Skip to content

Instantly share code, notes, and snippets.

@tom-clickhouse
Created September 26, 2023 09:34
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save tom-clickhouse/2d3f292ee0aac762251626c7c3156966 to your computer and use it in GitHub Desktop.
Save tom-clickhouse/2d3f292ee0aac762251626c7c3156966 to your computer and use it in GitHub Desktop.
------------------------------------------------------------------------------------
Configuration
------------------------------------------------------------------------------------
keeper_cpu_cores: number of CPU cores for running each of the three Keeper instances
keeper_memory: amount of RAM for running each of the three Keeper instances
keeper_clients: number of simulated clients (e.g. ClickHouse servers) sending requests in parallel to Keeper
keeper_requests: total number of requests sent to Keeper
------------------------------------------------------------------------------------
Metrics
------------------------------------------------------------------------------------
runtime_chkeeper and runtime_zookeeper:
based on benchmark run timestamps, the runtime for ClickHouse Keeper or Zookeeper processing the requests
memory_chkeeper and memory_zookeeper:
based on cAdvisor’s `container_memory_working_set_bytes`, the peak container memory usage of a single ClickHouse Keeper or Zookeeper instance during request processing
cpu_chkeeper and cpu_zookeeper:
based on cAdvisor’s `container_cpu_usage_seconds_total`, the peak CPU usage of a single ClickHouse Keeper or Zookeeper instance during request processing
nodes_chkeeper and nodes_zookeeper:
based on Keeper’s `zk_znode_count`, the peak number of znodes stored in ClickHouse Keeper or Zookeeper from the requests; see here for reasons why these numbers can differ between Keepers in the same run.
data_chkeeper and data_zookeeper:
based on Keeper’s `zk_approximate_data_size`, the approximate peak amount of memory consumption for all znodes stored in ClickHouse Keeper or Zookeeper from the requests; see here for reasons why these numbers can differ between Keepers in the same run.
zk_jvm_memory_committed and zk_jvm_memory_used:
based on ZooKeeper’s built-in JVM metrics, the committed and actually used JVM heap + non-heap memory when running Zookeeper
zk_jvm_memory_committed =
zk_jvm_memory_bytes_committed_heap + zk_jvm_memory_bytes_committed_nonheap
zk_jvm_memory_used =
zk_jvm_memory_bytes_used_heap + zk_jvm_memory_bytes_used_nonheap
zk_jvm_gc_collection_count:
based on ZooKeeper’s built-in JVM metrics, the sum of garbage collection runs of G1 old generation, G1 young generation, and mark-sweep compact garbage collectors.
------------------------------------------------------------------------------------
Metric factors
------------------------------------------------------------------------------------
factor_runtime: Zookeeper runtime divided by ClickHouse Keeper runtime
factor_memory: Zookeeper container memory usage divided by ClickHouse Keeper container memory usage
factor_cpu: Zookeeper CPU usage divided by ClickHouse Keeper CPU usage
------------------------------------------------------------------------------------
Meta infos
------------------------------------------------------------------------------------
chkeeper_version and zookeeper_version:
used software version of ClickHouse Keeper or Zookeeper
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment