Same Docker alias for multiple staking containers can make RPC calls buggy

[Marketen]

Giving the same alias to multiple staker containers makes calls to their endpoints like http://beacon-chain.holesky.dncore.dappnode:3500 point to the wrong container under specific conditions.

When an user has two staking clients of the same type up & running (like nimbus holesky & lighthouse holesky), they both have the beacon-chain.holesky.dncore.dappnode alias set in docker's dncore_network. Because of this, when calling http://beacon-chain.holesky.dncore.dappnode:3500, the call will end up going to the container that has been up for more time. This is a problem because that container is not necessarily the same container that is selected in the stakers tab

Solution:

• Remove shared aliases like beacon-chain.holesky.dncore.dappnode from staking container when it stops

How to reproduce:

1. Select any EC/CC pair in the stakers tab.
2. Change CC to another one in the stakers tab.
3. Stop both installed CC. Start first the CC installed in point 1, then the other one.
4. Make a call to curl -v http://beacon-chain.holesky.dncore.dappnode:3500 from a dappnode package like obol.
5. Check curl's output: a line like * Trying 172.33.0.30:3500... will appear. Notice how this is the IP of the CC not currently selected in the stakers tab.

[Marketen]

Seems to be solved by this #2060