1 Technical notes on version 1.39
6 - Implement using pg_config for finding PostgreSQL files.
7 Fixes bug #600. Patch supplied by user.
8 - Remove -t option from mktemp in mtx_changer.in and use
9 working directory. Fixes bug #578.
10 - Update job start time after the any run before job so that
11 files created by the script are only backed up once. Fixes
13 - Strip trailing newline only from filenames entered in
14 the restore command when reading a file. This permits
15 the user to enter filenames with trailing spaces. Fixes
16 bug #549. The user supplied a patch that I modified slightly.
17 - Use the most recent time (st_mtime, st_ctime) in the dir
18 command in restore. This gives the user a better idea of what
19 the newest file really is. This fixes bug #574. The fix
20 was suggested by the user.
21 - Implement a compatible version of base64. This permits external
22 programs to duplicate Bacula's base64 algorithm using standard
23 routines. This fixes bugs #296, and 565. Patch submitted by
25 ================= Note ========================
26 Previous Signatures stored in the database are no longer
27 compatible with this. The main downside is for Verify jobs,
28 and doing an InitCatalog run will fix the problem. Also, the
29 authentication between the deamons is changed, so all daemons
30 must be simultaneously upgraded.
31 ==============================================
32 If you don't like this fix, set:
33 const bool compatible = true;
35 const bool compatible = false;
38 - Ensure that DB signature is never NULL.
39 - Ensure that DB table names are not translated.
40 - Fix scheduler to handle time skew (eg daylight savings).
41 - Fix scheduler to use lock_jobs() to avoid most problems
42 with reload. Window is now milliseconds.
44 - Apply patch from Christopher Hull
45 - Allow multiple connections to database with different
47 - Invalidate the scheduler when doing a reload. Fixes seg
48 fault, but still 60 second window.
49 - Additional info in Reschedule message.
50 - Use set_jcr_job_status() everywhere to prevent loss of
52 - Display peer IP in FD if error from connecting DIR.
53 - Don't increment file count for DIRBEGIN.
54 - Replace illegal characters in Win32 filename by _.
55 - Add SE_CREATE_PERMANENT_NAME privilege in Win32.
56 - Hash hard link filenames rather than linked list.
57 - Fix for security failure in chdir on Win32.
58 - Add CreateDirectoryA/W win32 API entry points.
59 - Add /silent option to Win32 FD for Install/Remove service.
60 - Always print Heap statistics in FD.
62 - Remove the -f option from the chown in Makefile.in for more
64 - Change setting the group in Makefile.in to use chgrp for
66 - Implement a write_store and read_store to replace
68 - Implement a Bacula read/write lock for Python rather
69 than using the Python lock to avoid recursive problems.
70 - Correct the uninstall directory names in filed/Makefile.in
71 as reported by a user.
73 - Correct some problems with database creation (new tables).
74 - Replace a bunch of old 0x%x by %p.
75 - Get first cut of Migration with Job selection working.
77 - Change store_bit() to store_bool() for the Spooling in
78 dird_conf.c -- bug reported by Robert Nelson.
80 - Correct Makefile bug found by Dan Langille.
82 - Correct Makefile for Solaris /bin/sh
83 - Correct mtx-changer.in for Solaris /bin/sh
84 - Abort if a conf resource does not have a Name =
86 - Change the name of the regex program to bregex.
87 - Add the bwild program to the tools directory. It is similar
88 to the bregex program.
89 - Implement create bregex.h and bregex.c in src/lib from the
90 Python regexp program.
91 - Use the new bregex.c to implement Regex expressions on Win32.
93 - More work on migration.
94 - Implement wild program in tools directory for testing
95 wild-cards. Almost identical to the regex program.
96 - Up port VSS single thread locking code.
97 - Replace N_(..) by NT_(...) to inhibit translation of
99 - Modify Makefile to change the permissions on Working Directory
100 to 770 if the directory is created.
101 - Do not fail the Makefile if changing the permissions or
102 owner/group on WorkingDir fails.
103 - Remove early selection of Scratch Volumes in the recycling
105 - Correct the old recycling algorithm so that Scratch Volumes
106 are selected when looking for a Volume in the changer.
107 - Correct a typo in the Verify SQL reported by Joe Park.
