Bug#519567: bacula doesn't use /etc/mailname to set the domain part of outgoing emails
bacula fails to use /etc/mailname when setting the sender address in outgoing status emails: It seems to me that the above suggestion is system dependent. bsmtp is a general mail sending program and attempts to use only Unix and Windows well defined OS calls. As such I don't think the above suggestion is appropriate. The bsmtp program is provided as an optional mail sending program, so you are free to use any other program. Unless someone can show me a system independent way to change this, from the upstream perspective, we don't plan any changes. Best regards, Kern Bacula project manager -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#544090: FTBFS on alpha: invalid pointer conversion
John,
I think the patch attached to this email (hopfully uploaded properly) will
fix the Alpha problem.
I've finally decided it was better to add detection code because various
machines use long unsigned int and others use int for the ioctl(fd,
req, ...) req argument. In addition, at least on my machine, the man pages
say it is int, but /usr/include/sys/ioctl.h says it is unsigned long
int ...
Kern
commit 78c64f1757419550411778b75f1018727d8f4e18
Author: Kern Sibbald k...@sibbald.com
Date: Sat Dec 26 14:50:29 2009 +0100
Attempt to autoconfig ioctl_req_t
diff --git a/bacula/autoconf/config.h.in b/bacula/autoconf/config.h.in
index 632f9c0..754a44d 100644
--- a/bacula/autoconf/config.h.in
+++ b/bacula/autoconf/config.h.in
@@ -452,6 +452,9 @@
declares uintmax_t. */
#undef HAVE_INTTYPES_H_WITH_UINTMAX
+/* Set if ioctl request is unsigned long int */
+#undef HAVE_IOCTL_ULINT_REQUEST
+
/* Whether to enable IPv6 support */
#undef HAVE_IPV6
@@ -617,6 +620,9 @@
/* Define to 1 if you have the `snprintf' function. */
#undef HAVE_SNPRINTF
+/* Set if socklen_t exists */
+#undef HAVE_SOCKLEN_T
+
/* Define to 1 if you have the stdarg.h header file. */
#undef HAVE_STDARG_H
diff --git a/bacula/autoconf/configure.in b/bacula/autoconf/configure.in
index 8a91a31..3147036 100644
--- a/bacula/autoconf/configure.in
+++ b/bacula/autoconf/configure.in
@@ -1859,7 +1859,31 @@ AC_CACHE_CHECK(for socklen_t, ba_cv_header_socklen_t,
)
]
)
-test $ba_cv_header_socklen_t = yes AC_DEFINE(HAVE_SOCKLEN_T)
+test $ba_cv_header_socklen_t = yes AC_DEFINE(HAVE_SOCKLEN_T, 1, [Set if socklen_t exists])
+
+dnl --
+dnl Check for ioctl request type
+dnl --
+AC_CACHE_CHECK(for ioctl_req_t, ba_cv_header_ioctl_req_t,
+ [
+ AC_TRY_COMPILE(
+ [
+ #include sys/types.h
+ #include sys/ioctl.h
+ ], [
+ unsigned long int req;
+ int fd;
+ ioctl(fd, req);
+ ], [
+ ba_cv_header_ioctl_req_t=yes
+ ], [
+ ba_cv_header_ioct_req_t_t=no
+ ]
+ )
+ ]
+)
+test $ba_cv_header_ioctl_req_t = yes AC_DEFINE(HAVE_IOCTL_ULINT_REQUEST, 1, [Set if ioctl request is unsigned long int])
+
dnl --
dnl Check for bigendian
diff --git a/bacula/configure b/bacula/configure
index b9c4dce..65d7dd1 100755
--- a/bacula/configure
+++ b/bacula/configure
@@ -32468,11 +32468,83 @@ rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
fi
{ echo $as_me:$LINENO: result: $ba_cv_header_socklen_t 5
echo ${ECHO_T}$ba_cv_header_socklen_t 6; }
-test $ba_cv_header_socklen_t = yes cat confdefs.h \_ACEOF
+test $ba_cv_header_socklen_t = yes
+cat confdefs.h \_ACEOF
#define HAVE_SOCKLEN_T 1
_ACEOF
+{ echo $as_me:$LINENO: checking for ioctl_req_t 5
+echo $ECHO_N checking for ioctl_req_t... $ECHO_C 6; }
+if test ${ba_cv_header_ioctl_req_t+set} = set; then
+ echo $ECHO_N (cached) $ECHO_C 6
+else
+
+ cat conftest.$ac_ext _ACEOF
+/* confdefs.h. */
+_ACEOF
+cat confdefs.h conftest.$ac_ext
+cat conftest.$ac_ext _ACEOF
+/* end confdefs.h. */
+
+ #include sys/types.h
+ #include sys/ioctl.h
+
+int
+main ()
+{
+
+ unsigned long int req;
+ int fd;
+ ioctl(fd, req);
+
+ ;
+ return 0;
+}
+_ACEOF
+rm -f conftest.$ac_objext
+if { (ac_try=$ac_compile
+case (($ac_try in
+ *\* | *\`* | *\\*) ac_try_echo=\$ac_try;;
+ *) ac_try_echo=$ac_try;;
+esac
+eval echo \\$as_me:$LINENO: $ac_try_echo\) 5
+ (eval $ac_compile) 2conftest.er1
+ ac_status=$?
