Add dictionary metrics to system

Why it matters

These metrics provide monitoring and diagnostic information for dictionary updates by reporting the maximum delay (in seconds) of dictionary update and the total number of failed updates since the last successful load. This helps users to better track dictionary health and performance.

How to use it

Users can query the new metrics by running SQL against the system.asynchronous_metrics table to monitor dictionary update delays and failures. For example:

SELECT metric, value FROM system.asynchronous_metrics WHERE metric IN ('DictionaryMaxUpdateDelay', 'DictionaryTotalFailedUpdates')