Open missaugustina opened 7 years ago
I'm wondering if it would be worth parsing the logs and storing them into structured output or possibly archiving the structured output from one of our other services that will already be doing log parsing in the near future.
Is it enough to include all the logs in our backups, and have a long-ish retention?
Currently log rotation happens weekly or via a cleanup job and the logs are deleted. In order to determine when an issue started happening or to review details of events triggered in Datadog, we should have a long term archive of log data. This will likely get pretty big, so whatever solution we come up with should take that into consideration.