109 - Remove automatic case folding on Windows FDs. You must
110 explictly use the 'Ignore Case = yes' option.
111 - Remove the code added to 1.38.6 and 1.38.7 that pulls a
112 scratch volume in an Autochanger early in the 'recycling'
115 - Make weof() and clrerror() methods of class DEVICE.
117 - Tweak license to include Microsoft restrictions.
118 - Move mysql.reconnect to after real_connect(). Thanks to
119 Frank Sweetser for the patch.
120 - Disallow a backslash in a File = directive (Windows junk)
121 unless the string is quoted.
122 - Apply Eric's patch to ua_label.c so that daemon protocol
124 - Add NT_ definition for strings that should not be translated.
125 - Apply NT_() to ua_label.c
127 - Remove timed wait for VSS on Win2K3 as it is not yet
129 - Correct bacula.in script to reference bacula-ctl-xx in the
130 sysconfig directory rather than the bin directory.
132 - Fix a long standing bug in the bacula start/stop scripts.
133 - Attempt to add the new bacula-ctl-* files to the rpm.
134 - Switch to using typeof() for list traversing to avoid
135 as many FC5 g++ compiler warnings as possible.
137 - Split the bacula start/start script into four files:
138 bacula -- starts and stops calling other scripts
139 bacula-ctl-dir -- starts/stops the director
140 bacula-ctl-fd -- starts/stops the File daemon
141 bacula-ctl-sd -- starts/stops the Storage daemon
143 - Create datestyle fix for PostgreSQL. Fixes bug #574.
144 - Correct editing of JobId from int to int64 in fd_cmds.c
145 - Eliminate FileSet name race with bash_spaces() and multiple
146 threads by bashing in a local.
147 - Fix error return from 'use storage' to print a correct error
148 message rather than nothing.
149 - Correct false re-read last block error message when two jobs
150 are simultaneously writing at the end of a tape.
151 - Simplify exit conditions in the reserve.c code to avoid
152 possible non-release of reservation_lock().
153 - Suffle lock order in reserve to avoid deadlock between
154 reservation lock and device mutex.
156 - Initialize jcr mutex before first use. Thanks to Thorsten for
157 tracking this down for me !!!! as it broke the Win32 build.
159 - Integrate addition of line count limitation to bsmtp -l from
160 Sebastian Stark <stark at tuebingen.mpg.de>
162 - Implement regex test program in tools directory.
163 - Attempt to fix time problem with bsmtp with foreign langs.
164 - Add strip_trailing_newline() submitted by user.
165 - Implement regex matching in migrate.c
167 - Fix bug #537 to allow arbitrary time to mount a volume for
168 restore, if polling is turned on.
169 - If dir_user or dir-group is specified in ./configure apply it to
170 the working-dir. Fixes bug #533.
171 - If rescheduling a job cancel the previous incarnation with the SD.
172 Fixes bugs #566 and 557.
173 - Fix bug #567 do_message() definition type conflict.
176 - Add more jcr methods and make mutex and use_count private.
177 - Create lock/unlock methods for jcr.
178 - Fix PostgreSQL bug doing sql_data_seeek() by explicitly reading
179 records to get to seek position.
180 - Integrate patch from bug #561 to correct conio.c signal definitions.
181 - Fix Rescheduling failed Jobs. Ensure that SD message thread
182 terminates correctly by doing pthread_kill(). Do not destroy
183 SD cond wait variable between executions of the job. Use local
184 mutex for cond variable to avoid blocking jcr chain. Fix poor
185 use of jcr use count in jobq.c for restarted jobs.
186 - Fix obsolete usage of foreach_dlist() to use foreach_jcr() in
187 lib/jcr.c -- prevents locking the jcr chaing.
188 - Apply patch from bug #564, which corrects listing volumes with
189 multiple autochangers. Apply same fix to next volume list.
190 - Fix bug #562 where restore bootstrap file is not unique.
191 - More details in implementing David's migration syntax proposal.
192 - Save and restore dcr when swapping drives so that if we have
193 a write dcr it is not lost.
194 - Use new routine lock_reservations() to lock the reservations
195 system, and call it while looking for a volume in askdir.c.
196 This could possibly fix bug #543.
197 - Stop SD command loop if job is canceled.
