1 Technical notes on version 1.39
6 kes Make bscan ignore ACL streams.
7 kes Update projects files with new Feature Requests.
8 kes Generate current English and French manuals and upload
10 kes Update the home page to include the new British/French flags to
11 switch between translations. Work done by Alexandre Baron.
12 kes Make disk-changer script print an error message if the changer
13 directory is not defined.
14 kes Create a generic DB routine for getting an int (32/64 bit) from
16 kes Add more migration code.
17 kes Add a few more files to .cvsignore here and there.
18 kes Minor fixes for SD write rate and despooling rate.
19 kes Turn off new test code in src/console.c
23 - Incremented the version number to signify that all of Robert's code
24 for Win32 that was in the branch is now integrated into the HEAD,
25 and that Eric has submitted code that we hope will provide limited
26 backward compatibility with 1.38 file daemons (more testing to be
28 - I rationalized a few of the names of the variables (mainly jcr names)
29 that are used during migration as keeping three different jcrs at the
30 same time clear in one's head is not easy.
34 - Tweak new runscript feature to work with 1.38.x FDs.
35 - Correct crypto.c to fix segfault when restoring an encrypted stream
36 on fd which don't use rsa keys.
37 - Add yes/no english keywords on localised yes/no questions.
39 - Several important commits from Robert Nelson for code cleanup and
41 - Modify console.c so that when conio is enabled, it converts \n to
42 \n\r before outputting to the console. Hopefully this will fix
43 the problems with expect in Perl that Eric saw.
45 - Change cats/sql.c to elimate %-*s format, which I think is turned
47 - Add Comment field to llist of a volume.
48 - Allow true/false in some yes/no questions.
49 - Implement update volume enable=(on|off|true|false|archived|0|1|2)
50 - Remove a few /n's when scanning commands from another daemon. This
51 permits adding additional fields later.
52 - Modify dbcheck to handle ctl-d and to delete Log table entries.
55 - Tweak authentication to work with 1.38.x FDs.
56 - Tweak catalog make scripts.
57 - Fix catalog upgrade scripts so that they work.
58 - Correct despooling debug code to eliminate race condition.
60 - Add additional fields as specified by Arno to LocationLog.
61 - Add comment field to the Media record.
62 - Add Time field to the Log record.
63 - Correct migration SQL (thanks to Bill Moran) so that it
64 runs with PostgreSQL also.
65 - Add spooling/despooling info in status output of SD.
67 - Spend a lot of time integrating mkcdrec with the rescue disk.
68 - Add VOLMGMT message class for volume management messages (none yet).
69 - Add CATALOG as a destination. It goes into the LOG table.
70 - Implement the Log table in the DB.
71 - Implement the Location Log table in the DB (for user use).
72 - At Eric's request add a run_scripts() just after blast_data in
75 - Add Enabled column to Location, correct some typos in DB schemas.
76 - Correct bug I introduced into RunScripts enum.
77 - Add log table to Catalog
78 - Purge original job migrated if the migration terminates normally.
79 - Cleanup purging of files/jobs so that it is done only in a
81 - Add new VOLMGMT message class.
82 - Add Catalog message destination (user conf code remains to be written).
84 - Modify read_record and friends to properly deal with migration
86 - Update debug code in read_record.
88 ======================= Warning ==========================
89 Separate read and write storage in Jobs in the Director. This
90 is a major rewrite of some of the low level code and hence has
91 a high probability of introducing bugs.
92 ==========================================================
94 - Move the job init code up to *before* the job is put into
95 the job queue (resource allocation queue). This permits the
96 job to establish whether it will read or write or do both.
97 - Add name() methods to all the resource classes.
98 - Implement separate read and write storage resource
99 allocation/deallocation in jobq.c
101 ======================= Warning ==========================
102 Implemented new method of getting the default Storage from the Media
103 record rather than from the MediaType for restore. As a fall
104 back, if no Storage is defined, use the MediaType. This
105 eliminates most all prompts if there are multiple MediaTypes
106 used. This is an important and fundamental change.
107 ==========================================================
109 - Add more detail (Storage, Device) to list of volumes printed
111 - More code in migrate. In particular start multiple jobs if
113 - Probably overkill, but ensure that the bsr is unique even within a
115 - Rework bsys.c drop() so that it now should work as expected.
116 - Update a bunch of old copyrights -- particularly in .h files.
117 - Remove src/lib/bshm.c/h
121 ======================= Warning ==========================
122 All hash codes in the database are now kept in world
123 compatible base64 format (no = filling). This means that
124 all FileSets will be updated and if you are running
125 verify jobs, you must do an InitCatalog.
126 Authentication uses the new algorithm, but should be
127 backward compatible with 1.38.x
128 ===========================================================
130 - Modify the authentication routines so that the 1.39.15 Director
131 always runs with the compatible binary to base64 code (same as
132 the rest of the world). It will however detect if an SD or
133 FD is running the old code and if that is the case, will use
134 the old algorithm for authentication. However, all consoles must
135 use the compatible algorithm, and all data entered into the database
136 (all hash signatures, MD5, SHA1, ..., and all FileSet hashes) will
137 use the compatible algorithm.
