freebsd-skq/sbin/hastd/hast.conf.5
pjd 1c1e2e8b71 Please welcome HAST - Highly Avalable Storage.
HAST allows to transparently store data on two physically separated machines
connected over the TCP/IP network. HAST works in Primary-Secondary
(Master-Backup, Master-Slave) configuration, which means that only one of the
cluster nodes can be active at any given time. Only Primary node is able to
handle I/O requests to HAST-managed devices. Currently HAST is limited to two
cluster nodes in total.

HAST operates on block level - it provides disk-like devices in /dev/hast/
directory for use by file systems and/or applications. Working on block level
makes it transparent for file systems and applications. There in no difference
between using HAST-provided device and raw disk, partition, etc. All of them
are just regular GEOM providers in FreeBSD.

For more information please consult hastd(8), hastctl(8) and hast.conf(5)
manual pages, as well as http://wiki.FreeBSD.org/HAST.

Sponsored by:	FreeBSD Foundation
Sponsored by:	OMCnet Internet Service GmbH
Sponsored by:	TransIP BV
2010-02-18 23:16:19 +00:00

268 lines
6.9 KiB
Groff

.\" Copyright (c) 2010 The FreeBSD Foundation
.\" All rights reserved.
.\"
.\" This software was developed by Pawel Jakub Dawidek under sponsorship from
.\" the FreeBSD Foundation.
.\"
.\" Redistribution and use in source and binary forms, with or without
.\" modification, are permitted provided that the following conditions
.\" are met:
.\" 1. Redistributions of source code must retain the above copyright
.\" notice, this list of conditions and the following disclaimer.
.\" 2. Redistributions in binary form must reproduce the above copyright
.\" notice, this list of conditions and the following disclaimer in the
.\" documentation and/or other materials provided with the distribution.
.\"
.\" THIS SOFTWARE IS PROVIDED BY THE AUTHORS AND CONTRIBUTORS ``AS IS'' AND
.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
.\" ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHORS OR CONTRIBUTORS BE LIABLE
.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
.\" SUCH DAMAGE.
.\"
.\" $FreeBSD$
.\"
.Dd February 1, 2010
.Dt HAST.CONF 5
.Os
.Sh NAME
.Nm hast.conf
.Nd configuration file for the
.Xr hastd 8
deamon and the
.Xr hastctl 8
utility.
.Sh DESCRIPTION
The
.Nm
file is used by both
.Xr hastd 8
daemon
and
.Xr hastctl 8
control utility.
Configuration file is designed in a way that exactly the same file can be
(and should be) used on both HAST nodes.
Every line starting with # is treated as comment and ignored.
.Sh CONFIGURATION FILE SYNTAX
General syntax of the
.Nm
file is following:
.Bd -literal -offset indent
# Global section
control <addr>
listen <addr>
replication <mode>
on <node> {
# Node section
control <addr>
listen <addr>
}
on <node> {
# Node section
control <addr>
listen <addr>
}
resource <name> {
# Resource section
replication <mode>
name <name>
local <path>
on <node> {
# Resource-node section
name <name>
# Required
local <path>
# Required
remote <addr>
}
on <node> {
# Resource-node section
name <name>
# Required
local <path>
# Required
remote <addr>
}
}
.Ed
.Pp
Most of the various available configuration parameters are optional.
If parameter is not defined in the particular section, it will be
inherited from the parent section.
For example, if the
.Ic listen
parameter is not defined in the node section, it will be inherited from
the global section.
In case the global section does not define the
.Ic listen
parameter at all, the default value will be used.
.Sh CONFIGURATION FILE DESCRIPTION
The
.Aq node
argument can be replaced either by a full hostname as obtained by
.Xr gethostname 3 ,
only first part of the hostname, or by node's UUID as found in the
.Va kern.hostuuid
.Xr sysctl 8
variable.
.Pp
