Skip to main content

Overview

The real-time overview shows your cluster workload as it happens. CPU, memory, disk I/O, and network are attributed to individual queries, merges, and background operations in real time.

From here you can deep dive into any metric to inspect the queries, merges, or parts driving it.

What You See​

  • CPU - Total CPU broken down by queries, merges, mutations, replication, and other
  • Memory - RSS decomposed into query buffers, mark cache, uncompressed cache, primary keys, dictionaries, and jemalloc overhead
  • Disk I/O - Read/write throughput attributed to queries, merges, and background operations
  • Network - Inter-node replication and distributed query traffic

Click any metric attribution to jump directly to the relevant detail view:

  • Click a query's CPU slice to open it in the Query Monitor
  • Click a merge's I/O bar to see it in the Merge Tracker
  • Click a table's disk usage to inspect it in the Database Explorer