Cluster Commands

sbctl cluster --help

Cluster Commands

Creates a new cluster.

Created a new control plane cluster with the current node as the primary control plane node.

sbctl cluster create
    --cap-warn=<CAP_WARN>
    --cap-crit=<CAP_CRIT>
    --prov-cap-warn=<PROV_CAP_WARN>
    --prov-cap-crit=<PROV_CAP_CRIT>
    --ifname=<IFNAME>
    --mgmt-ip=<MGMT_IP>
    --tls-secret-name=<TLS_SECRET_NAME>
    --log-del-interval=<LOG_DEL_INTERVAL>
    --metrics-retention-period=<METRICS_RETENTION_PERIOD>
    --contact-point=<CONTACT_POINT>
    --grafana-endpoint=<GRAFANA_ENDPOINT>
    --data-chunks-per-stripe=<DATA_CHUNKS_PER_STRIPE>
    --parity-chunks-per-stripe=<PARITY_CHUNKS_PER_STRIPE>
    --ha-type=<HA_TYPE>
    --is-single-node
    --mode=<MODE>
    --ingress-host-source=<INGRESS_HOST_SOURCE>
    --dns-name=<DNS_NAME>
    --enable-node-affinity
    --fabric=<FABRIC>
    --strict-node-anti-affinity
    --enable-failure-domain
    --device-mode=<DEVICE_MODE>
    --enable-inline-checksum
    --4k_atomic
    --name=<NAME>
    --qpair-count=<QPAIR_COUNT>
    --client-qpair-count=<CLIENT_QPAIR_COUNT>
    --client-data-nic=<CLIENT_DATA_NIC>
    --use-backup=<USE_BACKUP>
    --nvmf-base-port=<NVMF_BASE_PORT>
    --rpc-base-port=<RPC_BASE_PORT>
    --snode-api-port=<SNODE_API_PORT>
    --max-subsys=<MAX_SUBSYS>
    --hugepages-mem=<HUGEPAGES_MEM>
    --vcpu-count=<VCPU_COUNT>
    --hashicorp-vault-url=<HASHICORP_VAULT_URL>
    --alerting-config-path=<ALERTING_CONFIG_PATH>
    --cluster-vip=<CLUSTER_VIP>

