feat(sharding): add command to sync tables onto new nodes

[macobo]

clickhouse-operator only syncs some tables onto new nodes. This new
command ensures that when adding new shards, they are automatically
synced up on redeploying

Note that there might be timing concerns here as resharding on altinity
cloud does not redeploy automatically. In practice however what this
means is that new nodes just won't ingest any data until another deploy

Closes #8904, related to #8652

How did you test this code?

Custom-built an image and deployed it in a sharded setting.

[guidoiaquinti]

LGTM and few non-blocking comments:

• should we add a specific test for scenarios in sync_replicated_schema?
• should we handle as well the case of ghost tables? (e.g. tables that were dropped on some nodes but not on all of them)

[macobo]

should we handle as well the case of ghost tables? (e.g. tables that were dropped on some nodes but not on all of them)

Automatically dropping client data is a huge no-no in my book so not something I want to touch with a 20-foot pole. This is not an issue right now IMO.

should we add a specific test for scenarios in sync_replicated_schema?

Not really given our current CI setup or without mocking the DB. The mocking route wouldn't really give us anything either. Given how light-weight this script is, I don't feel too uncomfortable with this as-is, but do argue if you disagree!

[hazzadous]

I think it's worth adding a test here given it's critical for production deployments that this runs as expected.

The other comments are non-blocking

[hazzadous]

I assume these are ON CLUSTER queries, and that will not cause issues if tables are already defined on a node?

[macobo]

Exactly - you can also check schema.py for the queries.