200 - Remove old code from findlib/create_file.c
201 - Rename mac.c migrate.c
202 - Add user friendly display of VolBytes in job report.
203 - Rename target... to previous... to make it a bit easier to
205 - Add selection type and selection pattern to Migration (idea
206 given by David Boyes).
209 - The po files should now be current.
210 - Fix new sql_use_result() code to properly release the
211 buffers in all cases.
212 - Use the keyword ujobid to mean the unique job id; job or jobname
213 to mean the Job name given on the Name directive, and jobid to
214 be the numeric (non-unique) job id.
215 - Allow listing by any of the above.
216 - Add the user friendly job report code for reporting job elapsed time
217 and rates with suffexes from John Kodis <kodis at comcast.net>.
218 - Add Priority and JobLevel as Python settable items.
219 - Use TEMPORARY table creation where the table is created by
221 - Add new code submitted by Eric for waiting on specific jobid.
222 - Add ACL checking for the dot commands.
223 - Fix restore of writable FIFOs.
224 - Fix a bug in bpipe where the string was freed too early.
226 - Modify the Python class examples to inherit object -- new way
227 of defining classes. Patch from Felix Schwarz.
228 - Implement jobuid to replace old usage of job in keywords as
229 suggested by Eric Bollengier.
230 - Apply patch for enhancing wait from Eric Bollengier. On can now:
231 wait (wait for all jobs to stop)
233 wait jobuid=unique id
235 - Implement write variables for Python to set Priority (anytime), and
236 Job Level, only during JobInit event.
238 - Fix the block listing bug pointed out by Arno.
239 - Update the po files at Eric's request.
241 - Fix Maximum Changer Wait, Maximum Open Wait, Maximum Rewind Wait to
242 accept time qualifiers.
243 - Make catalog respect ACL.
244 - Add recycle count to Media record.
246 Changes to 1.39.6-beta1 24Feb04
248 - Add Rudolf Cejka's new rc-chio-changer.
249 - Implement pulling Volume from Scratch Pool if the
250 Volume is in the autochanger.
251 - Implement additional command arguments for update Volume.
255 - Back port changes to 1.38.5
256 - Fix recycle SQL for StorageId.
257 - Fix misplaced quote in mtx-changer script.
259 - Note. Your database must be updated, or you must create
260 a new database. I have not yet incremented the database level.
261 - Add Location table.
262 - Add LocationId, DeviceId, and MediaTypeId to Media record.
264 - Implement create/get mac record in database for adding extended
265 Migration data to the job record.
266 - Add new MAC table to update/make database scripts.
267 - Return Storage name used when getting VolumeNames for a job.
268 - Change bsr file keyword Storage to Device, which is more accurate.
269 - Ensure that Mac records are pruned/purged.
270 - Tweak SD tools to deal with changing media type.
271 - Integrate more dev.c subroutines as methods (e.g. strerror, bsr, ...)
272 - Pass pointer to dcr pointer to acquire_device_for_read() so
273 that the subroutine can switch devices, and hence dcrs.
274 - Modify the multiple MediaType read code to re-use the same
275 dcr when switching devices. This makes the code much more
277 - Integrate patch from Karl Hakimian that reads JobIds, FileIndexes
278 from a table for restore.
279 - Add Storage name to VolParams, but it really should be Device.
281 - Add disk-changer to scripts directory + configure/Makefile
282 - Eliminate PoolId from jcr -- it is in jcr->jr.PoolId
283 - Implement store_bit scanner to replace store_yesno. Mostly done.
284 - Implement new store_bool that stores in a bool.
285 - Add true/false to yes/no conf directives.
286 - Make first cut at changing appropriate store_yesno to store_bool.
287 - Complete implementation of Pool storage devices.
288 - Move starting clones to job.c
289 - Move create_restore_bootstrap_file() to job.c
290 - Make copy_storage() more general to be able to handle
292 - Cleanup a lot of migration code for manual running, including
294 - Move getting a scratch Volume into a subroutine.
295 - Make all places a Volume that is added to a pool to
297 - Fix bug in autochanger recycle code (improper edit 64 bit).
298 - Fix segfault in restore command when no value specified.
299 - Start adding code to handle multiple MediaTypes in restore.
300 - Eliminate race condition in getting Volume name for
301 two drive autochanger.