Parameter	Description	Data Type	Required	Default
--cap-warn	The capacity warning level in percent. Default: `89`.	integer	False	89
--cap-crit	The capacity critical level in percent. Default: `99`.	integer	False	99
--prov-cap-warn	The capacity warning level in percent. Default: `250`.	integer	False	250
--prov-cap-crit	The capacity critical level in percent. Default: `500`.	integer	False	500
--ifname	Management interface name, e.g. eth0.	string	False	-
--mgmt-ip	Management IP address to use for the node (e.g., 192.168.1.10).	string	False	-
--tls-secret-name	Name of the Kubernetes TLS Secret to be used by the Ingress for HTTPS termination (e.g., my-tls-secret).	string	False	-
--log-del-interval	The logging retention policy. Default: `3d`.	string	False	3d
--metrics-retention-period	Retention period for I/O statistics (Prometheus). Default: `7d`.	string	False	7d
--contact-point	The email or slack webhook url to be used for alerting.	string	False
--grafana-endpoint	The endpoint url for Grafana.	string	False
--data-chunks-per-stripe	The erasure coding schema parameter k (distributed raid). Default: `1`.	integer	False	1
--parity-chunks-per-stripe	The erasure coding schema parameter n (distributed raid). Default: `1`.	integer	False	1
--ha-type	Logical volume HA type (single, ha), default is cluster ha type. Default: `ha`. Available Options: - single - ha	string	False	ha
--is-single-node	For single-node clusters only. Default: `false`.	marker	False	False
--mode	The environment to deploy management services. Default: `docker`. Available Options: - docker - kubernetes	string	False	docker
--ingress-host-source	Ingress host source: 'hostip' for node IP, 'loadbalancer' for external LB, or 'dns' for custom domain. Default: `hostip`. Available Options: - hostip - loadbalancer - dns	string	False	hostip
--dns-name	Fully qualified DNS name to use as the Ingress host (required if --ingress-host-source=dns).	string	False
--enable-node-affinity	Enable node affinity for storage nodes.	marker	False	-
--fabric	The NVMe fabric to use (specify: `tcp`, `rdma`, `tcp,rdma`). Default: `tcp`. Available Options: - tcp - rdma - tcp,rdma	string	False	tcp
--strict-node-anti-affinity	Enable strict node anti affinity for storage nodes. Never more than one chunk is placed on a node. This requires a minimum of data-chunks-in-stripe + parity-chunks-in-stripe + 1 nodes in the cluster.	marker	False	-
--enable-failure-domain	Enable failure-domain anti-affinity. Each storage node must then be added with a --failure-domain tag (rack/cabinet/DC); data, journal and secondary/tertiary copies are spread across distinct failure domains (best-effort). Deploy-time only: a cluster cannot be upgraded into this feature, it must be redeployed.	marker	False	-
--device-mode	Storage-device mode for the whole cluster. 'nvme' (default): NVMe PCIe devices auto-detected and attached via the SPDK nvme driver. 'lblk': arbitrary Linux block devices wrapped in SPDK AIO bdevs; devices are selected at 'sn configure' by name or serial number. Deploy-time only; inter-node fabric (nvme-tcp/rdma) is unaffected. Available Options: - nvme - lblk	string	False	nvme
--enable-inline-checksum	Enable inline CRC checksum validation on every IO for silent-data-error protection. Cannot be enabled or disabled after cluster creation. Per-device alceml mode (md-on-device vs fallback) is auto-detected at add-node.	marker	False	-
--4k_atomic	Declare that devices guarantee 4K write atomicity even with a <4K logical block size (e.g. AWS NVMe is 512B but atomic at 4K). Allows fallback-mode inline checksum on such devices by skipping the data-plane 4K block-size requirement. Only meaningful with --enable-inline-checksum. Cannot be changed after cluster creation.	marker	False	-
--name, -n	Assigns a name to the newly created cluster.	string	False	-
--qpair-count	Increase for clusters with few but very large logical volumes or decrease for clusters with a large number of very small logical volumes.	range(0..128)	False	32
--client-qpair-count	Increase for clusters with few but very large logical volumes or decrease for clusters with a large number of very small logical volumes.	range(0..128)	False	3
--client-data-nic	Network interface name from client to use for logical volume connection.	string	False	-
--use-backup	The path to JSON file with S3/MinIO backup configuration.	string	False	-
--nvmf-base-port	Base port for all NVMe-oF listeners (lvol, hublvol, device). Default: `4420`.	integer	False	4420
--rpc-base-port	The base port for SPDK JSON-RPC. Default: `8080`.	integer	False	8080
--snode-api-port	The SNodeAPI/firewall port (one per host IP). Default: `50001`.	integer	False	50001
--max-subsys	Cluster-wide ceiling on nvmf subsystems per storage node. 75 is a hard product limit and larger values are rejected. Set here rather than per node so every node in the cluster is sized alike; a node adopts a change on its next restart.	integer	False	0
--hugepages-mem	Huge-page memory floor per storage node, e.g. 4G. Cluster-wide; a node adopts it on its next restart.	string	False
--vcpu-count	Absolute core budget for SPDK on every storage node, replacing the old cores-percentage: a percentage meant different things on different hardware, a count does not. One core beyond this is required for the system, so a node with fewer than count+1 vCPUs is refused rather than silently under-provisioned.	integer	False	0
--hashicorp-vault-url	Hashicorp vault URL for storing encryption keys for this cluster	string	False	-
--alerting-config-path	Configures the Grafana contact points from a file, so that a cluster can notify Slack, Microsoft Teams, PagerDuty, Opsgenie and a generic webhook at once. The file has the same shape as the controlplane.observability.grafana.notifications block of the simplyblock-operator Helm chart values. When given, --contact-point is ignored.	string	False	-
--cluster-vip	Cluster Virtual IP, this is the load balancer IP or domain, would be used for the cluster logging ingress.	string	False	-

