feat(l1): docker compose parallel snapsync run in loop

[rodrigo-o]

Motivation

Have a single server continusly syncing hoodi, sepolia and mainnet

Description

This PR takes inspiration in the already present server_runner.py and creates a new docker_monitor.py accompanied with make targets and a new docker compose that spawn by default 3 nodes in parallel, hoodi, sepolia and mainnet, and monitors them in the following way:

• Monitor snapsync time (with a timeout of 4 hours globally)
• Once a network is synced it makes sure that it process blocks for at least 20 minutes
• Once all of the above pass the ntwork is marked as successful
• Once all networks succeed we send a notification, update a history log and store consensus/ethrex logs for all networks. Then restarts the containers and starts again
• On failure, the containers aren't stopped, to be able to debug DB issues if needed, failures are also notified and stored in the history of runs

Status while running

Notification

History Log

Next Steps:

This is far from perfect but its working and adds a lot of value in it's current form, next we may want to:

• Be able to validate the state after the syncs
• tweak how we update or not the branch on each run
• Refactor the script
• Unify the way to run nodes in loop (server_runner.py)
• Add additional information to the history log and notifications (some ideas were discussed, like peer_count, etc)

Closes #5718