Quote:
Originally Posted by
solaris_1977
2- How do I track, at what level it is failing ? I checked AWS instance and it was always up and no downtime on that Centos linux.
I don't think it is a matter of tools but a matter of organised debugging: let us first consider what could have gone wrong:
1) the server('s OS) - you ruled that out
2) the application software, i.e. Apache and/or whatever works on top of it.
3) the network connection of your server: possible reasons include network congestion, broadcast storms, intermittent hardware outage, ...
4) the connection between you and your server: caching mechanisms like Cloudflare may influence the connectivity until the cache is filled.
This off the top of my head list is probably neither complete nor detailed enough. You are welcome to edit it until it fits your environment. Once you have done that you start ruling out one point or sub-point after the other: for instance the application stack you use could be tested by a client working from within the server automatically and so foregoing the network connections otherwise necessary. Once you have established that you move on to the next point in the list.
Debugging is just the organised application of logic and a few usually rather simple tests once you have properly envisioned how things are supposed to work and what depends on what.
I hope this helps.
bakunin