Adds a new cluster.

sbctl cluster add
    --cap-warn=<CAP_WARN>
    --cap-crit=<CAP_CRIT>
    --prov-cap-warn=<PROV_CAP_WARN>
    --prov-cap-crit=<PROV_CAP_CRIT>
    --data-chunks-per-stripe=<DATA_CHUNKS_PER_STRIPE>
    --parity-chunks-per-stripe=<PARITY_CHUNKS_PER_STRIPE>
    --ha-type=<HA_TYPE>
    --enable-node-affinity
    --fabric=<FABRIC>
    --is-single-node
    --qpair-count=<QPAIR_COUNT>
    --client-qpair-count=<CLIENT_QPAIR_COUNT>
    --strict-node-anti-affinity
    --enable-failure-domain
    --device-mode=<DEVICE_MODE>
    --enable-inline-checksum
    --4k_atomic
    --name=<NAME>
    --client-data-nic=<CLIENT_DATA_NIC>
    --use-backup=<USE_BACKUP>
    --nvmf-base-port=<NVMF_BASE_PORT>
    --rpc-base-port=<RPC_BASE_PORT>
    --snode-api-port=<SNODE_API_PORT>
    --max-subsys=<MAX_SUBSYS>
    --hugepages-mem=<HUGEPAGES_MEM>
    --vcpu-count=<VCPU_COUNT>
    --hashicorp-vault-url=<HASHICORP_VAULT_URL>

Parameter	Description	Data Type	Required	Default
--cap-warn	The capacity warning level in percent. Default: `89`.	integer	False	89
--cap-crit	The capacity critical level in percent. Default: `99`.	integer	False	99
--prov-cap-warn	The capacity warning level in percent. Default: `250`.	integer	False	250
--prov-cap-crit	The capacity critical level in percent. Default: `500`.	integer	False	500
--data-chunks-per-stripe	The erasure coding schema parameter k (distributed raid). Default: `1`.	integer	False	1
--parity-chunks-per-stripe	The erasure coding schema parameter n (distributed raid). Default: `1`.	integer	False	1
--ha-type	Logical volume HA type (single, ha), default is cluster single type. Default: `ha`. Available Options: - single - ha	string	False	ha
--enable-node-affinity	Enables node affinity for storage nodes.	marker	False	-
--fabric	Fabric: tcp, rdma or both (specify: tcp, rdma). Default: `tcp`. Available Options: - tcp - rdma - tcp,rdma	string	False	tcp
--is-single-node	For single-node clusters only. Default: `false`.	marker	False	False
--qpair-count	Increase for clusters with few but very large logical volumes or decrease for clusters with a large number of very small logical volumes.	range(0..128)	False	32
--client-qpair-count	Increase for clusters with few but very large logical volumes or decrease for clusters with a large number of very small logical volumes.	range(0..128)	False	3
--strict-node-anti-affinity	Enable strict node anti affinity for storage nodes. Never more than one chunk is placed on a node. This requires a minimum of data-chunks-in-stripe + parity-chunks-in-stripe + 1 nodes in the cluster."	marker	False	-
--enable-failure-domain	Enable failure-domain anti-affinity. Each storage node must then be added with a --failure-domain tag (rack/cabinet/DC); data, journal and secondary/tertiary copies are spread across distinct failure domains (best-effort). Deploy-time only: a cluster cannot be upgraded into this feature, it must be redeployed.	marker	False	-
--device-mode	Storage-device mode for the whole cluster. 'nvme' (default): NVMe PCIe devices auto-detected and attached via the SPDK nvme driver. 'lblk': arbitrary Linux block devices wrapped in SPDK AIO bdevs; devices are selected at 'sn configure' by name or serial number. Deploy-time only; inter-node fabric (nvme-tcp/rdma) is unaffected. Available Options: - nvme - lblk	string	False	nvme
--enable-inline-checksum	Enable inline CRC checksum validation on every IO for silent-data-error protection. Cannot be enabled or disabled after cluster creation.	marker	False	-
--4k_atomic	Declare that devices guarantee 4K write atomicity even with a <4K logical block size (e.g. AWS NVMe is 512B but atomic at 4K). Allows fallback-mode inline checksum on such devices by skipping the data-plane 4K block-size requirement. Only meaningful with --enable-inline-checksum. Cannot be changed after cluster creation.	marker	False	-
--name, -n	Assigns a name to the newly created cluster.	string	False	-
--client-data-nic	Network interface name from client to use for logical volume connection.	string	False	-
--use-backup	The path to JSON file with S3/MinIO backup configuration.	string	False	-
--nvmf-base-port	Base port for all NVMe-oF listeners (lvol, hublvol, device). Default: `4420`.	integer	False	4420
--rpc-base-port	The base port for SPDK JSON-RPC. Default: `8080`.	integer	False	8080
--snode-api-port	The SNodeAPI/firewall port (one per host IP). Default: `50001`.	integer	False	50001
--max-subsys	Cluster-wide ceiling on nvmf subsystems per storage node. 75 is a hard product limit and larger values are rejected. Set here rather than per node so every node in the cluster is sized alike; a node adopts a change on its next restart.	integer	False	0
--hugepages-mem	Huge-page memory floor per storage node, e.g. 4G. Cluster-wide; a node adopts it on its next restart.	string	False
--vcpu-count	Absolute core budget for SPDK on every storage node, replacing the old cores-percentage: a percentage meant different things on different hardware, a count does not. One core beyond this is required for the system, so a node with fewer than count+1 vCPUs is refused rather than silently under-provisioned.	integer	False	0
--hashicorp-vault-url	Hashicorp vault URL for storing encryption keys for this cluster	string	False	-