The following statements are available:
.Bl -tag -width ".Ic xxxx"
.It Ic control Aq addr
.Pp
Address for communication with
.Xr hastctl 8 .
Each of the following examples defines the same control address:
.Bd -literal -offset indent
uds:///var/run/hastctl
unix:///var/run/hastctl
/var/run/hastctl
.Ed
.Pp
The default value is
.Pa uds:///var/run/hastctl .
.It Ic listen Aq addr
.Pp
Address to listen on in form of:
.Bd -literal -offset indent
protocol://protocol-specific-address
.Ed
.Pp
Each of the following examples defines the same listen address:
.Bd -literal -offset indent
0.0.0.0
0.0.0.0:8457
tcp://0.0.0.0
tcp://0.0.0.0:8457
tcp4://0.0.0.0
tcp4://0.0.0.0:8457
.Ed
.Pp
The default value is
.Pa tcp4://0.0.0.0:8457 .
.It Ic replication Aq mode
.Pp
Replication mode should be one of the following:
.Bl -tag -width ".Ic xxxx"
.It Ic memsync
.Pp
Report the write operation as completed when local write completes and
when the remote node acknowledges the data receipt, but before it
actually stores the data.
The data on remote node will be stored directly after sending
acknowledgement.
This mode is intended to reduce latency, but still provides a very good
reliability.
The only situation where some small amount of data could be lost is when
the data is stored on primary node and sent to the secondary.
Secondary node then acknowledges data receipt and primary reports
success to an application.
However, it may happen that the seconderay goes down before the received
data is really stored locally.
Before secondary node returns, primary node dies entirely.
When the secondary node comes back to life it becomes the new primary.
Unfortunately some small amount of data which was confirmed to be stored
to the application was lost.
The risk of such a situation is very small, which is the reason for this
mode to be the default.
.It Ic fullsync
.Pp
Mark the write operation as completed when local as well as remote
write completes.
This is the safest and the slowest replication mode.
The
.Ic fullsync
replication mode is currently not implemented.
.It Ic async
.Pp
The write operation is reported as complete right after the local write
completes.
This is the fastest and the most dangerous replication mode.
This mode should be used when replicating to a distant node where
latency is too high for other modes.
The
.Ic async
replication mode is currently not implemented.
.El
.It Ic name Aq name
.Pp
GEOM provider name that will appear as
.Pa /dev/hast/<name> .
If name is not defined, resource name will be used as provider name.
.It Ic local Aq path
.Pp
Path to the local component which will be used as backend provider for
the resource.
This can be either GEOM provider or regular file.
.It Ic remote Aq addr
.Pp
Address of the remote
.Nm hastd
daemon.
Format is the same as for the
.Ic listen
statement.
When operating as a primary node this address will be used to connect to
the secondary node.
When operating as a secondary node only connections from this address
will be accepted.
.El
.Sh EXAMPLES
The example configuration file can look as follows:
.Bd -literal -offset indent
resource shared {
local /dev/da0
on hasta {
remote tcp4://10.0.0.2
}
on hastb {
remote tcp4://10.0.0.1
}
}
resource tank {
on hasta {
local /dev/mirror/tanka
remote tcp4://10.0.0.2
}
on hastb {
local /dev/mirror/tankb
remote tcp4://10.0.0.1
}
}
.Ed
.Sh FILES
.Bl -tag -width ".Pa /var/run/hastctl" -compact
.It Pa /etc/hast.conf
The default
.Nm
configuration file.
.It Pa /var/run/hastctl
Control socket used by the
.Xr hastctl 8
control utility to communicate with the
.Xr hastd 8
daemon.
.El
.Sh SEE ALSO
.Xr gethostname 3 ,
.Xr geom 4 ,
.Xr hastctl 8 ,
.Xr hastd 8 .
.Sh AUTHORS
The
.Nm
was written by
.An Pawel Jakub Dawidek Aq pjd@FreeBSD.org
under sponsorship of the FreeBSD Foundation.