stratumctl admin dns

Per-VPC DNS recursors: the replica setting, where each replica runs, and the upstream servers

stratumctl admin dns [flags]

Subcommands

• stratumctl admin dns recursor-settings: Show or set how many replicas of each VPC’s recursor the cluster runs
• stratumctl admin dns recursors: List every VPC’s recursor replicas: node, switch address, health, placement
• stratumctl admin dns upstream-servers: Show or set where every VPC’s resolver forwards the names the cloud does not serve

stratumctl admin dns recursor-settings

Show or set how many replicas of each VPC’s recursor the cluster runs

stratumctl admin dns recursor-settings [flags]

The cluster’s per-VPC recursor replica count: 1, or 2 for active/active (a single-node install runs 1). Setting it records the desired count and returns at once; the network-dns-recursor converger does the work, and “get” shows how far it has got.

A live change is safe in both directions. Going to 2, each VPC’s resolver starts sending queries to the second replica only once that replica answers. Going to 1, every resolver stops sending queries to the second replica before it is removed.

Subcommands

• stratumctl admin dns recursor-settings get: Show the setting and how far the cluster has converged on it
• stratumctl admin dns recursor-settings set: Set the per-VPC recursor replica count (1 or 2)

stratumctl admin dns recursor-settings get

Show the setting and how far the cluster has converged on it

stratumctl admin dns recursor-settings get [flags]

Examples

stratumctl admin dns recursor-settings get
stratumctl admin dns recursor-settings get --output json

stratumctl admin dns recursor-settings set

Set the per-VPC recursor replica count (1 or 2)

stratumctl admin dns recursor-settings set [flags]

Examples

stratumctl admin dns recursor-settings set --replicas 2

Flags

Flag	Type	Default	Description
--replicas	int	0	replicas of each VPC’s recursor: 1, or 2 for active/active

stratumctl admin dns recursors

List every VPC’s recursor replicas: node, switch address, health, placement

stratumctl admin dns recursors [flags]

Every VPC’s recursor as the operator sees it: each replica’s node and address on the VPC’s internal DNS switch, whether the VPC’s.3 balances to it and whether it answers, the placement inputs, and whether a client on each subnet got an answer from its.3. Tenants see only the last part.

Examples

stratumctl admin dns recursors
stratumctl admin dns recursors --project p-team-a --output json

Flags

Flag	Type	Default	Description
--project	string		only this project’s VPCs
--vpc	string		only this VPC

stratumctl admin dns upstream-servers

Show or set where every VPC’s resolver forwards the names the cloud does not serve

stratumctl admin dns upstream-servers [flags]

The cluster’s upstream DNS servers: where every VPC’s recursor forwards the names the cloud does not serve (public names, and the names your own resolvers serve). With none, the recursors resolve from the internet’s root servers, which needs DNS out to the internet.

Setting them records the list and returns at once, with no render and no control-plane restart: every recursor’s forward zone follows at once, and every node routes the new servers to its recursors within one sweep (30 s), in place. “get” shows how far the cluster has got.

When every server fails, a recursor fails the query. With the fallback on it resolves the name from the root servers instead, which hides the failure from tenants and sends names only your own resolvers serve out to public resolution. Either way “get” shows whether each server answers the recursors, and the dns-upstreams-unreachable alarm fires when one does not.

Until they are set, each control plane uses its own environment (node.yaml dns.upstream_servers, rendered as STRATUM_NETWORK_DEFAULT_DNS_SERVERS), which seeds the setting once.

Subcommands

• stratumctl admin dns upstream-servers fallback: Turn the root-server fallback on or off, keeping the servers
• stratumctl admin dns upstream-servers get: Show the upstream DNS servers and how far the cluster has converged on them
• stratumctl admin dns upstream-servers set: Set the upstream DNS servers (up to 4 IPv4 addresses), or –none to turn forwarding off

stratumctl admin dns upstream-servers fallback

Turn the root-server fallback on or off, keeping the servers

stratumctl admin dns upstream-servers fallback on|off [flags]

Off (the default): when every upstream server fails, the recursors fail the query. On: they resolve the name from the root servers instead, which needs DNS out to the internet, hides the failure from tenants and cannot resolve names only your own resolvers serve. The servers must have been set.

Examples

stratumctl admin dns upstream-servers fallback on

stratumctl admin dns upstream-servers get

Show the upstream DNS servers and how far the cluster has converged on them

stratumctl admin dns upstream-servers get [flags]

Examples

stratumctl admin dns upstream-servers get
stratumctl admin dns upstream-servers get --output json

stratumctl admin dns upstream-servers set

Set the upstream DNS servers (up to 4 IPv4 addresses), or –none to turn forwarding off

stratumctl admin dns upstream-servers set [SERVER...] [flags]

Set the cluster’s upstream DNS servers: up to 4 IPv4 addresses, in the order the recursors try them, none twice and none a node of this cluster (the same rule as node.yaml dns.upstream_servers). –none sets an empty list: no forwarding, the recursors resolve from the root servers. –fallback sets the root-server fallback in the same request; without it the fallback is kept.

Examples

stratumctl admin dns upstream-servers set 172.24.2.3 172.24.2.4
stratumctl admin dns upstream-servers set 172.24.2.3 --fallback=false
stratumctl admin dns upstream-servers set --none

Flags

Flag	Type	Default	Description
--fallback	bool	false	resolve from the root servers when every server fails (kept when not given)
--none	bool	false	set an empty list: no forwarding, resolve from the root servers