+ grep -v '^ *+' conftest.er1 conftest.err
+ rm -f conftest.er1
+ cat conftest.err 5
+ echo $as_me:$LINENO: \$? = $ac_status 5
+ (exit $ac_status); } {
+ test -z $ac_c_werror_flag ||
+ test ! -s conftest.err
+ } test -s conftest.$ac_objext; then
+
+ ba_cv_header_ioctl_req_t=yes
+
+else
+ echo $as_me: failed program was: 5
+sed 's/^/| /' conftest.$ac_ext 5
+
+
+ ba_cv_header_ioct_req_t_t=no
+
+
+fi
+
+rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
+
+
+fi
+{ echo $as_me:$LINENO: result: $ba_cv_header_ioctl_req_t 5
+echo ${ECHO_T}$ba_cv_header_ioctl_req_t 6; }
+test $ba_cv_header_ioctl_req_t = yes
+cat confdefs.h \_ACEOF
+#define HAVE_IOCTL_ULINT_REQUEST 1
+_ACEOF
+
+
+
{ echo $as_me:$LINENO: checking for bigendian 5
echo $ECHO_N checking for bigendian... $ECHO_C 6; }
if test ${ba_cv_bigendian+set} = set; then
diff --git a/bacula/src/baconfig.h b/bacula/src/baconfig.h
index f725630..e93c98c 100644
--- a/bacula/src/baconfig.h
+++ b/bacula/src/baconfig.h
@@ -53,7 +53,11 @@
#define ETIME ETIMEDOUT
#endif
-#define ioctl_req_t long unsigned int
+#ifdef HAVE_IOCTL_ULINT_REQUEST
+#define ioctl_req_t unsigned long int
+#else
+#define ioctl_req_t int
+#endif
#ifdef PROTOTYPES
# define __PROTO(p) p
@@ -643,16 +647,9 @@ int m_msg(const char *file, int line
Bug#544090: FTBFS on alpha: invalid pointer conversion
Thank you for the most recent information. That and what John wrote made me realize that currently distinguish OS types, but in most of the code not machine architecture. In your case, it is apparently a different definition that depends on architecture. I think this is rather easy to fix, but I need one more confirmation in order to do it correctly. After doing your ./configure on the alpha, would you please copy the contents of bacula-source/src/host.h into the bug report. If it has #define DISTNAME alpha then I understand the problem and can create a patch. Regards, Kern -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#453940: bacula - default label format does not match documentation
The documentation was not correct in that it did not take into account the
fact that volumes may be deleted from the pool. In fact, what the code
attempts to do is to create unique volume names without reusing old ones.
I have updated the manual to say:
If no variable expansion characters are found in the string, the Volume
name will be formed from the {\bf format} string appended with the
a unique number that increases. If you do not remove volumes from the
pool, this number should be the number of volumes plus one, but this
is not guaranteed. The unique number will be edited as four
digits with leading zeros. For example, with a {\bf Label Format =
File-}, the first volumes will be named {\bf File-0001}, {\bf
File-0002}, ...
This will appear in version 2.4.3 and later versions.
A number of changes have been made to the algorithm in between your version
and the current 2.4.2 version. Note, the unique number is not necessarily the
MediaId as the MediaId pertains to all pools.
I recommend to close this bug report.
Kern
Bacula Project Manager.