302 - More debug code in autochanger.
303 - Add storage keyword to bootstrap file, add parsing.
304 - Move slot in bsr file into Volume record as there will be
305 a different slot for each Volume.
308 - Implement Pool storage overrides.
310 - Implement first cut of Migration.
311 - Implement mysql_use_result() from patch by Karl Hakimian.
312 This reduces significantly the memory consumption during
313 the restore tree building, and hence runs faster too.
314 - Implement StorageId patch supplied by user (reggie) in bug
315 #536. This should permit Bacula to work correctly with two
317 - Implement Job listing variations suggested by a user.
318 - Move updating bootstrap code in backup.c to subroutine
319 update_bootstrap_file().
320 - Add new job status elapsed time and bytes written user
321 friendly job report output patch sent by John Kodis
322 <kodis at comcast.ne>.
323 - Implement a storage list in Pools.
324 - Separate out setup_job() code from run_job().
325 - Get migration working -- lots of changes in mac.c in both
327 - Apply patch from user (Eric Bollinger I think) that fixes a
328 DIR crash when no arguments are supplied to a dot command.
329 - Fix typo (strcpy->strcmp) in ua_prune.c as reported by Martin.
330 - Fix command arg name->volume in label command of gnome-console.
331 - Fix SD acquire.c to release correct DCR by explicitly testing
332 on the dcr address rather than trying to devine if it is a read
333 or write dcr. This failed in error conditions when the device was
337 - Apply user supplied patch for more readable rate output
339 - Continue implementing migration.
340 - Implement support for removable filesystems in SD.
341 - Ensure that btraceback scripts can be read by anyone.
342 - Replace dvd-freespace and dvd-writepart by dvd-handler.
343 - Correct bug where canceling restore before the FD contacts
344 the SD causes the drive to be left in read mode.
345 - Move ofline_or_rewind into DEVICE::close().
346 - Eliminate close_device.
347 - Convert several dev subroutines to methods (e.g. bsf,
349 - Eliminate force_close_device().
350 - Implement Device Type directive in Device resource that
351 can have values File, Tape, Fifo, DVD, or Prog.
352 - Add has_cap() method to Device.
355 - Add patch from bug #527 to allow RedHat user to specify
356 options/user/group for starting each daemon in
359 - Add two new queries to query.sql provided by Arno. One
360 list volumes known to the Storage device, and the other
361 lists volumes possibly needing replacement (error, ...).
363 - Add periodic (every 24 hours) garbage collection of memory
364 pool by releasing free buffers.
366 - Correct bug counting sized (for display only) in smartall.c
367 - Print FD mempool stats if debug > 0 rather than 5.
369 - Make db_lock() mutex error fail the job rather than abort
370 Bacula. Canceling the job caused the mutex to fail.
371 - Correct bug in alist.c that re-allocated the list if the
372 number of items goes to zero.
373 - Move the reservation system thread locking to the top level
374 so that one job at a time tries all possible drives before
376 - Implement a reservation 'fail' message queue that is built
377 and destroyed on each pass through the reservation system.
378 These messages are displayed in a 'Jobs waiting to reserve
379 a drive' list during a 'status storage='. Note, multiple
380 messages will generally print for each JobId because they
381 represent the different problems with either the same drive
382 or different drives. If this output proves too confusing
383 of voluminous, I will display it only when debug level 1
384 or greater is enabled in the SD.
386 - Add enable/disable job=<job-name>. This command prevents
387 the specified job from being scheduled. Even when disabled,
388 the job can be manually started from the console.
389 - During 'update slots' clear all InChanger flags where the
390 StorageId is zero (old Media records).
394 - Fix autochanger code to strip leading spaces from returned
395 slots number. Remove bc from chio-changer.
396 - Back port a bit of 1.39 crypto code to reduce diffs.
397 - Fix first call to autochanger that missed close()ing the
398 drive. Put close() just before each run_program(). Fixes
401 - Add PoolId to Job record when updating it at job start time.
403 - Pull in more code from 1.39 so that there are fewer file
404 differences (the new ua_dotcmds.c, base64.h, crypto.h
405 hmac.c jcr.c (dird and lib) lib.h md5.h parse_conf.c
406 util.c. Aside from ua_dotcmds.c these are mostly crypto
408 - Implement new method of walking the jcr chain. The
409 incr/dec of the use_count is done within the walking
410 routines. This should prevent a jcr from being freed
411 from under the walk routines.