Stops the cluster accepting object lifecycle operations: creation, deletion and modification of volumes, snapshots, clones and pools. Read paths and the cluster's own maintenance are unaffected.

sbctl cluster op-stop
    <CLUSTER_ID>

Argument	Description	Data Type	Required
CLUSTER_ID	The cluster id.	string	True

Resumes object lifecycle operations on the cluster.

sbctl cluster op-start
    <CLUSTER_ID>

Argument	Description	Data Type	Required
CLUSTER_ID	The cluster id.	string	True

Activates a cluster.

Once a cluster has sufficient nodes added, it needs to be activated. Can also be used to re-activate a suspended cluster.

sbctl cluster activate
    <CLUSTER_ID>
    --force
    --force-lvstore-create

Argument	Description	Data Type	Required
CLUSTER_ID	The cluster id.	string	True

Parameter	Description	Data Type	Required	Default
--force	Force recreate distr and lv stores.	marker	False	-
--force-lvstore-create	Force recreate lv stores.	marker	False	-

Shows the cluster list.

sbctl cluster list
    --json

Parameter	Description	Data Type	Required	Default
--json	Print json output.	marker	False	-

Shows a cluster's status.

sbctl cluster status
    <CLUSTER_ID>

Argument	Description	Data Type	Required
CLUSTER_ID	The cluster id.	string	True

Create lvstore on newly added nodes to the cluster.

sbctl cluster complete-expand
    <CLUSTER_ID>

Argument	Description	Data Type	Required
CLUSTER_ID	The cluster id.	string	True

Shows a cluster's statistics.

sbctl cluster show
    <CLUSTER_ID>

Argument	Description	Data Type	Required
CLUSTER_ID	The cluster id.	string	True

Gets a cluster's information.

sbctl cluster get
    <CLUSTER_ID>

Argument	Description	Data Type	Required
CLUSTER_ID	The cluster id.	string	True

Gets a cluster's capacity.

