Now I'm trying to have some practice with docker and working on lightsquid containerization. There is a practical task of setting up several Squid servers for different purposes so I realize that there will be a need to have different lightsquid containers. I'd like to create as more common container as I can and my proposed solution here is:
- use lightsquid container only as interpreter of input data to output data;
- map one part of input data (configs) to docker host and use local git to maintain it;
- another part of input data (squid logs) will be generated by respective squid container;
- output data (lightsquid reports) will be mapped to respective docker host directory (and backuped/archived).
I attached a picture to make more sense on this question.
So wouldn't you like to share your opinion about this scheme: is there any concerns or it is a standard way of containerization?
Each answer will be greatly appreciated!
/srv/squid-container/etc
,/srv/squid-container/log
etc. Because if you maintain a system as proposed, you are interleaving the regular system files with docker-related files which you will have to remember (or document e.g. with the picture) more thoroughly to not foreget parts if you are e.g. doing a migration or "uninstalling" the container.