Project

General

Profile

Actions
Copy link

Bug #392

open

OAM: etcd heartbeat/election timeout tuning for spinning-disk cnodes

Bug #392: OAM: etcd heartbeat/election timeout tuning for spinning-disk cnodes

Added by Charles N 1 day ago.

Status:
New
Priority:
Normal
Assignee:
-
Target version:
-
Start date:
08/06/2026
Due date:
% Done:

0%

Estimated time:

Description

Even with writeback cache (fsync 2-3ms normal), occasional sub-second stalls (888ms observed) exceed k3s default etcd heartbeat of 500ms, destabilizing leader election during high-churn periods.

Fix: Tune etcd heartbeat-interval from 500ms to 1000ms, election-timeout from default to 5000ms via k3s --etcd-arg flags. This gives etcd 5x margin to absorb transient fsync stalls on spinning-disk storage.

Changes: k8s/install-cp.sh (add --etcd-arg flags to bootstrap + join commands)
Related: #391 (writeback cache), #367 (k8s rebuild)

No data to display

Actions
Copy link

Also available in: PDF Atom