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#475139: Inconsistencies should not cause a tape to be marked in error
Whenever Bacula marks a tape in error, it clearly prints that fact and why in the Job Report, so it is not quite accurate to say that it is impossible to distinguish between such a case and a case of a /real/ tape error.. Note, more recent versions of Bacula permit logging job reports to the database and easy access to that info using the GUI bat program. Just the same, having an additional state such as Inconsistent is a possibility. I will add it to our TODO list, but have not really decided whether or not to do it. In any case, it is a good deal of work to add a new state, including a major update of the database, so even if it is approved, don't count on it any time soon. If you want to ensure that this will become a project, please see: www.bacula.org - Feature Requests and follow the procedure outlined there. Kern Bacula Project Manager -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#461942: bacula-traymonitor: no manpage for bacula-tray-monitor
This is possibly a confusion due to the Debian package name being bacula-traymonitor, and the Bacula upstream named being bacula-tray-monitor. The upstream code does have a bacula-tray-monitor manpage that is installed during make install Kern Bacula Project Manager -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#442185: Debian Bug report #442185 bacula: configuration files in the Debian manner
In re-reading my response to your requests, I realize that I misunderstood. Sorry. What you are requesting is indeed possible. From the upstream point of view, the problem I have is that although, it is a very good way of organizing the conf files for larger organizations, it complicates the install/configuration for new users, and it does not correspond to the current documentation examples. I will consider adding documentation of this kind of a configuration setup in a future version and even possibly adding an alternate default configuration file that uses it (on TODO list). However until that is done, upstream, we cannot make this change. Of course, the John might consider it appropriate for the Debian package. Kern Bacula Project Manager -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#445867: Bug #445867 bacula-director-pgsql: crash apparently caused by bad config file
The crash is not a bug but the only system independent way I have found to cause Bacula to dump when there is some error. In this case, it is a user syntax error that seems to be reasonably clear from the error messages. Once the error messages are printed, Bacula induces a Seg Fault. This bug report should be closed. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#435627: Debian Bug report#435627 bacula-sd: change suggestion for mtx-changer script
This is corrected in version 2.2.6 which will be released in the next few weeks. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#441621: Debian Bug report #441621 All bacula daemons do not release STDOUT/STDERR on daemonization
Fixed upstream. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#442884: Debian Bug report #442884 Default MySQL table format contains a redundant index
Fixed upstream in 2.2.6. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#391949: Debian Bug report #391949 Please have text console ignore ConsoleFont, not error out
I'm not planning to implement this request since ConsoleFont is not a valid directive for bconsole. If you don't want error messages, please use two different conf files. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#440276: Debian Bug report #440276 bconsole errors when monitoring systems connect to bacula ports
Those messages are included so that users can detect security violations. It is a daemon message so if you want you can direct it anywhere you want, but I have no intention of removing it. At some time I may move them into a security message class so that they can be bit bucketed if you are not worried about security. You should definitely prevent non-bacula or at least non-bacula compatible programs from probing its ports. No change planned in the near future. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#442185: Debian Bug report #442185 bacula: configuration files in the Debian manner
I don't particularly like programs that include every file in a subdirectory, so it is not likely that I'll be implementing this personally, because Bacula already has an include method using @. However, if you really want this, I suggest that someone submit a patch. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#446809: Debian Bug report logs - #446809 CVE-2007-5626 unauthorized disclosure of information via clear-text passwords used in command line arguments
There is no simple solution that works in all cases that I am aware of. This potential security problem (no worse than configuring your conf file with world read permission) has been documented for quite a long time in the manual. We've added additional documentation to the scripts in question warning the user of security problems. No other changes planned. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#386920: bacula-director-common: do not require exim4 or mail-transport-agent pkgs, just suggest instead.
This appears to be a packaging problem because Bacula by default uses only mail or its own bsmtp program for emailing Job reports. The only thing necessary to have Bacula work out of the box is a valid user id and an smtp address. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#396913: Schedule WeeklyCycle should be MonthlyCycle
Technically you are correct if you look from the stand point of Full backups. However, it has been called that a long time and no one really seems to be confused. I'll accept a patch for it on the bacula-devel list, but I don't think it is worth tweaking the files, and *all* the documentation that would need to be changed. I've taken note of this, but promise nothing. I recommend closing this bug report. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#396921: bacula-director-pgsql depends on sqlite3?!
I agree, the PostgreSQL version of Bacula should not need SQLite3. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#391966: Control-Z in text console hangs console, not suspends
The Bacula console does not support ctl-z for suspending it. This is not a bug, but a feature, and implementation of suspending is highly unlikely unless some user contributes an appropriate patch. If you want to submit a patch, please submit it to the bacula-devel list. If you wish to submit a feature request, then please see www.bacula.org and click on the Feature Request menu item. I recommend closing this bug report. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#383357: bacula-director-pgsql: make_catalog_backup has hardcoded path to wrong version of pg_dump
This is very likely some packaging problem somewhere because the Bacula configure script automatically updates the make_catalog_backup script to have the correct path to the appropriate database dump program. The full path is not put on an environment variable, so it may seem that it is hard coded, but that is not the case. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#371890: lower priority for mtx
Bacula by default installs a script that uses mtx. The problem for a packager is that there is no easy way to know if the user really has an autochanger or not. Most packages including all the rpms require mtx by default to avoid any problem the user might have later. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#391492: Fails to init supplemental groups
There are two possible issues here. One is a packaging issue. If you package Bacula to run as non-root, which can be done for both the Director and the Storage daemon, then the packager must ensure that any new userid or group is created when installing the package. For example, often the Director runs as bacula:bacula, and the SD as bacula:disk. The second issue here is Bacula itself. It must be started as root, then drops to the user/group specified on the command line. Due to the excessively complex Unix API for dropping to a different group, the code in version 1.38.11 did not get it completely correct. This is fixed in 1.39.x. The result is that it is often easier to run the Storage daemon in group root otherwise you can run into permission problems attempting to access tape devices. In any case, if Bacula needs access to certificate files, then you *must* either run the appropriate component as root or modify the permissions on the certificate directories and files so that Bacula can access them. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#388027: Auto reconnect to DB when connection dropped?