sbctl cluster get-capacity
    <CLUSTER_ID>
    --json
    --history=<HISTORY>

Argument	Description	Data Type	Required
CLUSTER_ID	The cluster id.	string	True

Parameter	Description	Data Type	Required	Default
--json	Print json output.	marker	False	-
--history	(XXdYYh), list history records (one for every 15 minutes) for XX days and YY hours (up to 10 days in total).	string	False	-

Gets a cluster's I/O statistics.

sbctl cluster get-io-stats
    <CLUSTER_ID>
    --records=<RECORDS>
    --history=<HISTORY>

Argument	Description	Data Type	Required
CLUSTER_ID	The cluster id.	string	True

Parameter	Description	Data Type	Required	Default
--records	The number of records. Default: `20`.	integer	False	20
--history	(XXdYYh), list history records (one for every 15 minutes) for XX days and YY hours (up to 10 days in total).	string	False	-

Returns a cluster's status logs.

sbctl cluster get-logs
    <CLUSTER_ID>
    --json
    --limit=<LIMIT>

Argument	Description	Data Type	Required
CLUSTER_ID	The cluster id.	string	True

Parameter	Description	Data Type	Required	Default
--json	Return JSON formatted logs.	marker	False	-
--limit	Show last number of logs, default 50. Default: `50`.	integer	False	50

Gets a cluster's secret.

sbctl cluster get-secret
    <CLUSTER_ID>

Argument	Description	Data Type	Required
CLUSTER_ID	The cluster id.	string	True

Updates a cluster's secret.

sbctl cluster update-secret
    <CLUSTER_ID>
    <SECRET>

Argument	Description	Data Type	Required
CLUSTER_ID	The cluster id.	string	True
SECRET	The new 20 characters password.	unknown	True

Updates a cluster's fabric.

sbctl cluster update-fabric
    <CLUSTER_ID>
    <FABRIC>

Argument	Description	Data Type	Required
CLUSTER_ID	The cluster id.	string	True
FABRIC	Fabric: tcp, rdma or both (specify: tcp, rdma). Default: `tcp`.	string	True

Checks a cluster's health.

sbctl cluster check
    <CLUSTER_ID>

Argument	Description	Data Type	Required
CLUSTER_ID	The cluster id.	string	True

Updates a cluster to new version.

Updates a the control plane to a new version. To update the storage nodes, they have to be shutdown and restarted. This can be done in a rolling manner. Attention: verify that an upgrade path is available and has been tested!"

sbctl cluster update
    <CLUSTER_ID>
    --cp-only=<CP_ONLY>
    --spdk-image=<SPDK_IMAGE>
    --mgmt-image=<MGMT_IMAGE>
    --max-subsys=<MAX_SUBSYS>
    --hugepages-mem=<HUGEPAGES_MEM>

Argument	Description	Data Type	Required
CLUSTER_ID	The cluster id.	string	True

Parameter	Description	Data Type	Required	Default
--cp-only	Update the control plane only. Default: `false`.	boolean	False	False
--spdk-image	Restart the storage nodes using the provided image.	string	False	-
--mgmt-image	Restart the management services using the provided image.	string	False	-
--max-subsys	Change the cluster-wide max nvmf subsystems per storage node. Applied by each node on its next restart. Given alone, no image update runs.	integer	False	-
--hugepages-mem	Change the cluster-wide huge-page memory floor per storage node, e.g. 4G. Applied by each node on its next restart. Given alone, no image update runs.	string	False	-

Completes a cluster upgrade.

Runs the completion step of a cluster upgrade started with cluster update: resumes JC compression on all storage nodes (deferred to a resume task on nodes with running data migrations) and stamps the installed release. All storage nodes must be online.

sbctl cluster upgrade-complete
    <CLUSTER_ID>

Argument	Description	Data Type	Required
CLUSTER_ID	The cluster id.	string	True

Backfills the database's secondary indices.

Walks every indexed table and fills in the index entries of records that predate their index declaration, then marks each index usable by reads. Idempotent, and safe to run against a live cluster: normal writes already maintain every index, and the backfill only adds entries. An upgrade runs this automatically; use the command after an interrupted run, or to fill an index that sbctl cluster index-state --set building has just put back into service.

sbctl cluster build-indices

Shows or switches the state of the database's secondary indices.

