2 Bacula® - The Network Backup Solution
4 Copyright (C) 2003-2008 Free Software Foundation Europe e.V.
6 The main author of Bacula is Kern Sibbald, with contributions from
7 many others, a complete list can be found in the file AUTHORS.
8 This program is Free Software; you can redistribute it and/or
9 modify it under the terms of version two of the GNU General Public
10 License as published by the Free Software Foundation and included
13 This program is distributed in the hope that it will be useful, but
14 WITHOUT ANY WARRANTY; without even the implied warranty of
15 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
16 General Public License for more details.
18 You should have received a copy of the GNU General Public License
19 along with this program; if not, write to the Free Software
20 Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
23 Bacula® is a registered trademark of Kern Sibbald.
24 The licensor of Bacula is the Free Software Foundation Europe
25 (FSFE), Fiduciary Program, Sumatrastrasse 25, 8006 Zürich,
26 Switzerland, email:ftf@fsfeurope.org.
29 * Bacula File Daemon heartbeat routines
30 * Listens for heartbeats coming from the SD
31 * If configured, sends heartbeats to Dir
33 * Kern Sibbald, May MMIII
42 #define WAIT_INTERVAL 5
44 extern "C" void *sd_heartbeat_thread(void *arg);
45 extern "C" void *dir_heartbeat_thread(void *arg);
46 extern bool no_signals;
49 * Listen on the SD socket for heartbeat signals.
50 * Send heartbeats to the Director every HB_TIME
53 extern "C" void *sd_heartbeat_thread(void *arg)
56 JCR *jcr = (JCR *)arg;
58 time_t last_heartbeat = time(NULL);
61 pthread_detach(pthread_self());
63 /* Get our own local copy */
64 sd = dup_bsock(jcr->store_bsock);
65 dir = dup_bsock(jcr->dir_bsock);
68 jcr->hb_started = true;
69 jcr->hb_dir_bsock = dir;
71 /* Hang reading the socket to the SD, and every time we get
72 * a heartbeat or we get a wait timeout (1 minute), we
73 * check to see if we need to send a heartbeat to the
76 for ( ; !is_bnet_stop(sd); ) {
77 n = bnet_wait_data_intr(sd, WAIT_INTERVAL);
78 if (me->heartbeat_interval) {
80 if (now-last_heartbeat >= me->heartbeat_interval) {
81 bnet_sig(dir, BNET_HEARTBEAT);
85 if (n < 0 || is_bnet_stop(sd)) {
88 if (n == 1) { /* input waiting */
89 sd->recv(); /* read it -- probably heartbeat from sd */
90 if (sd->msglen <= 0) {
91 Dmsg1(100, "Got BNET_SIG %d from SD\n", sd->msglen);
93 Dmsg2(100, "Got %d bytes from SD. MSG=%s\n", sd->msglen, sd->msg);
96 Dmsg2(100, "wait_intr=%d stop=%d\n", n, is_bnet_stop(sd));
100 jcr->hb_bsock = NULL;
101 jcr->hb_started = false;
102 jcr->hb_dir_bsock = NULL;
106 /* Startup the heartbeat thread -- see above */
107 void start_heartbeat_monitor(JCR *jcr)
110 * If no signals are set, do not start the heartbeat because
111 * it gives a constant stream of TIMEOUT_SIGNAL signals that
112 * make debugging impossible.
115 jcr->hb_bsock = NULL;
116 jcr->hb_started = false;
117 jcr->hb_dir_bsock = NULL;
118 pthread_create(&jcr->heartbeat_id, NULL, sd_heartbeat_thread, (void *)jcr);
122 /* Terminate the heartbeat thread. Used for both SD and DIR */
123 void stop_heartbeat_monitor(JCR *jcr)
129 /* Wait max 10 secs for heartbeat thread to start */
130 while (!jcr->hb_started && cnt++ < 200) {
131 bmicrosleep(0, 50000); /* wait for start */
134 if (jcr->hb_started) {
135 jcr->hb_bsock->set_timed_out(); /* set timed_out to terminate read */
136 jcr->hb_bsock->set_terminated(); /* set to terminate read */
138 if (jcr->hb_dir_bsock) {
139 jcr->hb_dir_bsock->set_timed_out(); /* set timed_out to terminate read */
140 jcr->hb_dir_bsock->set_terminated(); /* set to terminate read */
142 Dmsg0(100, "Send kill to heartbeat id\n");
143 pthread_kill(jcr->heartbeat_id, TIMEOUT_SIGNAL); /* make heartbeat thread go away */
144 bmicrosleep(0, 50000);
146 /* Wait max 100 secs for heartbeat thread to stop */
147 while (jcr->hb_started && cnt++ < 200) {
148 pthread_kill(jcr->heartbeat_id, TIMEOUT_SIGNAL); /* make heartbeat thread go away */
149 bmicrosleep(0, 500000);
154 * Thread for sending heartbeats to the Director when there
155 * is no SD monitoring needed -- e.g. restore and verify Vol
156 * both do their own read() on the SD socket.
158 extern "C" void *dir_heartbeat_thread(void *arg)
160 JCR *jcr = (JCR *)arg;
162 time_t last_heartbeat = time(NULL);
164 pthread_detach(pthread_self());
166 /* Get our own local copy */
167 dir = dup_bsock(jcr->dir_bsock);
170 jcr->hb_started = true;
172 for ( ; !is_bnet_stop(dir); ) {
176 next = now - last_heartbeat;
177 if (next >= me->heartbeat_interval) {
178 dir->signal(BNET_HEARTBEAT);
179 last_heartbeat = now;
181 bmicrosleep(next, 0);
184 jcr->hb_bsock = NULL;
185 jcr->hb_started = false;
190 * Same as above but we don't listen to the SD
192 void start_dir_heartbeat(JCR *jcr)
194 if (me->heartbeat_interval) {
195 jcr->dir_bsock->set_locking();
196 pthread_create(&jcr->heartbeat_id, NULL, dir_heartbeat_thread, (void *)jcr);
200 void stop_dir_heartbeat(JCR *jcr)
202 if (me->heartbeat_interval) {
203 stop_heartbeat_monitor(jcr);