416 - Start implementing Verify list output.
417 - Move the suitable_drive flag to a better place to prevent
418 premature termination of the reservation if all drives
419 are busy -- should fix Arno's diff/inc pool failures.
421 - Add mutex to single thread VSS code in Win32.
423 - Simplify code in askdir.c that waits for creating an appendable
424 volume so that it can handle multiple returns from the wait
426 - Modify the wait code to permit multiple returns.
427 - Return a zero when "autochanger drives" is called and
428 it is not an autochanger.
429 - Make rewind_dev() a method taking a DCR as an argument.
430 This permits closing and reopening the drive if the
431 rewind fails as happens if the drive was loaded while the
432 file descriptor was open. This refreshes the file descriptor.
433 - Remove the ST_OPENED flag and always rely on fd < 0 for knowing
434 if the device is open or not. This should eliminate
436 - Return error if reserve cannot find at least one suitable device.
437 - Make wait_for_sysop() return correct state information.
438 - Fix Win32 state file problem. write was not using compat
439 code. This should fix bug #500.
441 - Modify gui on command to set only GUI mode and not batch.
442 - Modify .messages command to always print messages regardless
444 - If GUI mode is on, suppress automatic printing of
446 - Delete old bnet packet code.
447 - Ignore new BNET_START_SELECT and BNET_END_SELECT signals in
449 - Modify restore command in wx-console to set gui on and to use
450 only .messages instead of messages. Hopefully this fixes bug
452 - Fix seg fault in exit of acquire when canceling a job --
453 reported by Wolfgang Denk
454 - Pull in latest reservation system changes from 1.38
455 - Make .messages command always print messages regardless
456 of the automessages flag.
458 - Fix seg fault if user labels a drive directory bug #513
459 - Remove quotes around Version as it breaks things.
461 - Merge in Aleksandar Milivojevic's mods to the spec file.
462 - Apply sparse code fix for raw drives and fifos. Bug 506
463 - Thorsten fixed Unicode cd problem with wx-console bug 505.
465 - Correct reservation system to do a last ditch try
466 for any mounted volume, then anyone anywhere.
467 - Add quotes around table Version because of
468 error in MySQL 4.1.15 -- bug report submitted.
469 - Correct some minor problems with btape in the fill
471 - Updates to ssh-tunnel from Joshua Kugler.
472 - Added a report.pl program from Jonas Bjorklund.
473 - Simplify the O_NONBLOCK open() code for tape drives,
474 and always open nonblocking.
475 - Do not wait for open() if EIO returned (shouldn't happen).
476 - Eliminate 3 argument to tape open().
477 - Correct the slot # edited in the 3995 Bad autochanger unload
479 - With -S on bscan (show progress) do not divide by zero.
481 - Make cancel pthread_cond_signal() pthread_cond_broadcast().
482 - When dcr is freed, also broadcast dev->wait_next_vol signal.
483 - Remove unused code in wait_for_device.
484 - Make wait_for_device() always return after 60 seconds of wait.
488 - Add stubs for non-crypto build.
490 - Use localhost if no network configured
492 - Eliminated duplicate MaxVolBytes in cat update -- bug 509.
493 - Remove debug print.
494 - Add bail_out in error during state file reading.
496 - Merge changes made to 1.38.3 into HEAD
497 - Add stubs for pygtk-console code
498 - Create Makefile.in for pygtk-console code
500 - Merge updates into 1.38 branch
501 - Update specs to include mysql4 define.
502 - Fix when attributes are sent, must be after binit().
503 - Stop read_record() if status not ok in second loop.
504 - Return rec->FileIndex in dcr->VolLastIndex for normal
505 and partial records in read_record(). This allows bscan
506 to get FileIndex at EOT correct.
507 - Fix butil.c to correctly set dcr -- fixes seg fault in bls.
509 - Fix Win32 built to work with new crypto code.
510 - Apply patch supplied by user (slightly modified) to fix
511 correct detection of holes in block devices and FIFOs.
