Skip to content

OCPBUGS-114888: clarify that bare metal supports 3, 4, or 5 control plane nodes #119014

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

Closed
agawai wants to merge 1 commit into from
+14 −3
Closed

OCPBUGS-114888: clarify that bare metal supports 3, 4, or 5 control plane nodes #119014

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
2 changes: 2 additions & 0 deletions ...d_restore/control_plane_backup_and_restore/replacing-unhealthy-etcd-member.adoc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -20,6 +20,8 @@ If you have lost the majority of your control plane hosts, follow the steps in "
If the control plane certificates are not valid on the member being replaced, then you must follow the steps in "Recovering from expired control plane certificates" instead of this procedure.

If a control plane node is lost and a new one is created, the etcd cluster Operator handles generating the new TLS certificates and adding the node as an etcd member.

This procedure applies to clusters with 3, 4, or 5 control plane nodes. Most production clusters have 3 control plane nodes. Bare-metal clusters can have 3, 4, or 5; a 4-node control plane is supported but is not the standard production design.
====

// Identifying an unhealthy etcd member
Expand Down
2 changes: 2 additions & 0 deletions etcd/etcd-backup-restore/replace-unhealthy-etcd-member.adoc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -20,6 +20,8 @@ If you have lost the majority of your control plane hosts, follow the steps in "
If the control plane certificates are not valid on the member being replaced, then you must follow the steps in "Recovering from expired control plane certificates" instead of this procedure.

If a control plane node is lost and a new one is created, the etcd cluster Operator handles generating the new TLS certificates and adding the node as an etcd member.

This procedure applies to clusters with 3, 4, or 5 control plane nodes. Most production clusters have 3 control plane nodes. Bare-metal clusters can have 3, 4, or 5; a 4-node control plane is supported but is not the standard production design.
====

// Identifying an unhealthy etcd member
Expand Down
2 changes: 1 addition & 1 deletion ...ent/control_plane_machine_management/cpmso-manually-scaling-control-planes.adoc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,7 +7,7 @@ include::_attributes/common-attributes.adoc[]
toc::[]

[role="_abstract"]
Manually scale to 4 or 5 control plane nodes on bare-metal infrastructure to recover from a degraded state, perform deep-level debugging, or ensure control plane stability in complex scenarios.
On bare-metal infrastructure, you can manually scale a standard 3-node control plane to 4 or 5 control plane nodes to recover from a degraded state, perform deep-level debugging, or ensure control plane stability in complex scenarios. A cluster on bare metal can have 3, 4, or 5 control plane nodes.

[IMPORTANT]
====
Expand Down
2 changes: 1 addition & 1 deletion modules/architecture-machine-roles.adoc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -98,7 +98,7 @@ Extra controls apply to control plane machines to prevent you from deleting all
[NOTE]
====
ifndef::openshift-dedicated,openshift-rosa[]
Exactly three control plane nodes must be used for all production deployments. However, on bare metal platforms, clusters can be scaled up to five control plane nodes.
Production deployments typically use 3 control plane nodes. On bare metal platforms only, a cluster can have 3, 4, or 5 control plane nodes. 4-node and 5-node control planes are a non-standard high-availability option, not the default production design.
endif::openshift-dedicated,openshift-rosa[]
ifdef::openshift-dedicated,openshift-rosa[]
Single availability zone clusters and multiple availability zone clusters require a minimum of three control plane nodes.
Expand Down
9 changes: 8 additions & 1 deletion modules/etcd-node-scaling.adoc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,7 +7,9 @@
= Node scaling for etcd

[role="_abstract"]
In general, clusters must have 3 control plane nodes. However, if your cluster is installed on a bare metal platform, it can have up to 5 control plane nodes. If your bare-metal cluster has fewer than 5 control plane nodes, you can scale up the cluster as a postinstallation task.
In general, clusters must have 3 control plane nodes. A 3-node control plane is the standard production design.

If your cluster is installed on a bare metal platform, it can have 3, 4, or 5 control plane nodes. 4-node and 5-node control planes are a supported non-standard high-availability option; they are not the default production design. If your bare-metal cluster has fewer than 5 control plane nodes, you can scale up the cluster as a postinstallation task.

For example, to scale from 3 to 4 control plane nodes after installation, you can add a host and install it as a control plane node. Then, the etcd Operator scales accordingly to account for the additional control plane node.

Expand Down Expand Up @@ -44,4 +46,9 @@ The following table shows failure tolerance for clusters of different sizes:
|2
|===

[NOTE]
====
A 4-node control plane has the same failure tolerance as a 3-node control plane. Only a 5-node control plane can tolerate 2 simultaneous control plane node failures.
====

For more information about recovering from quorum loss, see "Restoring to an earlier cluster state".