Get the hot threads for nodes

  • GET /_nodes/{node_id}/hot_threads
  • GET /_nodes/hot_threads
Auth

Get a breakdown of the hot threads on each selected node in the cluster. The output is plain text with a breakdown of the top hot threads for each node.

Required authorization

Cluster privileges monitor manage

Path parameters

node_id union string | string[] optional

List of node IDs or names used to limit returned information.

Leave it out to use /_nodes/hot_threads.

Query parameters

ignore_idle_threads boolean · default: true

If true, known idle threads (e.g. waiting in a socket select, or to get a task from an empty queue) are filtered out.

interval string · default: 500ms

The interval to do the second sampling of threads.

snapshots number · default: 10

Number of samples of thread stacktrace.

threads number · default: 3

Specifies the number of hot threads to provide information for.

timeout string · default: 30s

Period to wait for a response. If no response is received before the timeout expires, the request fails and returns an error.

type enum · default: cpu

The type to sample.

Supported values include:

  • cpu: Threads that consume the most CPU time.
  • wait: Threads that have been in a waiting state the longest.
  • block: Threads that have been blocked the longest.
  • gpu: Threads that consume the most GPU time.
  • mem: Threads that allocate the most memory.
Values: cpu wait block gpu mem
sort enum · default: total

The sort order for 'cpu' type

Supported values include:

  • total: Sort by the time the thread was runnable, then by CPU time.
  • cpu: Sort by CPU time only.
Values: total cpu

Response

application/json

Response Type: object