working on and this infra upgrade has made possible :slight_smile: We’ll let you know when it is though!

4 Likes

Resurrecting this year old thread but this is just a feature recommendation, so it should be fine. I think logs would be a very neat addition to the web interface, along with metrics for quick observability to the apps.

Sometimes, in a pinch, where flyctl isn’t available, it can be really helpful to check out the logs. No fancy interface needed. Something as bare bones as dump of last 500 lines in <pre> would likely suffice as a great first iteration.

1 Like

This would be a neat addition! We just hired somebody to finally give the web UI some love and part of the hiring project was actually designing this. Hopefully we have something to show in a few months.

1 Like