Without arguments, lists every declared index with its state (building, ready or disabled), its backfill cursor and when the state last changed. --set disabled takes one index out of service: writes stop maintaining it and reads fall back to a full scan, which is how a misbehaving index is switched off without touching the rest of its class. Its entries are left in place, so readers that have not seen the change yet still get right answers, and the command waits a few seconds before returning, until every process has seen it. --set building puts it back, waiting the same way before discarding those entries, after which sbctl cluster build-indices refills and re-enables it. An index reaches ready only by completing a backfill, so it cannot be set here.

sbctl cluster index-state
    <INDEX>
    --set=<SET>
    --json

Argument	Description	Data Type	Required
INDEX	The index to act on, as . (e.g. LVol.node_id). Omit to list them all.	string	True

Parameter	Description	Data Type	Required	Default
--set	Switch the named index to this state. Available Options: - building - disabled	string	False	-
--json	Print outputs in json format.	marker	False	-

Verifies the database's secondary indices.

Walks every index in both directions - each record's derived keys must exist, and each index key must resolve to a live record that still carries the indexed value - and reports what does not match. Read-only unless --repair is given, and takes nothing offline. Fails while anything is left unresolved: every finding of a read-only run, and the one thing a repair cannot settle - a unique value that two live records both carry, which has to be resolved in the data itself.

sbctl cluster check-indices
    --repair

Parameter	Description	Data Type	Required	Default
--repair	Write back the entries that can be derived from the records, and clear orphaned ones, instead of only reporting them.	marker	False	-

Initiates a graceful shutdown of a cluster's storage nodes.

sbctl cluster graceful-shutdown
    <CLUSTER_ID>

Argument	Description	Data Type	Required
CLUSTER_ID	The cluster id.	string	True

Performs a full cluster restart: shuts down every node that is not offline, restarts all nodes in parallel and reactivates the cluster.

sbctl cluster restart
    <CLUSTER_ID>

Argument	Description	Data Type	Required
CLUSTER_ID	The cluster id.	string	True

Initiates a graceful startup of a cluster's storage nodes.

sbctl cluster graceful-startup
    <CLUSTER_ID>
    --clear-data
    --spdk-image=<SPDK_IMAGE>

Argument	Description	Data Type	Required
CLUSTER_ID	The cluster id.	string	True

Parameter	Description	Data Type	Required	Default
--clear-data	Clear Alceml data.	marker	False	-
--spdk-image	The SPDK image URI.	string	False	-

Lists tasks of a cluster.

sbctl cluster list-tasks
    <CLUSTER_ID>
    --limit=<LIMIT>

Argument	Description	Data Type	Required
CLUSTER_ID	The cluster id.	string	True

Parameter	Description	Data Type	Required	Default
--limit	Show last number of tasks, default 50. Default: `50`.	integer	False	50

Cancels task by task id.

sbctl cluster cancel-task
    <TASK_ID>

Argument	Description	Data Type	Required
TASK_ID	The cluster task id.	string	True

Get rebalancing subtasks list.

sbctl cluster get-subtasks
    <TASK_ID>

Argument	Description	Data Type	Required
TASK_ID	The cluster task id.	string	True

Deletes a cluster.

This is only possible, if no storage nodes and pools are attached to the cluster

sbctl cluster delete
    <CLUSTER_ID>

Argument	Description	Data Type	Required
CLUSTER_ID	The cluster id.	string	True

Enable cluster-wide per-chunk data placement-binding for distrib bdevs (forward-only upgrade; --disable is reserved for debug).

