Closed xenowits closed 2 years ago
Looks good to me, but still need to change what's happening in bootnode.yml init container.
My feel is we shouldn't have a shared PVC at all for these nodes, and instead we should inject everything in as configmaps + secrets. 1 config map for the definition. 4 secrets for the ENRs for each charon. and 4/8 secrets for each keystore.json and keystore.txt that need to be mounted into the VC
^ This could be something @eth2devops might explain to Abhishek and Dhruv on a call some time 😇 Good learning opportunity for config+secrets+volumes+lifecycle in k8s.
Looks good to me, but still need to change what's happening in bootnode.yml init container.
My feel is we shouldn't have a shared PVC at all for these nodes, and instead we should inject everything in as configmaps + secrets. 1 config map for the definition. 4 secrets for the ENRs for each charon. and 4/8 secrets for each keystore.json and keystore.txt that need to be mounted into the VC
^ This could be something @eth2devops might explain to Abhishek and Dhruv on a call some time 😇 Good learning opportunity for config+secrets+volumes+lifecycle in k8s.
That is great! I will explain the @dB2510 and @xenowits the current solution and how this cycle works, then we can agree on a new way without using a shared volume.
node3
) and a new teku vc (vc3-teku
) so as to bring the charon cluster node count to 4validator-api
andp2p-tcp