138 - Rename the cram_md5 routine names to be slightly more appropriate
139 to what they are doing (challenge and respond).
140 - Robert committed his #ifdef and header cleanup.
142 - Fix a complier warning in files/backup.c
143 - Cleanup NOT NULL vs DEFALT 0 in database creation.
145 - Robert committed his low impact changes.
146 - I removed a size_t from the jcr.h definition.
147 - Update the tape regression scripts (still to be tested).
148 - Add gcc and gcc-c++ requires to srm as suggested by a user.
149 - Require done or quit to exit from in memory tree restore.
151 - Write install/uninstall of new manpages. There are now a
152 pile of man pages. Some of which need a lot of work.
154 ======================= Warning ==========================
155 The --mandir ./configure option now points to the top level
156 man directory. The man files will be installed under
157 mandir/man8 and mandir/man1 as appropriate.
158 ==========================================================
160 - Add complete text of GPL to LICENSE.
161 - Implement job report that indicates where Storage and Pool
162 came from -- with overrides and Pool storage and NextPool,
163 it is all very complicated.
164 - Create a few common subroutines for backup.c and migrate.c
165 - Eliminate warning messages from tray-monitor.c
167 - Apply fix for scheduler for missing unlock_jobs().
168 - Clean out autoconf caches before rebuilding configure
169 - Fix the FreeBSD pthreads lib define updated by Dan.
170 - Add missing semicolon in src/console/console.c
171 - Add Robert's enhanced version of build-win32-cross-tools
172 as well as his patch file.
173 - Note, the new gcc options are likely to break building of
174 wx-console or the tray-monitor ...
176 - Implement Robert's suggestion to suppress printing compile lines.
177 - Detect GCC correctly (I hope)
178 - If using GCC (actually g++) add the following compiler flags
179 -fno-strict-aliasing -fno-exceptions -fno-rtti
181 ===================== Warning =============================
182 Turn on new bsnprintf() code. This could cause print output
184 ===========================================================
186 - Fix crypto when not using openssl. I previously overlooked this.
187 - Eliminate crypto type punning problems by eliminating void * and
188 using uint8_t * instead.
189 - Harden authentication failure in FD by single threading errors
190 and forcing a 6 second wait.
192 ===================== Warning =============================
193 Removed "Accept Any Volume" directive.
194 ===========================================================
196 - Major cleanup and simplification of regress using shell functions
197 (more tests to be converted to new format)
198 - Add detection of 64 bit Irix as indicated by user.
199 - Increment catalog database id from 9 to 10.
200 - Add a few of the new database fields to the various C++
202 - Implement code to fill in the most important new fields (e.g.
204 - Ensure that most of the new fields are in create/get/insert/update.
205 More work to be done.
207 - Update some very old licenses in src/cats.
208 - Modify the database format for handling Migration jobs:
209 Add PriorJobId, RealEndTime to Job table
211 Remove Stripe from JobMedia record (not used, wasting space)
212 Add ScratchPoolId, RecyclePoolId, Enabled to Media record
213 Add Cost to Location table.
214 - Delete mac_record subroutines.
215 - Add new table values to create/update/get routines as appropriate.
216 - Update the update_tables and make_tables scripts.
217 - Make a first cut in using a union where multiple pointers (actually
218 pointers to pointers) are stored.
219 - Make some changes to base64.c suggested by Robert and others and add
220 a small amount of documentation.
221 - Add some changes suggested by Peter for Solaris (bregex...)
222 - Make runscript.h always built because it needs to be pulled into
223 more places. In particular its RES was left out of a union.
224 - Renumber the enum{} list in runscript.h to avoid starting from zero.
225 This makes it slightly easier to detect uninitialized variables.
227 - Remove xpg4 lib from FreeBSD build as it is no longer needed and
228 sometimes creates install problems.
229 - Change a couple of %T in strftime to %H:%M:%S as seen in Robert
230 Nelson's code. More to do.
231 - Convert a few store_bit to store_bool in filed_conf.c.
232 - DVD tweak to block.c as supplied by Uwe Muessel.
234 - Modify configure.in to check for typeof().
235 - In findlib/attribs.c use typeof instead of templates if possible.
237 - Implement partial error messages for src/filed/acl.c
238 - Fix return code to fail when doing update barcodes and
239 Pool resource is not found. This previously caused Volumes
240 to be labeled as cleaning tapes.
241 - Small updates to projects.
243 - Add Eric Bollengier patch for new RunScript directive. It breaks
244 compatibility between dird and fd.
246 - Prevent DVD code from rewriting label if the device cannot be
248 - Used __GNUC__ to detect GNU g++ as suggested by John Goerzen to
249 fix Debian build bug.
250 - Fix Dir/SD run race in migrate.c, verify.c, and restore.c
251 - Integrate manpages contributed by Jose Luis.
252 - Update projects file.
253 - Add manpages Makefile.in to configure.in
254 - Remove installation of bacula.8 in scripts directory.