Preflight: every storage node must be ONLINE; the cluster must be ACTIVE and not rebalancing. The runtime distr_shared_placement RPC is dispatched to every online node, then the flag is persisted on the cluster row and on every node's lvstore_stack distrib entries so subsequent restarts re-create with the new mode.

sbctl cluster set-shared-placement
    <CLUSTER_ID>
    --disable
    --force

Argument	Description	Data Type	Required
CLUSTER_ID	The cluster id.	string	True

Parameter	Description	Data Type	Required	Default
--disable	Reverse transition (per-chunk -> per-page). Debug only; only safe on a balanced or empty bdev. Requires --force.	marker	False	-
--force	Bypass the rebalancing / non-online-node guards. Required when --disable is passed.	marker	False	-

Provisions the Grafana alert rules read from the cluster event log.

Adds eight alert rules that read the cluster event log through /api/v2/clusters//logs instead of the Thanos metrics the rules in alert_rules.yaml read. The event log carries the transition an entity made, which is what lets these rules tell an operator's shutdown apart from a fault that ended in the same state, and lets them report a device removal and the two journal-compression conditions that no metric carries. Opt-in: the rules need a REST data source, so Grafana downloads a plugin (~74 MB) on the next restart. Works on a running cluster -- the provisioning files are written to every management node and the Grafana task is recreated, so Grafana is unavailable for a few seconds. Run it once, from any management node; re-run it to change a setting.

sbctl cluster event-alerts
    <CLUSTER_ID>
    --disable
    --log-limit=<LOG_LIMIT>
    --interval=<INTERVAL>
    --pending-period=<PENDING_PERIOD>
    --plugin-url=<PLUGIN_URL>
    --plugin-preinstalled

Argument	Description	Data Type	Required
CLUSTER_ID	The cluster id.	string	True

Parameter	Description	Data Type	Required	Default
--disable	Remove the event log alert rules and the data source.	marker	False	-
--log-limit	Too low on a cluster that produces events quickly and an alert heals itself as soon as the transition that opened it scrolls out of the window.	integer	False	-
--interval	How often the rules run, as a Grafana duration. Default: `1m`.	string	False	-
--pending-period	How long a condition must hold before it notifies, as a Grafana duration. Default: `1m`.	string	False	-
--plugin-url	Where to fetch the Infinity data source plugin from.	string	False	-
--plugin-preinstalled	The data source plugin is already in the Grafana image; do not download it.	marker	False	-

Activate v2 distrib write protection cluster-wide.

Needed once after upgrading a cluster from a release without v2; new clusters are created on v2 already. Sends the runtime distr_write_protection_v2 RPC to every ONLINE storage node -- the only way an already-created distrib gains v2, since a create parameter cannot retrofit an existing bdev. Only if every online node succeeds is the generation recorded, on the cluster row and on every node's lvstore_stack distrib entries, so later (re-)creations use the v2 parameter. Partial success records nothing, so a retry is a plain re-run. Offline nodes are not an error: they have no running bdev to migrate and come back on v2 at their next restart.

sbctl cluster switch-write-protection
    <CLUSTER_ID>

Argument	Description	Data Type	Required
CLUSTER_ID	The cluster id.	string	True

Assigns or changes a name to a cluster

sbctl cluster change-name
    <CLUSTER_ID>
    <NAME>

Argument	Description	Data Type	Required
CLUSTER_ID	The cluster id.	string	True
NAME	The new cluster name.	string	True

DEPRECATED: use replication-target-add and replication-policy-add. Assigns the single snapshot replication target cluster.

sbctl cluster add-replication
    <CLUSTER_ID>
    <TARGET_CLUSTER_ID>
    --timeout=<TIMEOUT>
    --target-pool=<TARGET_POOL>

Argument	Description	Data Type	Required
CLUSTER_ID	Cluster id	string	True
TARGET_CLUSTER_ID	Target Cluster id	string	True

Parameter	Description	Data Type	Required	Default
--timeout	Snapshot replication network timeout	integer	False	3600
--target-pool	Target cluster pool ID or name	string	False	-

Adds a named replication destination to a cluster (several are allowed)

sbctl cluster replication-target-add
    <CLUSTER_ID>
    <NAME>
    <TARGET_CLUSTER_ID>
    --target-pool=<TARGET_POOL>
    --timeout=<TIMEOUT>

