stratumctl nodes
Inspect + manage hypervisor nodes (operator-only,)
stratumctl nodes [flags]Aliases: node, hypervisor-nodes
Subcommands
Section titled “Subcommands”stratumctl nodes capacity: How many more of each instance type fit, and what runs out firststratumctl nodes cordon: Refuse new placements while keeping running instancesstratumctl nodes disk-wipe: Wipe a disk’s signatures so it can join a pool (operator-only)stratumctl nodes disks: List a node’s block devices + usage (operator-only)stratumctl nodes drain: Refuse new placements + evacuate the node’s instancesstratumctl nodes evacuation-plan: Show what draining a node would do to each of its instancesstratumctl nodes get: Describe a hypervisor nodestratumctl nodes list: List hypervisor nodesstratumctl nodes maintenance: Mark a node for operator-driven maintenance (kicks every instance to error)stratumctl nodes reboot: Reboot a node, gracefullystratumctl nodes set-reservation: Set what a single-node install keeps back from VMs (or hand it back to the computed default)stratumctl nodes shutdown: Shut a node down (power off), gracefullystratumctl nodes uncordon: Return a cordoned/drained node to ready
stratumctl nodes capacity
Section titled “stratumctl nodes capacity”How many more of each instance type fit, and what runs out first
stratumctl nodes capacity [flags]Examples
Section titled “Examples”stratumctl nodes capacitystratumctl nodes capacity --type m1.large| Flag | Type | Default | Description |
|---|---|---|---|
--type |
string | Show one instance type per node (e.g. m1.large) |
stratumctl nodes cordon
Section titled “stratumctl nodes cordon”Refuse new placements while keeping running instances
stratumctl nodes cordon <node-id> [flags]Examples
Section titled “Examples”stratumctl nodes cordon n-AbCd1234stratumctl nodes disk-wipe
Section titled “stratumctl nodes disk-wipe”Wipe a disk’s signatures so it can join a pool (operator-only)
stratumctl nodes disk-wipe <node-id> <disk> [flags]Examples
Section titled “Examples”stratumctl nodes disk-wipe n-AbCd1234 vdb --force-wipe| Flag | Type | Default | Description |
|---|---|---|---|
--force-wipe |
bool | false |
Required to wipe a disk carrying data; never overrides mounted/active-PV guards |
stratumctl nodes disks
Section titled “stratumctl nodes disks”List a node’s block devices + usage (operator-only)
stratumctl nodes disks <node-id> [flags]Examples
Section titled “Examples”stratumctl nodes disks n-AbCd1234stratumctl nodes drain
Section titled “stratumctl nodes drain”Refuse new placements + evacuate the node’s instances
stratumctl nodes drain <node-id> [flags]Examples
Section titled “Examples”stratumctl nodes drain n-AbCd1234stratumctl nodes drain n-AbCd1234 --allow-disruption| Flag | Type | Default | Description |
|---|---|---|---|
--allow-disruption |
bool | false |
permit stopping running instances that cannot move any other way (their root volume is node-local) |
stratumctl nodes evacuation-plan
Section titled “stratumctl nodes evacuation-plan”Show what draining a node would do to each of its instances
stratumctl nodes evacuation-plan <node-id> [flags]Examples
Section titled “Examples”stratumctl nodes evacuation-plan n-AbCd1234stratumctl nodes get
Section titled “stratumctl nodes get”Describe a hypervisor node
stratumctl nodes get <node-id> [flags]Examples
Section titled “Examples”stratumctl nodes get n-AbCd1234stratumctl nodes get n-AbCd1234 --output jsonstratumctl nodes list
Section titled “stratumctl nodes list”List hypervisor nodes
stratumctl nodes list [flags]Examples
Section titled “Examples”stratumctl nodes liststratumctl nodes list --state ready --az lab-1a| Flag | Type | Default | Description |
|---|---|---|---|
--az |
string | Filter by availability zone | |
--region |
string | Filter by region | |
--role |
string | Filter by role (hybrid | hypervisor) | |
--state |
string | Filter by state (ready | cordoned | draining | maintenance | unknown) |
stratumctl nodes maintenance
Section titled “stratumctl nodes maintenance”Mark a node for operator-driven maintenance (kicks every instance to error)
stratumctl nodes maintenance <node-id> [flags]Examples
Section titled “Examples”stratumctl nodes maintenance n-AbCd1234stratumctl nodes reboot
Section titled “stratumctl nodes reboot”Reboot a node, gracefully
stratumctl nodes reboot <node-id> [flags]Reboot a node. The node’s own host-helper reboots it a few seconds after the answer; on the way down its VMs are shut down (ACPI, then power-off after the drain’s timeout) and its Ceph OSDs get noout, and on the way back the VMs it stopped start again. The node is held in maintenance until it has booted again.