It seems to me that there are a few issues here: 1. If you shutdown or upgrade a database while Bacula is running a Job, then Bacula will fail. The solution is: don't do that. Bacula has no explicit code to reconnect to a database after the database is successfully opened, and we don't plan to add any such code as it would render the data less secure. 2. It is possible that the network (if that is what you setup) connection between Bacula and the database server could be broken. In that case, since it is the database's client libraries that established the connection, it is their responsibility to manage the connection and reconnect if it is possible. If the database API supplies a call or flag that requests the client libraries to reconnect, we will (and in the case of MySQL where this exists do) use the API/flag. However, this still leaves it to the client libraries to manage the interface. 3. With most databases, you can set them up to use sockets rather than TCP/IP providing the database and the Director run on the same machine. This might reduce any disconnects that would occur via TCP/IP. 4. Bacula has a connection to the database open only when it is actually running a job, so if this error occurs, and the client library cannot reconnect by itself, Bacula will take the conservative point of view and fail the job. 5. There was one case of MySQL recently releasing a new version where they changed the API/flag that requests reconnect. This unfortunately resulted in a few dropped connections until we modified the Bacula code to correspond to the new MySQL coding. This should no longer be an issue if you are running any version of 1.38.11. 6. Some of the test results reported indicate that there may be a bug with specific versions of MySQL. The manual has pointers to how to resolve these disconnect issues for MySQL where they seem more common (I have never had one ...). Providing there are no Debian specific modifications to the cats directory source code, I recommend closing this bug report as there is no planned change to Bacula and no known bug in version 1.38.11 or greater. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#371883: bacula-director-pgsql: upgrade from 1.36 fails
I haven't read all the details of this bug report, but it seems to me that the basic problem is that normally in major Bacula upgrade, e.g. 1.36.x - 1.38, Bacula has new table formats. This has nothing to do with upgrading the SQL engine, but is due to the fact that the Bacula SQL table format has changed. Updating the table require running a special update script that is supplied with the new release of Bacula. If you do not run the script, you will get an error such as: bacula-dir: Fatal error: Version error for database bacula. Wanted 9, got 8 as reported in this bug. This indicates that Bacula was upgraded, but Bacula's tables were not upgraded. The upgrade can either be done manually by invoking the appropriate update script, or it can be automatically done when installing the new package (the rpms do an automatic update for the user). The procedure (what to do, where to find the script) is documented in the ReleaseNotes file. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#376679: bacula: Upgrade to 1.38.11 broke autochanger config.
I've taken another look at this bug report, and am convinced that my comment of 9 Oct remains valid. The user did not upgrade his Storage daemon configuration to include the new Autochanger resource, and thus Bacula fails to function correctly. I recommend that this bug be closed and marked as not a bug or user configuration error. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#379605: in file-selection mode bacula-console-wx issues .dir instead of dir
wx-console had a number of bugs which were fixed, and unfortunately some of those required making non-downward compatible changes, which is inevitable. The .dir is in fact correct, but for it to work correctly, you must use the same Director version (1.38.x) and not an older version. You will be better off running everything on version 1.38.11. I recommend closing this as a not a bug since we never can guarantee proper functioning of all components between different major versions (in general all 1.38.x components are compatible, but not 1.36.x to 1.38.x). Regards, Kern -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#387838: bacula-director-mysql: segfault
The Bacula manual recommends that you test your conf files with ./bacula-xx -t ... This will clearly indicate what is wrong and where (in 99.9% of the cases). If you run Bacula as a daemon, due to the flexibility in Bacula error message handling, it is very hard to get the user's attention, so Bacula attempts to produce a traceback, which for some reason in your case seems to be configured incorrectly. This also would have indicated what the problem is. Finally, yes Bacula does use SEGFAULT to generate tracebacks when it aborts, because it is the only portable way of doing so (the Unix abort() is not portable). This is not a bug and should be closed. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#376679: bacula: Upgrade to 1.38.11 broke autochanger config.
