The YB-Master hosts system metadata, records about tables in the system and locations of their tablets, users, roles, permissions, and so on. YB-Masters are also responsible for coordinating background operations such as schema changes, handling addition and removal of nodes from the cluster, automatic re-replication of data on permanent failures, and so on.
All handler latency metrics include additional attributes. Refer to Throughput and latency.
The following are key metrics for evaluating YB-Master performance. All metrics are counters and units are microseconds.
|Metric (Counter | microseconds)||Description|
||Time spent on fetching the replicas from the master servers. This metric includes the number of times the locations of the replicas are fetched from the master server.|
||Time to read the PostgreSQL system tables (during DDL). This metric includes the count or number of reads.|
||Time to write the PostgreSQL system tables (during DDL). This metric includes the count or number of writes.|
||Time to create a table (during DDL). This metric includes the count of create table operations.|
||Time to delete a table (during DDL). This metric includes the count of delete table operations.|
These metrics can be aggregated for nodes across the entire cluster using appropriate aggregations.