freebsd-src/usr.sbin/nvmfd/nvmfd.8
John Baldwin a8089ea5ae nvmfd: A simple userspace daemon for the NVMe over Fabrics controller
This daemon can operate as a purely userspace controller exporting one
or more simulated RAM disks or local block devices as NVMe namespaces
to a remote host.  In this case the daemon provides a discovery
controller with a single entry for an I/O controller.

nvmfd can also offload I/O controller queue pairs to the nvmft.ko
in-kernel Fabrics controller when -K is passed.  In this mode, nvmfd
still accepts connections and performs initial transport-specific
negotitation in userland.  The daemon still provides a userspace-only
discovery controller with a single entry for an I/O controller.
However, queue pairs for the I/O controller are handed off to the CTL
NVMF frontend.

Eventually ctld(8) should be refactored to to provide an abstraction
for the frontend protocol and the discovery and the kernel mode of
this daemon should be merged into ctld(8).  At that point this daemon
can be moved to tools/tools/nvmf as a debugging tool (mostly as sample
code for a userspace controller using libnvmf).

Reviewed by:	imp
Sponsored by:	Chelsio Communications
Differential Revision:	https://reviews.freebsd.org/D44731
2024-05-02 16:38:39 -07:00

127 lines
2.8 KiB
Groff

.\"
.\" SPDX-License-Identifier: BSD-2-Clause
.\"
.\" Copyright (c) 2024 Chelsio Communications, Inc.
.\"
.Dd May 2, 2024
.Dt NVMFD 8
.Os
.Sh NAME
.Nm nvmfd
.Nd "NVMeoF controller daemon"
.Sh SYNOPSIS
.Nm
.Fl K
.Op Fl dFGg
.Op Fl P Ar port
.Op Fl p Ar port
.Op Fl t Ar transport
.Op Fl n Ar subnqn
.Nm
.Op Fl dFGg
.Op Fl P Ar port
.Op Fl p Ar port
.Op Fl t Ar transport
.Op Fl n Ar subnqn
.Ar device
.Op Ar device ...
.Sh DESCRIPTION
.Nm
accepts incoming NVMeoF connections for both I/O and discovery controllers.
.Nm
can either implement a single dynamic I/O controller in user mode or hand
off incoming I/O controller connections to
.Xr nvmft 4 .
A dynamic discovery controller service is always provided in user mode.
.Pp
The following options are available:
.Bl -tag -width "-t transport"
.It Fl F
Permit remote hosts to disable SQ flow control.
.It Fl G
Permit remote hosts to enable PDU data digests for the TCP transport.
.It Fl g
Permit remote hosts to enable PDU header digests for the TCP transport.
.It Fl K
Enable kernel mode which hands off incoming I/O controller connections to
.Xr nvmft 4 .
.It Fl P Ar port
Use
.Ar port
as the listen TCP port for the discovery controller service.
The default value is 8009.
.It Fl d
Enable debug mode.
The daemon sends any errors to standard output and does not place
itself in the background.
.It Fl p Ar port
Use
.Ar port
as the listen TCP port for the I/O controller service.
By default an unused ephemeral port will be chosen.
.It Fl n Ar subnqn
The Subsystem NVMe Qualified Name for the I/O controller.
If an explicit NQN is not given, a default value is generated from the
current host's UUID obtained from the
.Vt kern.hostuuid
sysctl.
.It Fl t Ar transport
The transport type to use.
The default transport is
.Dq tcp .
.It Ar device
When implementing a user mode I/O controller,
one or more
.Ar device
arguments must be specified.
Each
.Ar device
describes the backing store for a namespace exported to remote hosts.
Devices can be specified using one of the following syntaxes:
.Bl -tag -width "ramdisk:size"
.It Pa pathname
File or disk device
.It ramdisk : Ns Ar size
Allocate a memory disk with the given
.Ar size .
.Ar size
may use any of the suffixes supported by
.Xr expand_number 3 .
.El
.El
.Sh FILES
.Bl -tag -width "/var/run/nvmfd.pid" -compact
.It Pa /var/run/nvmfd.pid
The default location of the
.Nm
PID file.
.El
.Sh EXIT STATUS
.Ex -std
.Sh SEE ALSO
.Xr ctl 4 ,
.Xr nvmft 4 ,
.Xr ctladm 8 ,
.Xr ctld 8
.Sh HISTORY
The
.Nm
module first appeared in
.Fx 15.0 .
.Sh AUTHORS
The
.Nm
subsystem was developed by
.An John Baldwin Aq Mt jhb@FreeBSD.org
under sponsorship from Chelsio Communications, Inc.
.Sh BUGS
The discovery controller and kernel mode functionality of
.Nm
should be merged into
.Xr ctld 8 .
.Pp
Additional paramters such as
.Va MAXR2T ,
.Va MAXH2CDATA ,
and queue sizes should be configurable.