1NVME-DISCOVER(1) NVMe Manual NVME-DISCOVER(1)
2
6 nvme-discover - Send Get Log Page request to Discovery Controller.
7
9 nvme discover
10 [--transport=<trtype> | -t <trtype>]
11 [--traddr=<traddr> | -a <traddr>]
12 [--trsvcid=<trsvcid> | -s <trsvcid>]
13 [--host-traddr=<traddr> | -w <traddr>]
14 [--hostnqn=<hostnqn> | -q <hostnqn>]
15 [--raw=<filename> | -r <filename>]
16 [--reconnect-delay=<#> | -c <#>]
17 [--ctrl-loss-tmo=<#> | -l <#>]
18
20 Send one or more Get Log Page requests to a NVMe-over-Fabrics Discovery
21 Controller.
22 If no parameters are given, then nvme discover will attempt to find a
24 /etc/nvme/discovery.conf file to use to supply a list of Discovery
25 commands to run. If no /etc/nvme/discovery.conf file exists, the
26 command will quit with an error.
27 Otherwise, a specific Discovery Controller should be specified using
29 the --transport, --traddr, and if necessary the --trsvcid flags. A
30 Diѕcovery request will then be sent to the specified Discovery
31 Controller.
32
34 The NVMe-over-Fabrics specification defines the concept of a Discovery
35 Controller that an NVMe Host can query on a fabric network to discover
36 NVMe subsystems contained in NVMe Targets which it can connect to on
37 the network. The Discovery Controller will return Discovery Log Pages
38 that provide the NVMe Host with specific information (such as network
39 address and unique subsystem NQN) the NVMe Host can use to issue an
40 NVMe connect command to connect itself to a storage resource contained
41 in that NVMe subsystem on the NVMe Target.
42 Note that the base NVMe specification defines the NQN (NVMe Qualified
44 Name) format which an NVMe endpoint (device, subsystem, etc) must
45 follow to guarantee a unique name under the NVMe standard. In
46 particular, the Host NQN uniquely identifies the NVMe Host, and may be
47 used by the the Discovery Controller to control what NVMe Target
48 resources are allocated to the NVMe Host for a connection.
49 A Discovery Controller has it’s own NQN defined in the
51 NVMe-over-Fabrics specification, nqn.2014-08.org.nvmexpress.discovery.
52 All Discovery Controllers must use this NQN name. This NQN is used by
53 default by nvme-cli for the discover command.
54
56 -t <trtype>, --transport=<trtype>
57 This field specifies the network fabric being used for a
58 NVMe-over-Fabrics network. Current string values include:
59 ┌──────┬────────────────────────────┐
61 │Value │ Definition │
62 ├──────┼────────────────────────────┤
63 │rdma │ The network fabric is an │
64 │ │ rdma network (RoCE, iWARP, │
65 │ │ Infiniband, basic rdma, │
66 │ │ etc) │
67 ├──────┼────────────────────────────┤
68 │fc │ WIP The network fabric is │
69 │ │ a Fibre Channel network. │
70 ├──────┼────────────────────────────┤
71 │loop │ Connect to a NVMe over │
72 │ │ Fabrics target on the │
73 │ │ local host │
74 └──────┴────────────────────────────┘
75 -a <traddr>, --traddr=<traddr>
77 This field specifies the network address of the Discovery
78 Controller. For transports using IP addressing (e.g. rdma) this
79 should be an IP-based (ex. IPv4) address.
80 -s <trsvcid>, --trsvcid=<trsvcid>
82 This field specifies the transport service id. For transports using
83 IP addressing (e.g. rdma) this field is the port number. By
84 default, the IP port number for the RDMA transport is 4420.
85 -q <hostnqn>, --hostnqn=<hostnqn>
87 Overrides the default host NQN that identifies the NVMe Host. If
88 this option is not specified, the default is read from
89 /etc/nvme/hostnqn first. If that does not exist, the autogenerated
90 NQN value from the NVMe Host kernel module is used next.
91 -w <traddr>, --host-traddr=<traddr>
93 This field specifies the network address used on the host to
94 connect to the Discovery Controller.
95 -r <filename>, --raw=<filename>
97 This field will take the output of the nvme discover command and
98 dump it to a raw binary file. By default nvme discover will dump
99 the output to stdout.
100 -c <#>, --reconnect-delay=<#>
102 Overrides the default delay (in seconds) before reconnect is
103 attempted after a connect loss.
104 -l <#>, --ctrl-loss-tmo=<#>
106 Overrides the default controller loss timeout period (in seconds).
107
109 · Query the Discover Controller with IP4 address 192.168.1.3 for all
110 resources allocated for NVMe Host name host1-rogue-nqn on the RDMA
111 network. Port 4420 is used by default:
112 # nvme discover --transport=rdma --traddr=192.168.1.3 \
114 --hostnqn=host1-rogue-nqn
115 · Issue a nvme discover command using a /etc/nvme/discovery.conf
117 file:
118 # Machine default 'nvme discover' commands. Query the
120 # Discovery Controller's two ports (some resources may only
121 # be accessible on a single port). Note an official
122 # nqn (Host) name defined in the NVMe specification is being used
123 # in this example.
124 -t rdma -a 192.168.69.33 -s 4420 -q nqn.2014-08.com.example:nvme:nvm-subsystem-sn-d78432
125 -t rdma -a 192.168.1.4 -s 4420 -q nqn.2014-08.com.example:nvme:nvm-subsystem-sn-d78432
126 At the prompt type "nvme discover".
128
130 nvme-connect(1) nvme-connect-all(1)
131
133 This was written by Jay Freyensee[1] for Keith Busch[2].
134
136 Part of the nvme-user suite
137
139 1. Jay Freyensee
140 mailto:james.p.freyensee@intel.com
141 2. Keith Busch
143 mailto:keith.busch@intel.com
144NVMe 01/08/2019 NVME-DISCOVER(1)