On a single-node install the console and the API are down until it is back.
Refused for the same reasons as ‘stratumctl nodes shutdown’; –force overrides the quorum refusal only.
Examples
Section titled “Examples”stratumctl nodes reboot n-msst0002 --reason "kernel update"stratumctl nodes reboot n-msst0002 --yes -o json| Flag | Type | Default | Description |
|---|---|---|---|
--force |
bool | false |
Go ahead when the cluster would be left without a control-plane majority |
--reason |
string | Why (recorded and audited) | |
-y, --yes |
bool | false |
Do not ask for confirmation |
stratumctl nodes set-reservation
Section titled “stratumctl nodes set-reservation”Set what a single-node install keeps back from VMs (or hand it back to the computed default)
stratumctl nodes set-reservation <node-id> [flags]Set the node’s reservation: the vCPUs and memory the substrate keeps for itself, taken off the total before the overcommit ratio, and enforced as system.slice’s MemoryMin/MemoryLow.
The default is computed from the node’s size: max(2, 6% of the threads) vCPUs and max(6 GiB, 10% of RAM). A value set here is kept from then on. It may go below the default for a development box, down to 1 vCPU and 4 GiB: that floor is the minimum for development and not recommended for production.
stratumctl nodes set-reservation n-0123456789 --vcpus 2 --memory-gib 8stratumctl nodes set-reservation n-0123456789 --computedSingle-node installs only.
Examples
Section titled “Examples”stratumctl nodes set-reservation n-AbCd1234 --vcpus 4 --memory-gib 12stratumctl nodes set-reservation n-AbCd1234 --computed| Flag | Type | Default | Description |
|---|---|---|---|
--computed |
bool | false |
hand the node back to the computed default |
--memory-gib |
int64 | 0 |
GiB of memory to keep back from VMs (at least 4) |
--vcpus |
int | 0 |
vCPUs to keep back from VMs (at least 1) |
stratumctl nodes shutdown
Section titled “stratumctl nodes shutdown”Shut a node down (power off), gracefully
stratumctl nodes shutdown <node-id> [flags]Shut a node down. The node’s own host-helper powers it off a few seconds after the answer; on the way down its VMs are shut down (ACPI, then power-off after the drain’s timeout) and its Ceph OSDs get noout. The node is held in maintenance until it has booted again, so nothing is placed on it and nothing evacuates it. It stays off until someone powers it on.
On a single-node install the console and the API go down with it.
Refused while an upgrade or a cluster shutdown runs, while an instance migrates to or from the node, and when the cluster would be left without a majority of its control-plane nodes up (–force overrides that one only).
Examples
Section titled “Examples”stratumctl nodes shutdown n-msst0002 --reason "replace a fan"stratumctl nodes shutdown n-msst0002 --yes -o json| Flag | Type | Default | Description |
|---|---|---|---|
--force |
bool | false |
Go ahead when the cluster would be left without a control-plane majority |
--reason |
string | Why (recorded and audited) | |
-y, --yes |
bool | false |
Do not ask for confirmation |
stratumctl nodes uncordon
Section titled “stratumctl nodes uncordon”Return a cordoned/drained node to ready
stratumctl nodes uncordon <node-id> [flags]Examples
Section titled “Examples”stratumctl nodes uncordon n-AbCd1234