NVME-DISCONNECT(1) NVMe Manual NVME-DISCONNECT(1) NAME nvme-disconnect - Disconnect one or more Fabrics controller(s). SYNOPSIS nvme [] disconnect [--device= | -d ] [--exclude | -x] [] DESCRIPTION Disconnects and removes one or more existing NVMe over Fabrics controllers. If the --device option is specified, the controller with that name will be removed. If --nqn is given by itself, every controller connected to the subsystem identified by subnqn will be removed. If --nqn is combined with any of the other fabrics options below (the same ones nvme connect accepts, such as --transport, --traddr, --trsvcid, --host-traddr, --host-iface, --hostnqn or --hostid), only the controller(s) matching all of the specified options are removed -- useful for disambiguating between multiple controllers connected to the same subsystem. --device cannot be combined with --nqn or any of the other fabrics options. With --exclude, a matching entry is written to the system-wide NVMe-oF exclusion list before the controller (or subsystem) is torn down, so that a cooperating orchestrator sees the exclusion in place before the removal event fires and does not immediately reconnect. See nvme-exclusion-add(1) and the exclusion list overview for details. OPTIONS -d , --device Indicates that the controller with the specified name should be removed. Cannot be combined with --nqn or any of the fabrics options below. -x, --exclude Write a matching entry to the user exclusion list before disconnecting, so orchestrators do not auto-reconnect the controller. The entry is built from the controller's transport parameters (or, with a bare --nqn, from the subsystem NQN). FABRICS OPTIONS The following options are common to NVMe over Fabrics commands such as nvme connect: -a , --traddr= Specify the network address of the controller. For transports using IP addressing (e.g. rdma, tcp), this should be an IP address. --concat Enable secure concatenation (TCP). -c <#>, --reconnect-delay=<#> Set the delay in seconds before reconnect is attempted after a connection loss. -C , --kxchap-ctrl-secret= Controller authentication secret for bi-directional authentication. If not specified, bi-directional authentication is not attempted. -S , --kxchap-secret= Host authentication secret (KX-HMAC-CHAP). Must be provided in ASCII format as defined in the NVMe specification. -C , --dhchap-ctrl-secret= Controller authentication secret for bi-directional authentication. If not specified, bi-directional authentication is not attempted. (deprecated, see --kxchap-ctrl-secret) -S , --dhchap-secret= Host authentication secret (DH-HMAC-CHAP). Must be provided in ASCII format as defined in the NVMe specification. (deprecated, see --kxchap-secret) --disable-sqflow Disable submission queue flow control. -G, --data-digest Enable data digest generation/verification (TCP). -D, --duplicate-connect Allow duplicate connections to the same subsystem. -g, --hdr-digest Enable header digest generation/verification (TCP). -f , --host-iface= Specify the network interface to use for the connection. -I , --hostid= Specify the host UUID. -q , --hostnqn= Override the default host NQN. -w , --host-traddr= Specify the source address on the host side of the connection. -k <#>, --keep-alive-tmo=<#> Set the keep-alive timeout in seconds. --keyring= Keyring to use for TLS key lookup. -n , --nqn= Specify the NVMe subsystem NQN to connect to. -i <#>, --nr-io-queues=<#> Number of I/O queues to create. -P <#>, --nr-poll-queues=<#> Number of polling queues to create. -W <#>, --nr-write-queues=<#> Number of write queues to create. -Q <#>, --queue-size=<#> Queue depth for I/O queues. -l <#>, --ctrl-loss-tmo=<#> Maximum time in seconds to retry reconnect attempts after controller loss. -s , --trsvcid= Transport service identifier (e.g. TCP/rdma port). Default is 4420 for RDMA. -T <#>, --tos=<#> Type of service value for the connection (TCP). -t , --transport= Specify the transport type. Supported values include: + +------+--------------------------+ |Value | Description | +------+--------------------------+ |rdma | RDMA (RoCE, iWARP, | | | InfiniBand) | +------+--------------------------+ |tcp | TCP/IP | +------+--------------------------+ |fc | Fibre Channel | | | (experimental) | +------+--------------------------+ |loop | Local loopback transport | +------+--------------------------+ --tls Enable TLS encryption (TCP). --tls-key= TLS key for the connection. It is recommended to preload keys into the system keyring instead of passing them via the command line. --tls-key-identity= Identity associated with the TLS key. GLOBAL OPTIONS The following options are defined at the top-level nvme command and are available to this subcommand: --dry-run Print the command that would be executed, but do not actually execute it. --no-ioctl-probing Disable probing for 64-bit IOCTL support. --no-retries Disable retry logic on transient errors. -o , --output-format= Set the reporting format to normal, tabular, 'json, or binary. Only one output format may be used at a time. --output-format-version= Select the output format version. Version 1 uses the original field naming, while version 2 (default) provides more consistent and script-friendly field names. --timeout= Set the timeout for the command in milliseconds. -v, --verbose Increase the level of detail in the output. May be specified multiple times to further increase verbosity. These options can also be set as machine-wide defaults in nvme-cli.conf(5). A command-line flag always overrides the file. EXAMPLES o Disconnect all controllers for a subsystem named nqn.2014-08.com.example:nvme:nvm-subsystem-sn-d78432: # nvme disconnect --nqn=nqn.2014-08.com.example:nvme:nvm-subsystem-sn-d78432 o Disconnect the controller nvme4 # nvme disconnect --device=nvme4 o Disconnect only the tcp controller for a subsystem at a given traddr, leaving other controllers of the same subsystem (e.g. over a different transport address) connected: # nvme disconnect --nqn=nqn.2014-08.com.example:nvme:nvm-subsystem-sn-d78432 \ --transport=tcp --traddr=192.168.1.100 o Disconnect nvme4 and exclude it so orchestrators do not reconnect it: # nvme disconnect --device=nvme4 --exclude SEE ALSO nvme-connect(1) nvme-exclusion-add(1) NVME Part of the nvme-user suite NVMe 09/07/2026 NVME-DISCONNECT(1)