At the time of writing, we suggest these troubleshooting steps for many I.T. tasks. Some of these will not apply to every problem. But if you do not know where to begin, this can help you with "DevOps block."
1. Find the word "error" in the logs. Look at a log with vi (e.g., copy the log to a location where it will not be continually written to). Open the copy with vi. Run this command::set ignorecase
Search for the word "error"
2. Reboot the server and try again.
3. Find one server or program that works, see how it is different from the one that fails.
4. Start over from the beginning (redeploy).
5. Find what changed since it was last working. If it never worked, try to deploy it on a different distribution of Linux.
6. If the problem is reproducible, try to isolate the problem. A back-to-basics approach can involve reducing problems to atomic parts. From there you can diagnose an underlying component having a problem and thus find a solution.