256 - Implement patch submitted by cesarb in bug #606 to implement O_NOATIME
258 O_NOATIME is a open() flag which makes it possible to read a file without
259 updating the inode atime (and also without the inode ctime update which
260 happens if you try to set the atime back to its previous value). It also
261 prevents a race condition when two programs are reading the same file, but
262 only one does not want to change the atime. It's most useful for backup
263 programs and file integrity checkers (and bacula can fit on both
266 Recent versions of the Linux kernel and glibc have support for it (the
267 glibc support being mostly copying the O_NOATIME definition to
268 bits/fcntl.h). If there's no support for it on the kernel, trying to use
269 it does nothing (since the kernel ignores unknown flags).
271 If the kernel has support for it, trying to use it either works, fails
272 silently (mostly in remote filesystems), or returns errno=EPERM (if you
273 are not either the owner of the file or root). A simple way to prevent the
274 failure is to open the file without the flag and set it later with
275 fcntl(F_SETFL), ignoring any EPERM errors.
276 - Applied patch posted by Jaime Ventura to bug #570 to correct failure
277 of restarted jobs to complete (due to a destroyed pthreads conditional
278 variable used by the message thread).
279 - Fixed bug #619 where Bacula would not restart jobs with MaxRestartTimes
281 - Apply patch from John Goerzen bug #611 to fix bad manual links in the code.
282 - Apply patch from Rudolf Cejka bug #614 to removed trailing
283 colons in query.sql file.
284 - Apply patch from Rudolf Cejka bug #613 to correct prunning of files
285 not to create orphans.
286 - Apply patch from Rudolf Cejka bug #617 to use mtime instead of ctime
287 in restore long listing.
288 - Apply patch from Rudolf Cejka bug #609 to use MarkId in verify.c instead
289 of incorrect MarkedId.
290 - Apply patch from adioso bug #616 to correct text in manual (replace
291 Director with File daemon).
292 - Rework a few of the zlib changes so that they build properly.
295 - Move DIR "run" command to SD before starting the message thread
296 as suggested by Cristopher Hull to avoid a race deadlock from
297 two threads using the bsock structure.
298 - Modify LICENSE to correct some problems pointed out by Debian.
300 - Correct EOF with no \n lex.c problem.
301 - Tweak compat.h compat.cpp and merge with apcupsd
302 - Add additional INFO messages if autochanger cannot be used.
305 - Simplify Win32 Makefile
306 - Correct count of buffers/bytes used by smartall.c
307 - Updated compat.h after porting apcupsd to MinGW.
309 - Work on getting wx-console building on MinGW. wxWidgets now builds
310 started adding Makefile ...
311 - Moved MinGW library objects to src/win32/lib to reduce clutter.
313 - Reloading a bad configuration file doesn't kill director any more.
314 Thanks to fix from Eric Bollengier.
316 - Fix problem of accents with new Win32 code.
317 - Integrate Howard's VSS patch. Tweak it a bit. VSS now
318 works in the MinGW build with the exception that there
319 is a problem with accented characters -- i.e. there is
320 some mapping problem with wide characters.
321 - The code no longer compiles with Microsoft VC++.
323 - Start implementing Christopher's St.Bernard code.
324 - Add Christopher's mods for opening files.
325 - Strip \r \n from Win32 error messages (Christopher)
326 - Add more info about Win32 system if unknown (Christopher)
327 - Use bstrn(cpy, cat) routines to avoid overrun in compat.cpp
328 - Fix a number of incorrect strings marked for translation, when
329 they should not be (e.g. commands).
330 - Fix StartTime bug reported for ClientRunBeforeJob and RunBeforeJob
331 Patches supplied by user in bugs #608 and #599.
332 - Use most recent of st_ctime and st_mtime in restore tree DIR
333 listing. This fixes some problems with Win32 printing incorrect
335 - Correct the clock shift comm time calculation between the DIR
336 and the FD. It was added rather than subtracted.
337 - Strip all backslashes in filenames to be restored on Win32
338 machines -- the names should all be in Unix format, and if
339 there are any backslashes it is most likely due to an improper
340 FileSet definition, and the files will not otherwise be
342 - Remove lib/regex.c (replaced by bregex.c).
343 - Delete src/win32/License.txt
344 - Modify installer to install LICENSE
346 - Fix bug in build-win32-cross-tools script reported by Howard
347 - Implement cross compiled bconsole
349 - Fix barcode test, eliminate BOOL (doesn't exist on Linux).
350 - Write first cut of README.mingw32
351 - Add back ua_label barcode fix from Rufolf Cejka.
352 - Integrate the majority of the Mingw cross-tools submission
353 made by Howard Thomson. Everything compiles, with zlib, openssl,
354 and VSS turned off. Linking needs a pthreads library ...
356 - Add 'Dir Status' button to the gnome console.
357 - More work on migration.
358 - Restore non-compatible base64 coding to permit compatibility.
359 - Display heap stats in Storage daemon without debug level.
361 - Integrate most of the MinGW changes -- more to do.
363 - Implement using pg_config for finding PostgreSQL files.
