xref: /freebsd/sbin/hastd/hastd.8 (revision 7aa65846327fe5bc7e5961c2f7fd0c61f2ec0b01)
1.\" Copyright (c) 2010 The FreeBSD Foundation
2.\" All rights reserved.
3.\"
4.\" This software was developed by Pawel Jakub Dawidek under sponsorship from
5.\" the FreeBSD Foundation.
6.\"
7.\" Redistribution and use in source and binary forms, with or without
8.\" modification, are permitted provided that the following conditions
9.\" are met:
10.\" 1. Redistributions of source code must retain the above copyright
11.\"    notice, this list of conditions and the following disclaimer.
12.\" 2. Redistributions in binary form must reproduce the above copyright
13.\"    notice, this list of conditions and the following disclaimer in the
14.\"    documentation and/or other materials provided with the distribution.
15.\"
16.\" THIS SOFTWARE IS PROVIDED BY THE AUTHORS AND CONTRIBUTORS ``AS IS'' AND
17.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
18.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
19.\" ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHORS OR CONTRIBUTORS BE LIABLE
20.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
21.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
22.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
23.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
24.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
25.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
26.\" SUCH DAMAGE.
27.\"
28.\" $FreeBSD$
29.\"
30.Dd February 1, 2010
31.Dt HASTD 8
32.Os
33.Sh NAME
34.Nm hastd
35.Nd "Highly Available Storage daemon"
36.Sh SYNOPSIS
37.Nm
38.Op Fl dFh
39.Op Fl c Ar config
40.Op Fl P Ar pidfile
41.Sh DESCRIPTION
42The
43.Nm
44daemon is responsible for managing highly available GEOM providers.
45.Pp
46.Nm
47allows to transparently store data on two physically separated machines
48connected over the TCP/IP network.
49Only one machine (cluster node) can actively use storage provided by
50.Nm .
51This machine is called primary.
52The
53.Nm
54daemon operates on block level, which makes it transparent for file
55systems and applications.
56.Pp
57There is one main
58.Nm
59daemon which starts new worker process as soon as a role for the given
60resource is changed to primary or as soon as a role for the given
61resource is changed to secondary and remote (primary) node will
62successfully connect to it.
63Every worker process gets a new process title (see
64.Xr setproctitle 3 ) ,
65which describes its role and resource it controls.
66The exact format is:
67.Bd -literal -offset indent
68hastd: <resource name> (<role>)
69.Ed
70.Pp
71When (and only when)
72.Nm
73operates in primary role for the given resource, corresponding
74.Pa /dev/hast/<name>
75disk-like device (GEOM provider) is created.
76File systems and applications can use this provider to send I/O
77requests to.
78Every write, delete and flush operation
79.Dv ( BIO_WRITE , BIO_DELETE , BIO_FLUSH )
80is send to local component and synchronously replicated
81to the remote (secondary) node if it is available.
82Read operations
83.Dv ( BIO_READ )
84are handled locally unless I/O error occurs or local version of the data
85is not up-to-date yet (synchronization is in progress).
86.Pp
87The
88.Nm
89daemon uses the GEOM Gate class to receive I/O requests from the
90in-kernel GEOM infrastructure.
91The
92.Nm geom_gate.ko
93module is loaded automatically if the kernel was not compiled with the
94following option:
95.Bd -ragged -offset indent
96.Cd "options GEOM_GATE"
97.Ed
98.Pp
99The connection between two
100.Nm
101daemons is always initiated from the one running as primary to the one
102running as secondary.
103When primary
104.Nm
105is unable to connect or connection fails, it will try to re-establish
106connection every few seconds.
107Once connection is established, primary
108.Nm
109will synchronize every extent that was modified during connection outage
110to the secondary
111.Nm .
112.Pp
113It is possible that in case of connection outage between the nodes
114.Nm
115primary role for the given resource will be configured on both nodes.
116This in turn leads to incompatible data modifications.
117Such condition is called split-brain and cannot be automatically
118resolved by the
119.Nm
120daemon as this will lead most likely to data corruption or lost of
121important changes.
122Even though it cannot be fixed by
123.Nm
124itself, it will be detected and further connection between independently
125modified nodes will not be possible.
126Once this situation is manually resolved by an administrator, resource
127on one of the nodes can be initialized (erasing local data), which makes
128connection to the remote node possible again.
129Connection of freshly initialized component will trigger full resource
130synchronization.
131.Pp
132The
133.Nm
134daemon itself never picks his role up automatically.
135The role has to be configured with the
136.Xr hastctl 8
137control utility by additional software like
138.Nm ucarp
139or
140.Nm heartbeat
141that can reliably manage role separation and switch secondary node to
142primary role in case of original primary failure.
143.Pp
144The
145.Nm
146daemon can be started with the following command line arguments:
147.Bl -tag -width ".Fl P Ar pidfile"
148.It Fl c Ar config
149Specify alternative location of the configuration file.
150The default location is
151.Pa /etc/hast.conf .
152.It Fl d
153Print or log debugging information.
154This option can be specified multiple times to raise the verbosity
155level.
156.It Fl F
157Start the
158.Nm
159daemon in the foreground.
160By default
161.Nm
162starts in the background.
163.It Fl h
164Print the
165.Nm
166usage message.
167.It Fl P Ar pidfile
168Specify alternative location of a file where main process PID will be
169stored.
170The default location is
171.Pa /var/run/hastd.pid .
172.El
173.Sh FILES
174.Bl -tag -width ".Pa /var/run/hastctl" -compact
175.It Pa /etc/hast.conf
176The configuration file for
177.Nm
178and
179.Xr hastctl 8 .
180.It Pa /var/run/hastctl
181Control socket used by the
182.Xr hastctl 8
183control utility to communicate with
184.Nm .
185.It Pa /var/run/hastd.pid
186The default location of the
187.Nm
188PID file.
189.El
190.Sh EXIT STATUS
191Exit status is 0 on success, or one of the values described in
192.Xr sysexits 3
193on failure.
194.Sh EXAMPLES
195Launch
196.Nm
197on both nodes.
198Set role for resource
199.Nm shared
200to primary on
201.Nm nodeA
202and to secondary on
203.Nm nodeB .
204Create file system on
205.Pa /dev/hast/shared
206provider and mount it.
207.Bd -literal -offset indent
208nodeB# hastd
209nodeB# hastctl role secondary shared
210
211nodeA# hastd
212nodeA# hastctl role primary shared
213nodeA# newfs -U /dev/hast/shared
214nodeA# mount -o noatime /dev/hast/shared /shared
215.Ed
216.Sh SEE ALSO
217.Xr sysexits 3 ,
218.Xr geom 4 ,
219.Xr hast.conf 5 ,
220.Xr ggatec 8 ,
221.Xr ggated 8 ,
222.Xr ggatel 8 ,
223.Xr hastctl 8 ,
224.Xr mount 8 ,
225.Xr newfs 8 ,
226.Xr g_bio 9 .
227.Sh AUTHORS
228The
229.Nm
230was developed by
231.An Pawel Jakub Dawidek Aq pjd@FreeBSD.org
232under sponsorship of the FreeBSD Foundation.
233