It is hard to be sure, but most likely you did not upgrade your bacula-sd.conf file as is required for autochangers in switching from 1.38.11. As very clearly indicated in the Release Notes and the manual, Bacula 1.38.11 requires a new Autochanger resource in order to function correctly with Autochangers. If you didn't add an Autochanger resource, that is most surely your problem. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#380220: bacula-director-mysql: Start fails with *** glibc detected *** free(): invalid pointer: 0x080ea648
This kind of problem is almost always due to the fact that Bacula was linked against one version of mysql (e.g. 4.x) and you are using a different version (e.g. 5.x). I'm not a Debian user, so I cannot give the best advice, but in general, either check what was linked in the original package you loaded, or rebuild Bacula on your current system linking against your current mysql libraries. Another similar possiblity, but probably pretty unlikely on Debian, is that you have several different mysqls loaded on your system, possibly in different locations, as often happens when upgrading from 4.x to 5.x. Bacula may be picking up or using header files from an old installation but using the shared objects from a newer installation. Regards, Kern -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#387838: bacula-director-mysql: segfault
On Monday 09 October 2006 18:15, John Goerzen wrote: On Mon, Oct 09, 2006 at 05:57:00PM +0200, Kern Sibbald wrote: Finally, yes Bacula does use SEGFAULT to generate tracebacks when it aborts, because it is the only portable way of doing so (the Unix abort() is not portable). This is not a bug and should be closed. Thank you for the clear explanation. I have closed the report. Just to clarify for the submitter: Kern is saying that this is not a bug unless you also get segfaults in the test mode. If you try it and do get segfaults there, let me know, and we'll pursue it further. Yes, thanks for the clarification -- you are correct. The Bacula conf file scanner has never been what I would call super good, but it should not seg fault on a valid conf file or it is a bug. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#383332: bacula: backup is slow
From the statistics you show, the backup does not appear slow to me. The reason you might think it is slow is because you are comparing apples and oranges. On the one hand, you measure the time to to a non-compressed tar on a local machine sending the output down an extremely hi-speed bit bucket. On the other hand, you measure the time of Bacula using compression sending real data to another process via TCP/IP (even though it might be on the same machine). To do a better comparison, you could run tar including the z option so that it does compression. In addition, you should send the output of tar across the network and write it to either a file or a tape (whatever Bacula is using). At that point, providing you are always doing Full backups on Bacula (and not Incremental of Differential), you will probably find that the total Bacula time is not terribly greater than tar. That said, Bacula will amost always be slower than tar because it does a whole lot more -- in addition to checksuming all the data Bacula writes to the Volume, which I am not sure tar does, Bacula also interfaces to a database and stores a lot of information about the job. If you want to do additional performance testing you can look at bacula-source/src/version.h. There are various configuration parameters that you can turn on/off and then re-build Bacula and measure the performance of particular parts. Performance testing is a highly evolved science as well as an art, and it is not always easy. For example, if you are going to do any timing experiments as you did, you *must* on Unix systems re-run the test at least 10 times, throw out the first two timings, then take an average of the remaining 8. If you don't do this, your timing tests will have no meaning due to the memory cacheing that Unix does. Best regards, Kern
Bug#374427: bacula-director-pgsql: Does not work with ident sameuser
Hello, This is indeed a problem, and not one that I consider an upstream bug. The basic problem is that Bacula cecks if it can open the database prior to becoming a daemon and thus prior to dropping root privilege. It is the only way, it can properly print the message and exit. When it attempts to connect to the database the first time, Bacula will be running as root, but it will use the correct database name and the correct user id and password. If you have set the Bacula database to *require* that the user id specified corresponds to the Unix user id that is running, then you will need to configure the Bacula database to either accept root or bacula logins. Otherwise, a simpler method is to allow any user to log into the bacula database providing he has the userid and password for the database. This makes more sense because there are a lot of applications such as bacula-web that need to be able to read/write the Bacula database. I don't consider this an upstream bug, but rather a problem of administration of database access to be resolved at installation time. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#360530: bacula-common: tcpwrapper config not documented
Hello, Yes, the tcpwrappers name that the daemons look for in libwrap is not the daemon name. This is because there can be multiple simultaneous Bacula daemons running on the same machine. They simply use different .conf files. The fact that you must use the daemon Name as defined in the .conf file has long been documented in several places in the manual (the --with-tcpwrapper section and the Security chapter). The security chapter goes into detail on how to setup and test tcpwrappers. Since the manual is some 500 pages, the information is not always easy to find, but it is now indexed. I don't think it appropriate to add this level of detail to the man pages. I recommend closing this bug report. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#342111: bacula-sd: Bad autochanger loaded drive 0 command: ERR=Child exited with code 1
Hello, This kind of problem happens frequently if the SD is not being run as root and there are not sufficient permissions for the SD to access the script or the directory in which the script resides. Make sure that the user (don't count on the group as changing the group on the command line does not function correctly) that the SD is running under can actually access the directory and the script. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#367643: man page for dbcheck is wrong
Hello, I have recently integrated all the man pages into the upstream Bacula source code, and I noticed this problem and corrected it. I've also added more substance to some of the man pages and hope to get through them all in the next few months. This correction will appear upstream in version 1.40.0 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#363273: bacula should be able to schedule commands in advance
Hello, Bacula is fully multiple threaded, which means you can issue Bacula commands at any time. Each command must complete (e.g. database listing, ...) prior to issuing the next command, and in some cases, if you attempt to use a tape drive to label a tape and the drive is busy, the command will fail (i.e. return an error). If you want such an item as you are suggestion (and I don't really think it is needed), please submit a Feature Request to the bacula-users email list. Note, the precise details of a Feature Request are described on the www.bacula.org web site. I recommend closing this bug. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#369541: bacula: split parts from bacula-dir.conf
Hello, I think you have some good ideas here that a number of users would want. The best way to get something like this implemented is to do it yourself (after discussing it on the bacula-users list), then submit a patch against the current CVS to the bacula-devel list. If you wish to submit this as a Feature Request, the best procedure is to follow the Feature request procedures as outlined on the www.bacula.org web site then submit it to the bacula-users list and either defend your ideas or wrap in their suggestions. When there seems to be a consensus (more or less) I will accept the Feature request and add it to our projects list. I recommend closing this bug report and submitting a Bacula feature request. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#352287: /usr/sbin/btape: btape seg faults when handling a device error
Hello, The programs such as btape that need a drive to work ABORT when it cannot find/open the drive. The only portable way of ensuring the program ABORTs is to generate a seg fault, so the behavior you have noted is a bit unusual but normal. I recommend closing this bug report. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#252305: bacula-console-gnome: Looks in wrong place for configuration file Date:
Hello, No configuration paths of any kind are programmed into the binaries. Thus they look where you have specified on the command line for their configuration file. If you don't specify one, they look in the current directory, and then die if one is not found. If you want a simple one line command that runs the console, simply create a shell script that adds the proper path to the configuration file, or create an appropriate icon on your desktop. No change upstream is necessary. I recommend closing this bug report. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#314301: bacula-doc missing section on FileSet
Hello,
There was indeed a problem in producing the document because under certain
circumstances LaTeX does not permit includes within includes. This caused
the missing section.