Argument	Description	Data Type	Required
CLUSTER_ID	Source cluster id	string	True
NAME	Name of the replication target, unique per source cluster	string	True
TARGET_CLUSTER_ID	Destination cluster id	string	True

Parameter	Description	Data Type	Required	Default
--target-pool	Pool on the destination cluster (ID or name). Stored as a UUID.	string	False	-
--timeout	Replication network timeout in seconds. Default: `600`.	integer	False	-

Lists the replication targets of a cluster

sbctl cluster replication-target-list
    --cluster-id=<CLUSTER_ID>
    --json

Parameter	Description	Data Type	Required	Default
--cluster-id	Source cluster id	string	False	-
--json	Print outputs in json format.	marker	False	-

Removes a replication target. Refused while a policy still uses it.

sbctl cluster replication-target-remove
    <TARGET_ID>

Argument	Description	Data Type	Required
TARGET_ID	Replication target id	string	True

Fails over EVERY volume replicating to this target (site loss)

sbctl cluster replication-target-failover
    <TARGET_ID>
    --json

Argument	Description	Data Type	Required
TARGET_ID	Replication target id	string	True

Parameter	Description	Data Type	Required	Default
--json	Print outputs in json format.	marker	False	-

Adds a replication policy on a target, defining the replication cadence

sbctl cluster replication-policy-add
    <CLUSTER_ID>
    <NAME>
    --target=<TARGET>
    --interval-min=<INTERVAL_MIN>
    --mode=<MODE>
    --keep=<KEEP>
    --retention-schedule=<RETENTION_SCHEDULE>
    --rpo-target-sec=<RPO_TARGET_SEC>
    --consistency-group

Argument	Description	Data Type	Required
CLUSTER_ID	Source cluster id	string	True
NAME	Name of the policy, unique per source cluster	string	True

Parameter	Description	Data Type	Required	Default
--target	Replication target id or name	string	True	-
--interval-min	Cadence: minutes between internal replication snapshots. 0 replicates user snapshots only. Default: `1`.	integer	False	-
--mode	Replication mode. Default: `failover`. Available Options: - failover - migration	string	False	-
--keep	Replicated internal snapshots to retain on each side. Minimum (and default): `2`.	integer	False	-
--retention-schedule	Tiered retention, e.g. `15m:2h,1h:11h,1d:7d` - one snapshot every 15 minutes for the last 2 hours, then hourly for 11 hours, then daily for 7 days. Snapshots older than the total span are pruned. Empty (default) keeps the flat --keep behaviour.	string	False	-
--rpo-target-sec	Declared recovery point objective in seconds. RPO compliance is computed against this target instead of the derived lag budget. Unset means no declared objective.	integer	False	-
--consistency-group	All volumes attached to this policy form ONE consistency group: they must share an LVS (creation pins them to it), cadence snapshots are taken as one frozen group, and fail-over generations resolve group-wide.	marker	False	-

Lists the replication policies of a cluster

sbctl cluster replication-policy-list
    --cluster-id=<CLUSTER_ID>
    --json

Parameter	Description	Data Type	Required	Default
--cluster-id	Source cluster id	string	False	-
--json	Print outputs in json format.	marker	False	-

Removes a replication policy. Refused while a volume still follows it.

sbctl cluster replication-policy-remove
    <POLICY_ID>

Argument	Description	Data Type	Required
POLICY_ID	Replication policy id	string	True

Fails over EVERY volume following this policy

sbctl cluster replication-policy-failover
    <POLICY_ID>
    --json

Argument	Description	Data Type	Required
POLICY_ID	Replication policy id	string	True

Parameter	Description	Data Type	Required	Default
--json	Print outputs in json format.	marker	False	-

Takes ONE crash-consistent snapshot of every volume in the policy's consistency group, as a new group generation

sbctl cluster replication-policy-snapshot
    <POLICY_ID>

Argument	Description	Data Type	Required
POLICY_ID	Replication policy id (must be a consistency-group policy)	string	True