4 \section*{What To Do When Bacula Crashes (Kaboom)}
5 \label{_ChapterStart47}
6 \index[general]{Kaboom!What To Do When Bacula Crashes }
7 \index[general]{What To Do When Bacula Crashes (Kaboom) }
8 \addcontentsline{toc}{section}{What To Do When Bacula Crashes (Kaboom)}
10 If you are running on a Linux system, and you have a set of working
11 configuration files, it is very unlikely that {\bf Bacula} will crash. As with
12 all software, however, it is inevitable that someday, it may crash,
13 particularly if you are running on another operating system or using a new or
16 This chapter explains what you should do if one of the three {\bf Bacula}
17 daemons (Director, File, Storage) crashes.
19 \subsection*{Traceback}
20 \index[general]{Traceback }
21 \addcontentsline{toc}{subsection}{Traceback}
23 Each of the three Bacula daemons has a built-in exception handler which, in
24 case of an error, will attempt to produce a traceback. If successful the
25 traceback will be emailed to you.
27 For this to work, you need to ensure that a few things are setup correctly on
31 \item You must have an installed copy of {\bf gdb} (the GNU debugger), and it
32 must be on {\bf Bacula's} path. On some systems such as Solaris, {\bf
33 gdb} may be replaced by {\bf dbx}.
34 \item The Bacula installed script file {\bf btraceback} must be in the same
35 directory as the daemon which dies, and it must be marked as executable.
36 \item The script file {\bf btraceback.gdb} must have the correct path to it
37 specified in the {\bf btraceback} file.
38 \item You must have a {\bf mail} program which is on {\bf Bacula's} path.
39 By default, this {\bf mail} program is set to {\bf bsmtp}, so it must
40 be correctly configured.
43 If all the above conditions are met, the daemon that crashes will produce a
44 traceback report and email it to you. If the above conditions are not true,
45 you can either run the debugger by hand as described below, or you may be able
46 to correct the problems by editing the {\bf btraceback} file. I recommend not
47 spending too much time on trying to get the traceback to work as it can be
50 The changes that might be needed are to add a correct path to the {\bf gdb}
51 program, correct the path to the {\bf btraceback.gdb} file, change the {\bf
52 mail} program or its path, or change your email address. The key line in the
53 {\bf btraceback} file is:
57 gdb -quiet -batch -x /home/kern/bacula/bin/btraceback.gdb \
58 $1 $2 2>\&1 | bsmtp -s "Bacula traceback" your-address@xxx.com
62 Since each daemon has the same traceback code, a single btraceback file is
63 sufficient if you are running more than one daemon on a machine.
65 \subsection*{Testing The Traceback}
66 \index[general]{Traceback!Testing The }
67 \index[general]{Testing The Traceback }
68 \addcontentsline{toc}{subsection}{Testing The Traceback}
70 To "manually" test the traceback feature, you simply start {\bf Bacula} then
71 obtain the {\bf PID} of the main daemon thread (there are multiple threads).
72 Unfortunately, the output had to be split to fit on this page:
76 [kern@rufus kern]$ ps fax --columns 132 | grep bacula-dir
77 2103 ? S 0:00 /home/kern/bacula/k/src/dird/bacula-dir -c
78 /home/kern/bacula/k/src/dird/dird.conf
79 2104 ? S 0:00 \_ /home/kern/bacula/k/src/dird/bacula-dir -c
80 /home/kern/bacula/k/src/dird/dird.conf
81 2106 ? S 0:00 \_ /home/kern/bacula/k/src/dird/bacula-dir -c
82 /home/kern/bacula/k/src/dird/dird.conf
83 2105 ? S 0:00 \_ /home/kern/bacula/k/src/dird/bacula-dir -c
84 /home/kern/bacula/k/src/dird/dird.conf
88 which in this case is 2103. Then while Bacula is running, you call the program
89 giving it the path to the Bacula executable and the {\bf PID}. In this case,
94 ./btraceback /home/kern/bacula/k/src/dird 2103
98 It should produce an email showing you the current state of the daemon (in
99 this case the Director), and then exit leaving {\bf Bacula} running as if
100 nothing happened. If this is not the case, you will need to correct the
101 problem by modifying the {\bf btraceback} script.