This has been *long* ago corrected. If someone would like to correct this,
edit docs/manual/dirdconf.tex and change the line that reads:
\include{fileset}
to
\input{fileset}
then rebuild the document.
I recommend closing the bug.
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#367424: [Bacula-devel] FW: Bug#367424: bacula-sd: Debian build with debug symbols
Hello,
The debug listing below confirms that the problem is coming from the
foreach... loop that is examining the Devices attached to the Autochanger.
It is impossible for the pointer to be NULL, which is apparently what is
happening, which indicates to me a g++ bug.
Could you tell me what version of g++ you are using?
Possible solutions:
1. Verify that #define HAVE_GCC 1 is set in src/config.h
2. add -O0 (- oh zero) or rather replace the -02 compile option with -O0.
3. add -fno-strict-aliasing to the compile options, in which case the -O2
should be able to be used.
Concerning the impact of -O0. Well, obviously if -fno-strict-aliasing -O2
works, it will be more efficient. However, between having the SD crash
and running with no optimization, I personally would quickly choose -O0
:-)
All this is quite frustrating to me, because in more than 5 years, I have
never seen a compiler problem (bug), and in the last 6 months, there have
been a rash of these kinds of strict aliasing problems that arise
because of optimization or over optimization. Every case I have seen has
been with a 4.x compiler ...
- Forwarded message from Marcus Meyer [EMAIL PROTECTED] -
From: Marcus Meyer [EMAIL PROTECTED]
Date: Tue, 23 May 2006 15:57:27 +0200
Reply-To: Marcus Meyer [EMAIL PROTECTED], [EMAIL PROTECTED]
To: Debian Bug Tracking System [EMAIL PROTECTED]
Subject: Bug#367424: bacula-sd: Debian build with debug symbols
Package: bacula-sd
Version: 1.38.9-9
Followup-For: Bug #367424
Hi!
Now I've compiled the debian source with debug symbols.
Here is, wat i did:
apt-get source bacula
cd bacula-1.38.9/
DEB_BUILD_OPTIONS=noopt nostrip dpkg-buildpackage -rfakeroot -us -uc -B
After the build I copy the bacula-sd from
bacula-1.38.9/debian/tmp-build-mysql/src/stored/bacula-sd to
/usr/sbin/bacula-sd then I started the daemon with /etc/init.d/bacula-sd
start.
The error was:
Starting Bacula Storage daemon: 23-Mai 15:45 bacula-sd: Fatal Error
because: Bacula interrupted by signal 11: Segmentation violation
Kaboom! bacula-sd, bacula-sd got signal 11. Attempting traceback.
Kaboom! exepath=/usr/sbin/
Calling: /usr/sbin/btraceback /usr/sbin/bacula-sd 6989
Traceback complete, attempting cleanup ...
And the traceback:
Using host libthread_db library /lib/tls/libthread_db.so.1.