364 Fixes bug #600. Patch supplied by user.
365 - Remove -t option from mktemp in mtx_changer.in and use
366 working directory. Fixes bug #578.
367 - Update job start time after the any run before job so that
368 files created by the script are only backed up once. Fixes
370 - Strip trailing newline only from filenames entered in
371 the restore command when reading a file. This permits
372 the user to enter filenames with trailing spaces. Fixes
373 bug #549. The user supplied a patch that I modified slightly.
374 - Use the most recent time (st_mtime, st_ctime) in the dir
375 command in restore. This gives the user a better idea of what
376 the newest file really is. This fixes bug #574. The fix
377 was suggested by the user.
378 - Implement a compatible version of base64. This permits external
379 programs to duplicate Bacula's base64 algorithm using standard
380 routines. This fixes bugs #296, and 565. Patch submitted by
382 ================= Note ========================
383 Previous Signatures stored in the database are no longer
384 compatible with this. The main downside is for Verify jobs,
385 and doing an InitCatalog run will fix the problem. Also, the
386 authentication between the deamons is changed, so all daemons
387 must be simultaneously upgraded.
388 ==============================================
389 If you don't like this fix, set:
390 const bool compatible = true;
392 const bool compatible = false;
395 - Ensure that DB signature is never NULL.
396 - Ensure that DB table names are not translated.
397 - Fix scheduler to handle time skew (eg daylight savings).
398 - Fix scheduler to use lock_jobs() to avoid most problems
399 with reload. Window is now milliseconds.
401 - Apply patch from Christopher Hull
402 - Allow multiple connections to database with different
404 - Invalidate the scheduler when doing a reload. Fixes seg
405 fault, but still 60 second window.
406 - Additional info in Reschedule message.
407 - Use set_jcr_job_status() everywhere to prevent loss of
409 - Display peer IP in FD if error from connecting DIR.
410 - Don't increment file count for DIRBEGIN.
411 - Replace illegal characters in Win32 filename by _.
412 - Add SE_CREATE_PERMANENT_NAME privilege in Win32.
413 - Hash hard link filenames rather than linked list.
414 - Fix for security failure in chdir on Win32.
415 - Add CreateDirectoryA/W win32 API entry points.
416 - Add /silent option to Win32 FD for Install/Remove service.
417 - Always print Heap statistics in FD.
419 - Remove the -f option from the chown in Makefile.in for more
421 - Change setting the group in Makefile.in to use chgrp for
423 - Implement a write_store and read_store to replace
424 dirstore in reserve.c
425 - Implement a Bacula read/write lock for Python rather
426 than using the Python lock to avoid recursive problems.
427 - Correct the uninstall directory names in filed/Makefile.in
428 as reported by a user.
430 - Correct some problems with database creation (new tables).
431 - Replace a bunch of old 0x%x by %p.
432 - Get first cut of Migration with Job selection working.
434 - Change store_bit() to store_bool() for the Spooling in
435 dird_conf.c -- bug reported by Robert Nelson.
437 - Correct Makefile bug found by Dan Langille.
439 - Correct Makefile for Solaris /bin/sh
440 - Correct mtx-changer.in for Solaris /bin/sh
441 - Abort if a conf resource does not have a Name =
443 - Change the name of the regex program to bregex.
444 - Add the bwild program to the tools directory. It is similar
445 to the bregex program.
446 - Implement create bregex.h and bregex.c in src/lib from the
447 Python regexp program.
448 - Use the new bregex.c to implement Regex expressions on Win32.
450 - More work on migration.
451 - Implement wild program in tools directory for testing
452 wild-cards. Almost identical to the regex program.
453 - Up port VSS single thread locking code.
454 - Replace N_(..) by NT_(...) to inhibit translation of
456 - Modify Makefile to change the permissions on Working Directory
457 to 770 if the directory is created.
458 - Do not fail the Makefile if changing the permissions or
459 owner/group on WorkingDir fails.
460 - Remove early selection of Scratch Volumes in the recycling
462 - Correct the old recycling algorithm so that Scratch Volumes
463 are selected when looking for a Volume in the changer.
464 - Correct a typo in the Verify SQL reported by Joe Park.
466 - Remove automatic case folding on Windows FDs. You must
467 explictly use the 'Ignore Case = yes' option.
468 - Remove the code added to 1.38.6 and 1.38.7 that pulls a
469 scratch volume in an Autochanger early in the 'recycling'
472 - Make weof() and clrerror() methods of class DEVICE.
474 - Tweak license to include Microsoft restrictions.
475 - Move mysql.reconnect to after real_connect(). Thanks to
476 Frank Sweetser for the patch.
477 - Disallow a backslash in a File = directive (Windows junk)
478 unless the string is quoted.
479 - Apply Eric's patch to ua_label.c so that daemon protocol
481 - Add NT_ definition for strings that should not be translated.
482 - Apply NT_() to ua_label.c
483 - Remove timed wait for VSS on Win2K3 as it is not yet
485 - Correct bacula.in script to reference bacula-ctl-xx in the
486 sysconfig directory rather than the bin directory.
488 - Fix a long standing bug in the bacula start/stop scripts.