103 Typical problems might be that {\bf gdb} is not on the default path. Fix this
104 by specifying the full path to it in the {\bf btraceback} file. Another common
105 problem is that the {\bf mail} program doesn't work or is not on the default
106 path. On some systems, it is preferable to use {\bf Mail} rather than {\bf
109 \subsection*{Getting A Traceback On Other Systems}
110 \index[general]{Getting A Traceback On Other Systems }
111 \index[general]{Systems!Getting A Traceback On Other }
112 \addcontentsline{toc}{subsection}{Getting A Traceback On Other Systems}
114 It should be possible to produce a similar traceback on systems other than
115 Linux, either using {\bf gdb} or some other debugger. Solaris with {\bf gdb}
116 loaded works quite fine. On other systems, you will need to modify the {\bf
117 btraceback} program to invoke the correct debugger, and possibly correct the
118 {\bf btraceback.gdb} script to have appropriate commands for your debugger. If
119 anyone succeeds in making this work with another debugger, please send us a
120 copy of what you modified.
121 \label{ManuallyDebugging}
123 \subsection*{Manually Running Bacula Under The Debugger}
124 \index[general]{Manually Running Bacula Under The Debugger }
125 \index[general]{Debugger!Manually Running Bacula Under The }
126 \addcontentsline{toc}{subsection}{Manually Running Bacula Under The Debugger}
128 If for some reason you cannot get the automatic traceback, or if you want to
129 interactively examine the variable contents after a crash, you can run Bacula
130 under the debugger. Assuming you want to run the Storage daemon under the
131 debugger (the technique is the same for the other daemons, only the name
132 changes), you would do the following:
135 \item Start the Director and the File daemon. If the Storage daemon also
136 starts, you will need to find its PID as shown above (ps fax | grep
137 bacula-sd) and kill it with a command like the following:
146 where you replace {\bf PID} by the actual value.
147 \item At this point, the Director and the File daemon should be running but
148 the Storage daemon should not.
149 \item cd to the directory containing the Storage daemon
150 \item Start the Storage daemon under the debugger:
159 \item Run the Storage daemon:
163 run -s -f -c ./bacula-sd.conf
168 You may replace the {\bf ./bacula-sd.conf} with the full path to the Storage
169 daemon's configuration file.
170 \item At this point, Bacula will be fully operational.
171 \item In another shell command window, start the Console program and do what
172 is necessary to cause Bacula to die.
173 \item When Bacula crashes, the {\bf gdb} shell window will become active and
174 {\bf gdb} will show you the error that occurred.
175 \item To get a general traceback of all threads, issue the following command:
185 After that you can issue any debugging command.
188 \subsection*{Getting Debug Output from Bacula}
189 \index[general]{Getting Debug Output from Bacula }
190 \addcontentsline{toc}{subsection}{Getting Debug Output from Bacula}
192 Each of the daemons normally has debug compiled into the program, but
193 disabled. There are two ways to enable the debug output. One is to add the
194 {\bf -d nnn} option on the command line when starting the debugger. The {\bf
195 nnn} is the debug level, and generally anything between 50 and 200 is
196 reasonable. The higher the number, the more output is produced. The output is
197 written to standard output.
199 The second way of getting debug output is to dynamically turn it on using the
200 Console using the {\bf setdebug} command. The full syntax of the command is:
204 setdebug level=nnn client=client-name storage=storage-name dir
208 If none of the options are given, the command will prompt you. You can
209 selectively turn on/off debugging in any or all the daemons (i.e. it is not
210 necessary to specify all the components of the above command).