[Thread debugging using libthread_db enabled]
[New Thread -1213532480 (LWP 6989)]
0xb7ea71fe in __waitpid_nocancel () from /lib/tls/libpthread.so.0
$1 = bacula-sd, '\0' repeats 20 times
$2 = 0x80b4870 bacula-sd
$3 = 0x80b4898 /usr/sbin/bacula-sd
$4 = MySQL
$5 = 0x80a57f8 1.38.9 (02 May 2006)
$6 = 0x80a270d i486-pc-linux-gnu
$7 = 0x80a2706 debian
$8 = 0x80a26f5 testing/unstable
#0 0xb7ea71fe in __waitpid_nocancel () from /lib/tls/libpthread.so.0
#1 0x080900c3 in signal_handler (sig=11) at signal.c:159
#2 signal handler called
#3 0x0804e70a in init_autochangers () at autochanger.c:48
#4 0x0804c8b0 in main (argc=value optimized out, argv=value optimized
out) at stored.c:401
Thread 1 (Thread -1213532480 (LWP 6989)):
#0 0xb7ea71fe in __waitpid_nocancel () from /lib/tls/libpthread.so.0
#1 0x080900c3 in signal_handler (sig=11) at signal.c:159
#2 signal handler called
#3 0x0804e70a in init_autochangers () at autochanger.c:48
#4 0x0804c8b0 in main (argc=value optimized out, argv=value optimized
out) at stored.c:401
#0 0xb7ea71fe in __waitpid_nocancel () from /lib/tls/libpthread.so.0
#0 0xb7ea71fe in __waitpid_nocancel () from /lib/tls/libpthread.so.0
No symbol table info available.
#1 0x080900c3 in signal_handler (sig=11) at signal.c:159
159 waitpid(pid, NULL, 0); /* wait for child to produce
dump */
Current language: auto; currently c++
be = {SMARTALLOC = {No data fields}, buf_ = 0x99816a54 Address
0x99816a54 out of bounds, berrno_ = 65522866}
sigdefault = {__sigaction_handler = {sa_handler = 0, sa_sigaction = 0},
sa_mask = {__val = {2147483647, 4294967294, 4294967295 repeats 30
times}},
sa_flags = 0, sa_restorer = 0xc5a58668}
argv = {0x0, 0x0, 0x0, 0x0}
pid_buf = 6989, '\0' repeats 15 times
buf =
/tmp/\000Qi\202\234·Óð\016-M^p\203\227¬ÂÙñ\n$?[x\226µÕæø\v\0374Jay\222¬Çã\000\036=]n\200\223§?Òé\001\0324Ok\210?Ååö\b\033/DZq\211¢?×ó\020.Mm~\220£·Ìâù\021*D_{\230¶Õõ\006\030+?
pid = 6990
btpath = /usr/sbin/btraceback, '\0' repeats 379 times
#2 signal handler called
No symbol table info available.
#3 0x0804e70a in init_autochangers () at autochanger.c:48
48 if (!device-changer_command changer-changer_command)
{
device = (DEVRES *) 0x0
OK = true
changer = (AUTOCHANGER *) 0x80b6938
#4 0x0804c8b0 in main (argc=value optimized out, argv=value optimized
out) at stored.c:401
401OK = init_autochangers();
ch = value optimized out
no_signals = 0
test_config = 0
thid = 0
uid = 0xbfdf9f0e bacula
gid = 0xbfdf9f18 tape
#0 0x in ?? ()
No symbol table info available.
#0 0x in ?? ()
No symbol table info
Bug#296648: [Apcupsd-users] debian bug request
On Friday 25 February 2005 11:01, Samuele Giovanni Tonon wrote: hi, A bug request was filled against debian: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=296648 i think that this is most a feature than a bug, however, what the user pointed out (that MAX_SLAVE is hardcoded instead of staying in conf file) is quite interesting. Yes, a limit of 20 slaves is complied into apcupsd. While this is not the way I would have coded it, one must remember that such practices were prevalent at the time apcupsd was written and are quite common today. A limit of 20 is actually reasonable since having more handling more than 20 slaves increases significantly the possibility that the master does not have the time to notify all the slaves before shutdown, particularly if there are network problems due to a router/hub being down due to a power failure. The recommended alternative is to use the NIS networking rather than the Master/Slave networking. This is documented in the manual. There is no limit to the number of NIS slaves nor the same problem of notification. If the user would like to submit a patch, and it is written in the current apcupsd style, the apcupsd project will be happy to accept it. Otherwise, this is not an item high on our priority list. I recommend closing the bug report. Best regards, Kern -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#295823: [Bacula-users] Bug#295823: bacula-director-mysql: Unknown table 'delCandidates'
Hello, Thanks for the bug report. In the future, it would be helpful to send them to [EMAIL PROTECTED] instead of the bacula-users list. This is actually a bug in MySQL, or at least the version you are using since, unless I am mistaken, the SQL standard requires table names to be case independent. Just the same, I have corrected the Bacula code in both the 1.36.2 stream (to be released shortly) and the 1.37 development stream. Thanks for the patch. On Fri, 2005-02-18 at 12:29 +0100, Lupe Christoph wrote: Package: bacula-director-mysql Version: 1.36.1-1 Severity: normal Tags: patch Subject: bacula-director-mysql: Unknown table 'delCandidates' Package: bacula-director-mysql Version: 1.36.1-1 Severity: normal A restore job just printed 18-Feb 11:51 antalya-dir: Begin pruning Jobs. 18-Feb 11:51 antalya-dir: Query failed: SELECT DISTINCT DelCandidates.JobId FROM Job,DelCandidates WHERE (Job.JobTdate1098355884 AND delCandidates.JobStatus!='T') OR (Job.JobTDate1098355884 AND Job.ClientId=1 AND Job.Type='R'): ERR=Unknown table 'delCandidates' in where clause 18-Feb 11:51 antalya-dir: Pruned 0 Jobs for client antalya-fd from catalog. 18-Feb 11:51 antalya-dir: Begin pruning Files. 18-Feb 11:51 antalya-dir: No Files found to prune. 