ashb commented on code in PR #73244:
URL: https://github.com/apache/airflow/pull/73244#discussion_r4027360437
##########
airflow-core/src/airflow/config_templates/config.yml:
##########
@@ -2097,13 +2097,19 @@ workers:
max_failed_heartbeats:
description: |
The maximum number of consecutive failed heartbeats before terminating
the task instance process.
+
+ Unlike other calls to the execution API, heartbeats are not retried.
This setting and
+ ``min_heartbeat_interval`` together determine how long a worker
tolerates an unreachable API server
+ before killing running tasks.
version_added: 3.0.0
type: integer
example: ~
- default: "3"
+ default: "20"
Review Comment:
Hmmmm, not sure if this is the right value here, I have a feeling setting
this so high might mean the task runs _well_ beyond the zombie threshold on the
scheduler side.
The original intent here was when we can't speak to the API server in a long
enough time to have the task stop itself, under the assumption that it can't
reach the API anymore, and thus the server will have likely said "this task
didn't heartbeat, I'll kill it as a zombie and start a new retry elsewhere".
API client retries were added later (somewhere early in 3.0.x series)
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
