Skip to content

fix(autoscaler): leave a Hetzner server free for the Talos snapshot build #268

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
devantler merged 1 commit into from
Oct 25, 2026
Merged

fix(autoscaler): leave a Hetzner server free for the Talos snapshot build #268

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
17 changes: 11 additions & 6 deletions docs/node-autoscaling.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -14,12 +14,12 @@ KSail natively installs and manages the Cluster Autoscaler when
```
KSail (static baseline)
├── 3 control planes (cx33, 4 vCPU / 8 GB, never autoscaled)
└── 3 static workers (cx33, 4 vCPU / 8 GB, guaranteed minimum, Longhorn storage nodes)
└── 4 static workers (cx33, 4 vCPU / 8 GB, guaranteed minimum, Longhorn storage nodes)

Cluster Autoscaler (dynamic workers, managed by KSail)
├── Pool: autoscale-small → 0-1 × CX23 (2 vCPU, 4 GB)
├── Pool: autoscale-medium → 0-2 × CX33 (4 vCPU, 8 GB)
├── maxNodesTotal: 10 (3 CPs + 3 workers + headroom for autoscaler nodes)
├── maxNodesTotal: 9 (3 CPs + 4 workers + up to 2 autoscaler nodes)
└── Expander: LeastWaste
```

Expand Down Expand Up @@ -65,7 +65,7 @@ spec:
node:
enabled: true
expander: LeastWaste
maxNodesTotal: 10
maxNodesTotal: 9
scaleDownUnneededTime: "10m"
pools:
- name: autoscale-small
Expand Down Expand Up @@ -96,9 +96,14 @@ spec:

- **Hard max per pool** — `pools[].max` caps each pool independently.
- **Hard max total** — `maxNodesTotal` caps the **total cluster node count**
(CPs + static workers + autoscaler workers). Set to `10` (3 CPs + 3
workers = 6 base, plus up to 3 autoscaler nodes from the current pool caps
of 1 small + 2 medium).
(CPs + static workers + autoscaler workers). Set to `9`: 3 CPs + 4
workers = 7 base, plus up to 2 autoscaler nodes, although the pool caps
(1 small + 2 medium) would allow 3.
- **Snapshot-build slot** — KSail starts a temporary server to build the
Talos snapshot that nodes boot from, and rejects the config unless that
server fits under the Hetzner server limit (`spec.provider.hetzner.serverLimit`,
default `10`). Keep `maxNodesTotal`, or the base plus every pool cap when it
is lower, at least one below the limit.
- **Expander** — `LeastWaste` (current) picks the node group left with the
least idle capacity after scheduling; `Price` instead picks the cheapest
eligible group.
Expand Down
6 changes: 5 additions & 1 deletion ksail.prod.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -23,7 +23,11 @@ spec:
node:
enabled: true
expander: LeastWaste
maxNodesTotal: 10
# Whole-cluster ceiling: 3 control planes + 4 static workers + up to
# 2 autoscaled nodes. Keep it one below the Hetzner server limit
# (default 10): KSail rejects the config unless one server stays free
# for the temporary server that builds the Talos snapshot.
maxNodesTotal: 9
scaleDownUnneededTime: "10m"
pools:
- name: autoscale-small
Expand Down
Loading