489 - Attempt to add the new bacula-ctl-* files to the rpm.
490 - Switch to using typeof() for list traversing to avoid
491 as many FC5 g++ compiler warnings as possible.
493 - Split the bacula start/start script into four files:
494 bacula -- starts and stops calling other scripts
495 bacula-ctl-dir -- starts/stops the director
496 bacula-ctl-fd -- starts/stops the File daemon
497 bacula-ctl-sd -- starts/stops the Storage daemon
499 - Create datestyle fix for PostgreSQL. Fixes bug #574.
500 - Correct editing of JobId from int to int64 in fd_cmds.c
501 - Eliminate FileSet name race with bash_spaces() and multiple
502 threads by bashing in a local.
503 - Fix error return from 'use storage' to print a correct error
504 message rather than nothing.
505 - Correct false re-read last block error message when two jobs
506 are simultaneously writing at the end of a tape.
507 - Simplify exit conditions in the reserve.c code to avoid
508 possible non-release of reservation_lock().
509 - Suffle lock order in reserve to avoid deadlock between
510 reservation lock and device mutex.
512 - Initialize jcr mutex before first use. Thanks to Thorsten for
513 tracking this down for me !!!! as it broke the Win32 build.
515 - Integrate addition of line count limitation to bsmtp -l from
516 Sebastian Stark <stark at tuebingen.mpg.de>
518 - Implement regex test program in tools directory.
519 - Attempt to fix time problem with bsmtp with foreign langs.
520 - Add strip_trailing_newline() submitted by user.
521 - Implement regex matching in migrate.c
523 - Fix bug #537 to allow arbitrary time to mount a volume for
524 restore, if polling is turned on.
525 - If dir_user or dir-group is specified in ./configure apply it to
526 the working-dir. Fixes bug #533.
527 - If rescheduling a job cancel the previous incarnation with the SD.
528 Fixes bugs #566 and 557.
529 - Fix bug #567 do_message() definition type conflict.
532 - Add more jcr methods and make mutex and use_count private.
533 - Create lock/unlock methods for jcr.
534 - Fix PostgreSQL bug doing sql_data_seeek() by explicitly reading
535 records to get to seek position.
536 - Integrate patch from bug #561 to correct conio.c signal definitions.
537 - Fix Rescheduling failed Jobs. Ensure that SD message thread
538 terminates correctly by doing pthread_kill(). Do not destroy
539 SD cond wait variable between executions of the job. Use local
540 mutex for cond variable to avoid blocking jcr chain. Fix poor
541 use of jcr use count in jobq.c for restarted jobs.
542 - Fix obsolete usage of foreach_dlist() to use foreach_jcr() in
543 lib/jcr.c -- prevents locking the jcr chaing.
544 - Apply patch from bug #564, which corrects listing volumes with
545 multiple autochangers. Apply same fix to next volume list.
546 - Fix bug #562 where restore bootstrap file is not unique.
547 - More details in implementing David's migration syntax proposal.
548 - Save and restore dcr when swapping drives so that if we have
549 a write dcr it is not lost.
550 - Use new routine lock_reservations() to lock the reservations
551 system, and call it while looking for a volume in askdir.c.
552 This could possibly fix bug #543.
553 - Stop SD command loop if job is canceled.
556 - Remove old code from findlib/create_file.c
557 - Rename mac.c migrate.c
558 - Add user friendly display of VolBytes in job report.
559 - Rename target... to previous... to make it a bit easier to
561 - Add selection type and selection pattern to Migration (idea
562 given by David Boyes).
565 - The po files should now be current.
566 - Fix new sql_use_result() code to properly release the
567 buffers in all cases.
568 - Use the keyword ujobid to mean the unique job id; job or jobname
569 to mean the Job name given on the Name directive, and jobid to
570 be the numeric (non-unique) job id.
571 - Allow listing by any of the above.
572 - Add the user friendly job report code for reporting job elapsed time
573 and rates with suffexes from John Kodis <kodis at comcast.net>.
574 - Add Priority and JobLevel as Python settable items.
575 - Use TEMPORARY table creation where the table is created by
577 - Add new code submitted by Eric for waiting on specific jobid.
578 - Add ACL checking for the dot commands.
579 - Fix restore of writable FIFOs.
580 - Fix a bug in bpipe where the string was freed too early.
582 - Modify the Python class examples to inherit object -- new way
583 of defining classes. Patch from Felix Schwarz.
584 - Implement jobuid to replace old usage of job in keywords as
585 suggested by Eric Bollengier.
586 - Apply patch for enhancing wait from Eric Bollengier. On can now:
587 wait (wait for all jobs to stop)
589 wait jobuid=unique id
591 - Implement write variables for Python to set Priority (anytime), and
592 Job Level, only during JobInit event.
594 - Fix the block listing bug pointed out by Arno.
595 - Update the po files at Eric's request.
597 - Fix Maximum Changer Wait, Maximum Open Wait, Maximum Rewind Wait to
598 accept time qualifiers.
599 - Make catalog respect ACL.
600 - Add recycle count to Media record.
