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 [--hostid=<hostid> | -I <hostid>]
16 [--raw=<filename> | -r <filename>]
17 [--keep-alive-tmo=<sec> | -k <sec>]
18 [--reconnect-delay=<#> | -c <#>]
19 [--ctrl-loss-tmo=<#> | -l <#>]
20 [--hdr_digest | -g]
21 [--data_digest | -G]
22 [--nr-io-queues=<#> | -i <#>]
23 [--nr-write-queues=<#> | -W <#>]
24 [--nr-poll-queues=<#> | -P <#>]
25 [--queue-size=<#> | -Q <#>]
26 [--persistent | -p]
27 [--quiet | -S]
28
30 Send one or more Get Log Page requests to a NVMe-over-Fabrics Discovery
31 Controller.
32 If no parameters are given, then nvme discover will attempt to find a
34 /etc/nvme/discovery.conf file to use to supply a list of Discovery
35 commands to run. If no /etc/nvme/discovery.conf file exists, the
36 command will quit with an error.
37 Otherwise, a specific Discovery Controller should be specified using
39 the --transport, --traddr, and if necessary the --trsvcid flags. A
40 Diѕcovery request will then be sent to the specified Discovery
41 Controller.
42
44 The NVMe-over-Fabrics specification defines the concept of a Discovery
45 Controller that an NVMe Host can query on a fabric network to discover
46 NVMe subsystems contained in NVMe Targets which it can connect to on
47 the network. The Discovery Controller will return Discovery Log Pages
48 that provide the NVMe Host with specific information (such as network
49 address and unique subsystem NQN) the NVMe Host can use to issue an
50 NVMe connect command to connect itself to a storage resource contained
51 in that NVMe subsystem on the NVMe Target.
52 Note that the base NVMe specification defines the NQN (NVMe Qualified
54 Name) format which an NVMe endpoint (device, subsystem, etc) must
55 follow to guarantee a unique name under the NVMe standard. In
56 particular, the Host NQN uniquely identifies the NVMe Host, and may be
57 used by the the Discovery Controller to control what NVMe Target
58 resources are allocated to the NVMe Host for a connection.
59 A Discovery Controller has it’s own NQN defined in the
61 NVMe-over-Fabrics specification, nqn.2014-08.org.nvmexpress.discovery.
62 All Discovery Controllers must use this NQN name. This NQN is used by
63 default by nvme-cli for the discover command.
64
66 -t <trtype>, --transport=<trtype>
67 This field specifies the network fabric being used for a
68 NVMe-over-Fabrics network. Current string values include:
69 ┌──────┬────────────────────────────┐
71 │Value │ Definition │
72 ├──────┼────────────────────────────┤
73 │rdma │ The network fabric is an │
74 │ │ rdma network (RoCE, iWARP, │
75 │ │ Infiniband, basic rdma, │
76 │ │ etc) │
77 ├──────┼────────────────────────────┤
78 │fc │ WIP The network fabric is │
79 │ │ a Fibre Channel network. │
80 ├──────┼────────────────────────────┤
81 │loop │ Connect to a NVMe over │
82 │ │ Fabrics target on the │
83 │ │ local host │
84 └──────┴────────────────────────────┘
85 -a <traddr>, --traddr=<traddr>
87 This field specifies the network address of the Discovery
88 Controller. For transports using IP addressing (e.g. rdma) this
89 should be an IP-based (ex. IPv4) address.
90 -s <trsvcid>, --trsvcid=<trsvcid>
92 This field specifies the transport service id. For transports using
93 IP addressing (e.g. rdma) this field is the port number. By
94 default, the IP port number for the RDMA transport is 4420.
95 -w <traddr>, --host-traddr=<traddr>
97 This field specifies the network address used on the host to
98 connect to the Discovery Controller.
99 -q <hostnqn>, --hostnqn=<hostnqn>
101 Overrides the default host NQN that identifies the NVMe Host. If
102 this option is not specified, the default is read from
103 /etc/nvme/hostnqn first. If that does not exist, the autogenerated
104 NQN value from the NVMe Host kernel module is used next.
105 -I <hostid>, --hostid=<hostid>
107 UUID(Universally Unique Identifier) to be discovered which should
108 be formatted.
109 -r <filename>, --raw=<filename>
111 This field will take the output of the nvme discover command and
112 dump it to a raw binary file. By default nvme discover will dump
113 the output to stdout.
114 -k <#>, --keep-alive-tmo=<#>
116 Overrides the default dealy (in seconds) for keep alive. This
117 option will be ignored for the discovery, and it is only
118 implemented for completeness.
119 -c <#>, --reconnect-delay=<#>
121 Overrides the default delay (in seconds) before reconnect is
122 attempted after a connect loss.
123 -l <#>, --ctrl-loss-tmo=<#>
125 Overrides the default controller loss timeout period (in seconds).
126 -g, --hdr_digest
128 Generates/verifies header digest (TCP).
129 -G, --data_digest
131 Generates/verifies data digest (TCP).
132 -i <#>, --nr-io-queues=<#>
134 Overrides the default number of I/O queues create by the driver.
135 This option will be ignored for the discovery, and it is only
136 implemented for completeness.
137 -W <#>, --nr-write-queues=<#>
139 Adds additional queues that will be used for write I/O.
140 -P <#>, --nr-poll-queues=<#>
142 Adds additional queues that will be used for polling latency
143 sensitive I/O.
144 -Q <#>, --queue-size=<#>
146 Overrides the default number of elements in the I/O queues created
147 by the driver which can be found at drivers/nvme/host/fabrics.h.
148 This option will be ignored for the discovery, and it is only
149 implemented for completeness.
150 -p, --persistent
152 Persistent discovery connection.
153 -S, --quiet
155 Suppress already connected errors.
156
158 · Query the Discover Controller with IP4 address 192.168.1.3 for all
159 resources allocated for NVMe Host name host1-rogue-nqn on the RDMA
160 network. Port 4420 is used by default:
161 # nvme discover --transport=rdma --traddr=192.168.1.3 \
163 --hostnqn=host1-rogue-nqn
164 · Issue a nvme discover command using a /etc/nvme/discovery.conf
166 file:
167 # Machine default 'nvme discover' commands. Query the
169 # Discovery Controller's two ports (some resources may only
170 # be accessible on a single port). Note an official
171 # nqn (Host) name defined in the NVMe specification is being used
172 # in this example.
173 -t rdma -a 192.168.69.33 -s 4420 -q nqn.2014-08.com.example:nvme:nvm-subsystem-sn-d78432
174 -t rdma -a 192.168.1.4 -s 4420 -q nqn.2014-08.com.example:nvme:nvm-subsystem-sn-d78432
175 At the prompt type "nvme discover".
177
179 nvme-connect(1) nvme-connect-all(1)
180
182 This was written by Jay Freyensee[1]
183
185 Part of the nvme-user suite
186
188 1. Jay Freyensee
189 mailto:james.p.freyensee@intel.com
190NVMe 04/24/2020 NVME-DISCOVER(1)