18-Feb 11:51 antalya-dir: End auto prune. There is a typo in src/dird/sql_cmds.c (still there in CVS, file version 1.48). Patch attached. -- System Information: Debian Release: 3.1 APT prefers testing APT policy: (990, 'testing'), (100, 'unstable') Architecture: i386 (i686) Kernel: Linux 2.4.27-lpc.7.k7 Locale: LANG=en_US, LC_CTYPE=en_US (charmap=ISO-8859-1) Versions of packages bacula-director-mysql depends on: ii bacula-common 1.36.1-1 Network backup, recovery and verif ii bacula-director-common 1.36.1-1 Network backup, recovery and verif ii debconf 1.4.30.11Debian configuration management sy ii libc6 2.3.2.ds1-20 GNU C Library: Shared libraries an ii libgcc1 1:3.4.3-6GCC support library ii libmysqlclient124.0.23-4 mysql database client library ii libssl0.9.7 0.9.7e-2 SSL shared libraries ii libstdc++5 1:3.3.5-8The GNU Standard C++ Library v3 ii libwrap07.6.dbs-6Wietse Venema's TCP wrappers libra ii mysql-client4.0.23-4 mysql database client binaries ii zlib1g 1:1.2.2-3compression library - runtime -- debconf information: * bacula-director-mysql/remove_catalog_on_purge: false * bacula-director-mysql/db_host: localhost * bacula-director-mysql/create_tables: true * bacula-director-mysql/mysql_root_username: lupe * bacula-director-mysql/db_user: bacula -- System Information: Debian Release: 3.1 APT prefers testing APT policy: (990, 'testing'), (100, 'unstable') Architecture: i386 (i686) Kernel: Linux 2.4.27-lpc.7.k7 Locale: LANG=en_US, LC_CTYPE=en_US (charmap=ISO-8859-1) Versions of packages bacula-director-mysql depends on: ii bacula-common 1.36.1-1 Network backup, recovery and verif ii bacula-director-common 1.36.1-1 Network backup, recovery and verif ii debconf 1.4.30.11Debian configuration management sy ii libc6 2.3.2.ds1-20 GNU C Library: Shared libraries an ii libgcc1 1:3.4.3-6GCC support library ii libmysqlclient124.0.23-4 mysql database client library ii libssl0.9.7 0.9.7e-2 SSL shared libraries ii libstdc++5 1:3.3.5-8The GNU Standard C++ Library v3 ii libwrap07.6.dbs-6Wietse Venema's TCP wrappers libra ii mysql-client4.0.23-4 mysql database client binaries ii zlib1g 1:1.2.2-3compression library - runtime -- debconf information: * bacula-director-mysql/remove_catalog_on_purge: false * bacula-director-mysql/db_host: localhost * bacula-director-mysql/create_tables: true * bacula-director-mysql/mysql_root_username: lupe * bacula-director-mysql/db_user: bacula -- Best regards, Kern -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#274752: [Apcupsd-users] Re: apcupsd debian bugs and feature requested
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=285316 Docbook is no longer used for apcupsd. The directory will remain for a few releases for historical reference. The manual has now been converted to texinfo format, so it is found in apcupsd/doc/texi. To build it simply cd to that directory and type make. If you have the texinfo tools loaded (including TeX), you will get the manual -- see the Makefile for details. Note, TeX generates a good number of warning messages, which you can ignore. Please note that I am very unsatisfied with the texinfo manual, so will be converting it to LaTeX in the next several months. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#87728: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=87728
The problem here is that the only person who understands the po stuff left the project. Since you do not seem to be too sure about the problem (I think it's not necessary to make this distinction), and I don't know anything about this stuff the best tactic is to do nothing. Now, if someone understands how gettext works and can take the time to explain it to me ([EMAIL PROTECTED]), I'll be happy to try to clean up these annoying points. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#286612: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=286612
I may be missing something, but I don't see the need for this feature. You can do what you want today by simply modifying apccontrol to do your action script rather than complicating apcupsd. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#294256: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=294256
If the submitter of this bug report would like us to include his idea in the apcupsd release, we will be glad to do so. For that to happen, we will need a small document describing it and how to use it as well as the full script. I'm not sure we will integrate it into the mainstream code (I reserve judgment on this ...), but at a minimum we can include the script and the document in our examples directory. Send the full script and document to [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#286613: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=286613
This is a good idea. I've added it to the things to do for the next or subsequent release. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#287739: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=287739
I believe that we have given you the tools necessary to do what you want by using apcupsd.conf directives and event scripts. This is discussed in the release notes in a recent release -- if not the current release then look in doc/techlogs. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