602 Changes to 1.39.6-beta1 24Feb04
604 - Add Rudolf Cejka's new rc-chio-changer.
605 - Implement pulling Volume from Scratch Pool if the
606 Volume is in the autochanger.
607 - Implement additional command arguments for update Volume.
611 - Back port changes to 1.38.5
612 - Fix recycle SQL for StorageId.
613 - Fix misplaced quote in mtx-changer script.
615 - Note. Your database must be updated, or you must create
616 a new database. I have not yet incremented the database level.
617 - Add Location table.
618 - Add LocationId, DeviceId, and MediaTypeId to Media record.
620 - Implement create/get mac record in database for adding extended
621 Migration data to the job record.
622 - Add new MAC table to update/make database scripts.
623 - Return Storage name used when getting VolumeNames for a job.
624 - Change bsr file keyword Storage to Device, which is more accurate.
625 - Ensure that Mac records are pruned/purged.
626 - Tweak SD tools to deal with changing media type.
627 - Integrate more dev.c subroutines as methods (e.g. strerror, bsr, ...)
628 - Pass pointer to dcr pointer to acquire_device_for_read() so
629 that the subroutine can switch devices, and hence dcrs.
630 - Modify the multiple MediaType read code to re-use the same
631 dcr when switching devices. This makes the code much more
633 - Integrate patch from Karl Hakimian that reads JobIds, FileIndexes
634 from a table for restore.
635 - Add Storage name to VolParams, but it really should be Device.
637 - Add disk-changer to scripts directory + configure/Makefile
638 - Eliminate PoolId from jcr -- it is in jcr->jr.PoolId
639 - Implement store_bit scanner to replace store_yesno. Mostly done.
640 - Implement new store_bool that stores in a bool.
641 - Add true/false to yes/no conf directives.
642 - Make first cut at changing appropriate store_yesno to store_bool.
643 - Complete implementation of Pool storage devices.
644 - Move starting clones to job.c
645 - Move create_restore_bootstrap_file() to job.c
646 - Make copy_storage() more general to be able to handle
648 - Cleanup a lot of migration code for manual running, including
650 - Move getting a scratch Volume into a subroutine.
651 - Make all places a Volume that is added to a pool to
653 - Fix bug in autochanger recycle code (improper edit 64 bit).
654 - Fix segfault in restore command when no value specified.
655 - Start adding code to handle multiple MediaTypes in restore.
656 - Eliminate race condition in getting Volume name for
657 two drive autochanger.
658 - More debug code in autochanger.
659 - Add storage keyword to bootstrap file, add parsing.
660 - Move slot in bsr file into Volume record as there will be
661 a different slot for each Volume.
664 - Implement Pool storage overrides.
666 - Implement first cut of Migration.
667 - Implement mysql_use_result() from patch by Karl Hakimian.
668 This reduces significantly the memory consumption during
669 the restore tree building, and hence runs faster too.
670 - Implement StorageId patch supplied by user (reggie) in bug
671 #536. This should permit Bacula to work correctly with two
673 - Implement Job listing variations suggested by a user.
674 - Move updating bootstrap code in backup.c to subroutine
675 update_bootstrap_file().
676 - Add new job status elapsed time and bytes written user
677 friendly job report output patch sent by John Kodis
678 <kodis at comcast.ne>.
679 - Implement a storage list in Pools.
680 - Separate out setup_job() code from run_job().
681 - Get migration working -- lots of changes in mac.c in both
683 - Apply patch from user (Eric Bollengier I think) that fixes a
684 DIR crash when no arguments are supplied to a dot command.
685 - Fix typo (strcpy->strcmp) in ua_prune.c as reported by Martin.
686 - Fix command arg name->volume in label command of gnome-console.
687 - Fix SD acquire.c to release correct DCR by explicitly testing
688 on the dcr address rather than trying to devine if it is a read
689 or write dcr. This failed in error conditions when the device was
693 - Apply user supplied patch for more readable rate output
695 - Continue implementing migration.
696 - Implement support for removable filesystems in SD.
697 - Ensure that btraceback scripts can be read by anyone.
698 - Replace dvd-freespace and dvd-writepart by dvd-handler.
699 - Correct bug where canceling restore before the FD contacts
700 the SD causes the drive to be left in read mode.
701 - Move ofline_or_rewind into DEVICE::close().
702 - Eliminate close_device.
703 - Convert several dev subroutines to methods (e.g. bsf,
705 - Eliminate force_close_device().
706 - Implement Device Type directive in Device resource that
707 can have values File, Tape, Fifo, DVD, or Prog.
708 - Add has_cap() method to Device.
711 - Add patch from bug #527 to allow RedHat user to specify
712 options/user/group for starting each daemon in
715 - Add two new queries to query.sql provided by Arno. One
716 list volumes known to the Storage device, and the other
717 lists volumes possibly needing replacement (error, ...).
719 - Add periodic (every 24 hours) garbage collection of memory
720 pool by releasing free buffers.
722 - Correct bug counting sized (for display only) in smartall.c
723 - Print FD mempool stats if debug > 0 rather than 5.