513 - Apply patch supplied by user (slightly modified)
514 to fix SD hang with multiple pools and bad client
517 - Add nagios plugin to the examples directory. Submitted by
519 - Remove warning message about multiple saves of hardlinked files
520 from find_one.c as it can generate too many warning messages.
521 - Modify most restore error messages to be queued so that they
522 appear at the end of the job rather than mixted with the restore
523 listing where they could be "lost".
525 - Reset timeout values before select() per patch from
526 Frank Sweetser for problems with non-blocking sockets.
527 - Unlink the state file if either reading or writing it gets
528 errors. Hopefully this will fix Win32 exit problems.
529 - Add sanity check in append.c to ensure that dcr is not NULL.
530 This can happen if multiple drive autochanger SCSI control
531 channel and drive indicies do not correspond.
533 - Get next volume from Scratch pool before creating a volume.
534 - Set new Pool defaults in Vol when moved from Scratch Pool.
535 - Remove argument from create_bacula_database for SQLite as it
537 - Add back index code so that two drive autochangers can get
539 - Change a bunch of debug levels to aid debugging autochangers.
540 - Fix reservation so that mutexes are properly applied.
541 - Rework reservation algorithm so that two drives can be used
544 - Landon merged his data encription changes into the HEAD
545 - Apply days keyword patch from Alexander.Bergolth at wu-wien.ac.at
546 If this patch is applied, the number of days can be specified with
547 "list nextvol days=xx"
550 My use case is to be able to preview the next scheduled job (and the
551 next tape to be used) on fridays if there are no scheduled jobs during
555 - Fix font code in gnome2 console user patch. Fixes bug #501.
556 - Fix malformatted bnet error message that caused seg fault
558 - Applied user patch to improve README.vc8 in src/win32.
560 - Add Migrate, Copy, Archive Job types (some where there)
561 - Correct some more editing of JobId's (for 64 bit compatibility).
562 - Ensure that StorageId is stored in Media record when ever possible.
563 - Add Migration Job to Job.
564 - Add Migration Time, Migration High Bytes, Migration Low Bytes
565 Next Pool to Pool resource.
566 - Add more code to mac.c (migration archive copy).
567 - Change Start Storage daemon job to require read and write storage
569 - Pass read storage data to SD as well as write storage data.
570 - Remove old code from winservice.cpp
571 - Break on error in scan.
572 - Fix typo in signal.c
573 - Separate read/write DCR in SD. Add jcr->read_dcr.
574 - Cleanup how find_device() works.
575 - Add read output to Status in SD.
578 - Add red-black btree routines
580 - Remove abs() in bfile.c so that it compiles on Solaris.
583 - Fix crash in tray-monitor when daemon disconnects. Bug #479.
584 - Fix bnet-server bug found on OpenBSD. Bug #486
585 - Fix cancel failure bug. Bug #481
586 - Fix failure when Pool name has spaces. Bug #487
587 - Fix SD crash in autochanger code. Mutex failure. Bug #488
588 - Fix a couple of free()s in src/filed/acl.c
589 - Fix memory overrun in bfile.c in building OS X resource
590 fork filename. Bug #489
591 - Add Pool name to SD status output.
593 - Apply SunOS patch for ACLs submitted by David Duchscher.
594 - Make sure to set storage before trying to set drive.
595 - Add bacula_mail_summary.sh to examples directory. It makes
596 a single email summary of any number of jobs. Submitted
598 - Make sure when we do a mount to unblock the device even
599 if the drive could not be opened.
601 - Remove the USE_WIN32STREAMEXTRACTION #defines (always on)
602 and correct a few minor problems to make it build on Linux.
604 - Remove delete of CVS from all Makefiles
605 - Fix seg fault when clicking on Add button in wx-console
606 restore panel. Bug #470.
607 - Fix copyright date and URL typo -- bug #468.
608 - Change autostart install for FreeBSD to look for rc.conf
609 rather than rc.local as suggested fix for bug #466.
610 - Apply patch supplied by Eric Bollinger to fix PostgreSQL
611 grant on status. Bug #465
612 - Apply patch supplied by Eric Bollinger to fix PostgreSQL
613 update script. Bug #464
614 - Tweak #ifdefing a bit in new Win32 stream code.
615 - Fix #ifdeffing for FD_NO_SEND_TEST.
616 - Add documentation of performance #defines