Friction

Pain point · Self-hosting & upgrades · Blocks work

Orchestrator pod ignores configured Kubernetes tolerations

2 source threads · first seen 2023-07 · last seen 2024-09

Summary

With tolerations set for job pods, sync jobs get them but the orchestrator pod starts without them, so it cannot be scheduled on tainted nodes.

Affects
Kubernetes deployers using tainted nodes

Evidence

Excerpts are copied word for word from the source; follow the link to read it in full.

“In the new version, the tolerations in Helm values are not added to the job pods”

“sync jobs run with correct tolerations, but an orchestrator pod starts without tolerations on its pod”

Report this item