725 - Make db_lock() mutex error fail the job rather than abort
726 Bacula. Canceling the job caused the mutex to fail.
727 - Correct bug in alist.c that re-allocated the list if the
728 number of items goes to zero.
729 - Move the reservation system thread locking to the top level
730 so that one job at a time tries all possible drives before
732 - Implement a reservation 'fail' message queue that is built
733 and destroyed on each pass through the reservation system.
734 These messages are displayed in a 'Jobs waiting to reserve
735 a drive' list during a 'status storage='. Note, multiple
736 messages will generally print for each JobId because they
737 represent the different problems with either the same drive
738 or different drives. If this output proves too confusing
739 of voluminous, I will display it only when debug level 1
740 or greater is enabled in the SD.
742 - Add enable/disable job=<job-name>. This command prevents
743 the specified job from being scheduled. Even when disabled,
744 the job can be manually started from the console.
745 - During 'update slots' clear all InChanger flags where the
746 StorageId is zero (old Media records).
750 - Fix autochanger code to strip leading spaces from returned
751 slots number. Remove bc from chio-changer.
752 - Back port a bit of 1.39 crypto code to reduce diffs.
753 - Fix first call to autochanger that missed close()ing the
754 drive. Put close() just before each run_program(). Fixes
757 - Add PoolId to Job record when updating it at job start time.
759 - Pull in more code from 1.39 so that there are fewer file
760 differences (the new ua_dotcmds.c, base64.h, crypto.h
761 hmac.c jcr.c (dird and lib) lib.h md5.h parse_conf.c
762 util.c. Aside from ua_dotcmds.c these are mostly crypto
764 - Implement new method of walking the jcr chain. The
765 incr/dec of the use_count is done within the walking
766 routines. This should prevent a jcr from being freed
767 from under the walk routines.
772 - Start implementing Verify list output.
773 - Move the suitable_drive flag to a better place to prevent
774 premature termination of the reservation if all drives
775 are busy -- should fix Arno's diff/inc pool failures.
777 - Add mutex to single thread VSS code in Win32.
779 - Simplify code in askdir.c that waits for creating an appendable
780 volume so that it can handle multiple returns from the wait
782 - Modify the wait code to permit multiple returns.
783 - Return a zero when "autochanger drives" is called and
784 it is not an autochanger.
785 - Make rewind_dev() a method taking a DCR as an argument.
786 This permits closing and reopening the drive if the
787 rewind fails as happens if the drive was loaded while the
788 file descriptor was open. This refreshes the file descriptor.
789 - Remove the ST_OPENED flag and always rely on fd < 0 for knowing
790 if the device is open or not. This should eliminate
792 - Return error if reserve cannot find at least one suitable device.
793 - Make wait_for_sysop() return correct state information.
794 - Fix Win32 state file problem. write was not using compat
795 code. This should fix bug #500.
797 - Modify gui on command to set only GUI mode and not batch.
798 - Modify .messages command to always print messages regardless
800 - If GUI mode is on, suppress automatic printing of
802 - Delete old bnet packet code.
803 - Ignore new BNET_START_SELECT and BNET_END_SELECT signals in
804 wx-console. -- not implemented because bconsole must be updated.
805 - Modify restore command in wx-console to set gui on and to use
806 only .messages instead of messages. Hopefully this fixes bug
808 - Fix seg fault in exit of acquire when canceling a job --
809 reported by Wolfgang Denk
810 - Pull in latest reservation system changes from 1.38
811 - Make .messages command always print messages regardless
812 of the automessages flag.
814 - Fix seg fault if user labels a drive directory bug #513
815 - Remove quotes around Version as it breaks things.
817 - Merge in Aleksandar Milivojevic's mods to the spec file.
818 - Apply sparse code fix for raw drives and fifos. Bug 506
819 - Thorsten fixed Unicode cd problem with wx-console bug 505.
821 - Correct reservation system to do a last ditch try
822 for any mounted volume, then anyone anywhere.
823 - Add quotes around table Version because of
824 error in MySQL 4.1.15 -- bug report submitted.
825 - Correct some minor problems with btape in the fill
827 - Updates to ssh-tunnel from Joshua Kugler.
828 - Added a report.pl program from Jonas Bjorklund.
829 - Simplify the O_NONBLOCK open() code for tape drives,
830 and always open nonblocking.
831 - Do not wait for open() if EIO returned (shouldn't happen).
832 - Eliminate 3 argument to tape open().
833 - Correct the slot # edited in the 3995 Bad autochanger unload
835 - With -S on bscan (show progress) do not divide by zero.
837 - Make cancel pthread_cond_signal() pthread_cond_broadcast().
838 - When dcr is freed, also broadcast dev->wait_next_vol signal.
839 - Remove unused code in wait_for_device.
840 - Make wait_for_device() always return after 60 seconds of wait.
844 - Add stubs for non-crypto build.
846 - Use localhost if no network configured
848 - Eliminated duplicate MaxVolBytes in cat update -- bug 509.
849 - Remove debug print.
850 - Add bail_out in error during state file reading.
