I’ve been using LogDNA on Heroku for a long time and love how well they’ve integrated it.
With Render, there are a few issues that make it pretty much useless:
Every new deploy creates a new source
This means you can’t create any views or any alerts because the source will change next time you deploy something
This then trickles down to generate other problems…
You can’t choose which apps to log data from, so you basically have to upgrade to higher plans and have a huge data mess as the “everything” tab contains literally every log from every active worker/server.
The easiest fix for Render would be to stop creating new sources when a new deploy is done
(and on a side note, render does a call to check if the app is running every second, which creates tons of useless logs getting rid of those would definitely be a good nice to have
Thanks for the feedback on logdna sources. I’ve passed this on to the rest of the team and also opened a conversation with logdna to see if there’s a workaround.
As I said, the Health Check is a side thing, the real issue is creating a new source at every deploy. It works in Heroku so there is a way but it depends on the way render is built of course
Hi @webbiger!
I had the same problem and I discovered that you can use labels to create views, because the source changes but the app name remains the same.
Example for an app named “test-app”, you can use label.serviceSlug:test-app