Reduce ClickHouse log level

[bnaecker]

We're currently shipping a ClickHouse configuration file that uses the default log level of trace. That's helpful during local development, but we have never needed such detailed logs on a production system. We should reduce that to debug or probably even info.

[karencfv]

We'll be rolling out the replicated cluster alongside the single node once #7522 is merged. Wouldn't the current log level be useful for our long running testing that will happen afterwards?

[bnaecker]

Maybe! I have no plans to work on this until we have to, in any case.