852 - Merge changes made to 1.38.3 into HEAD
853 - Add stubs for pygtk-console code
854 - Create Makefile.in for pygtk-console code
856 - Merge updates into 1.38 branch
857 - Update specs to include mysql4 define.
858 - Fix when attributes are sent, must be after binit().
859 - Stop read_record() if status not ok in second loop.
860 - Return rec->FileIndex in dcr->VolLastIndex for normal
861 and partial records in read_record(). This allows bscan
862 to get FileIndex at EOT correct.
863 - Fix butil.c to correctly set dcr -- fixes seg fault in bls.
865 - Fix Win32 built to work with new crypto code.
866 - Apply patch supplied by user (slightly modified) to fix
867 correct detection of holes in block devices and FIFOs.
869 - Apply patch supplied by user (slightly modified)
870 to fix SD hang with multiple pools and bad client
873 - Add nagios plugin to the examples directory. Submitted by
875 - Remove warning message about multiple saves of hardlinked files
876 from find_one.c as it can generate too many warning messages.
877 - Modify most restore error messages to be queued so that they
878 appear at the end of the job rather than mixted with the restore
879 listing where they could be "lost".
881 - Reset timeout values before select() per patch from
882 Frank Sweetser for problems with non-blocking sockets.
883 - Unlink the state file if either reading or writing it gets
884 errors. Hopefully this will fix Win32 exit problems.
885 - Add sanity check in append.c to ensure that dcr is not NULL.
886 This can happen if multiple drive autochanger SCSI control
887 channel and drive indicies do not correspond.
889 - Get next volume from Scratch pool before creating a volume.
890 - Set new Pool defaults in Vol when moved from Scratch Pool.
891 - Remove argument from create_bacula_database for SQLite as it
893 - Add back index code so that two drive autochangers can get
895 - Change a bunch of debug levels to aid debugging autochangers.
896 - Fix reservation so that mutexes are properly applied.
897 - Rework reservation algorithm so that two drives can be used
900 - Landon merged his data encription changes into the HEAD
901 - Apply days keyword patch from Alexander.Bergolth at wu-wien.ac.at
902 If this patch is applied, the number of days can be specified with
903 "list nextvol days=xx"
906 My use case is to be able to preview the next scheduled job (and the
907 next tape to be used) on fridays if there are no scheduled jobs during
911 - Fix font code in gnome2 console user patch. Fixes bug #501.
912 - Fix malformatted bnet error message that caused seg fault
914 - Applied user patch to improve README.vc8 in src/win32.
916 - Add Migrate, Copy, Archive Job types (some where there)
917 - Correct some more editing of JobId's (for 64 bit compatibility).
918 - Ensure that StorageId is stored in Media record when ever possible.
919 - Add Migration Job to Job.
920 - Add Migration Time, Migration High Bytes, Migration Low Bytes
921 Next Pool to Pool resource.
922 - Add more code to mac.c (migration archive copy).
923 - Change Start Storage daemon job to require read and write storage
925 - Pass read storage data to SD as well as write storage data.
926 - Remove old code from winservice.cpp
927 - Break on error in scan.
928 - Fix typo in signal.c
929 - Separate read/write DCR in SD. Add jcr->read_dcr.
930 - Cleanup how find_device() works.
931 - Add read output to Status in SD.
934 - Add red-black btree routines
936 - Remove abs() in bfile.c so that it compiles on Solaris.
939 - Fix crash in tray-monitor when daemon disconnects. Bug #479.
940 - Fix bnet-server bug found on OpenBSD. Bug #486
941 - Fix cancel failure bug. Bug #481
942 - Fix failure when Pool name has spaces. Bug #487
943 - Fix SD crash in autochanger code. Mutex failure. Bug #488
944 - Fix a couple of free()s in src/filed/acl.c
945 - Fix memory overrun in bfile.c in building OS X resource
946 fork filename. Bug #489
947 - Add Pool name to SD status output.
949 - Apply SunOS patch for ACLs submitted by David Duchscher.
950 - Make sure to set storage before trying to set drive.
951 - Add bacula_mail_summary.sh to examples directory. It makes
952 a single email summary of any number of jobs. Submitted
954 - Make sure when we do a mount to unblock the device even
955 if the drive could not be opened.
957 - Remove the USE_WIN32STREAMEXTRACTION #defines (always on)
958 and correct a few minor problems to make it build on Linux.
960 - Remove delete of CVS from all Makefiles
961 - Fix seg fault when clicking on Add button in wx-console
962 restore panel. Bug #470.
963 - Fix copyright date and URL typo -- bug #468.
964 - Change autostart install for FreeBSD to look for rc.conf
965 rather than rc.local as suggested fix for bug #466.
966 - Apply patch supplied by Eric Bollengier to fix PostgreSQL
967 grant on status. Bug #465
968 - Apply patch supplied by Eric Bollengier to fix PostgreSQL
969 update script. Bug #464
970 - Tweak #ifdefing a bit in new Win32 stream code.
971 - Fix #ifdeffing for FD_NO_SEND_TEST.
972 - Add documentation of performance #defines
975 Release 1.38.11 (28Jun06) released 28JunO6