4 Documentation to do: (any release a little bit at a time)
5 - Document running a test version.
6 - Document query file format.
7 - Document static linking
8 - Document problems with Verify and pruning.
9 - Document how to use multiple databases.
10 - For FreeBSD typical /dev/nrsa0 and for mtx /dev/pass1
11 - VXA drives have a "cleaning required"
12 indicator, but Exabyte recommends preventive cleaning after every 75
14 - Lookup HP cleaning recommendations.
15 - Lookup HP tape replacement recommendations (see trouble shooting autochanger)
16 - Document FInclude ...
17 - Document need to add "-u root" to most of MySQL script calls
18 (./create_mys... ./make_my...).
19 - Document all the status codes JobLevel, JobType, JobStatus.
20 - Document dynamic DNS
22 Testing to do: (painful)
23 - that ALL console command line options work and are always implemented
24 - blocksize recognition code.
25 - Test if rewind at end of tape waits for tape to rewind.
27 - Test not zeroing Autochanger slot when it is wrong.
28 - Test recycling and purging (code changed in db_find_next_volume and
30 - Figure out how to use ssh or stunnel to protect Bacula communications.
33 - Increment DB version prior to releasing.
34 - Turn off FULL_DEBUG prior to releasing.
38 - Maybe remove multiple simultaneous devices code in SD.
39 - On Windows with very long path names, it may be impossible to create
40 a file (and thus restore it) because the total length is too long.
41 We must cd into the directory then create the file without the
43 - lstat() is not going to work on Win32 for testing date.
44 - Something is not right in last block of fill command.
45 - Implement a Recycle command
46 - Implement List Volume Job=xxx or List scheduled volumes or Status Director
47 - Add FileSet to command line arguments for restore.
48 - Do full check the command line args in update (e.g. VolStatus ...).
49 - Allow multiple Storage specifications (or multiple names on
50 a single Storage specification) in the Job record. Thus a job
51 can be backed up to a number of storage devices.
52 - Add client name to cram-md5 challenge so Director can immediately
53 verify if it is the correct client.
54 - Implement ClientRunBeforeJob and ClientRunAfterJob.
55 - Implement new alist in FileSet scanning.
56 - Add JobLevel in FD status (but make sure it is defined).
57 - Audit all UA commands to ensure that we always prompt where possible.
58 - Restrict characters permitted in a Resource name, and don't permit
60 - Prohibit backing up archive device (findlib/find_one.c:128)
61 - Make | and < work on FD side.
62 - Check Jmsg in bnet, may not work, must dup bsock.
63 - Suppress Job Name in Jmsg for console
64 - Create Pools that are referenced in a Run statement at startup if possible.
65 - Use runbeforejob to unload, then reload a volume previously used,
66 then the next job run gets an error reading the drive.
67 - Make bootstrap filename unique.
68 - Test a second language e.g. french.
69 - Start working on Base jobs.
70 - Make "make binary-release" work from any directory.
71 - Unsaved Flag in Job record (use JobMissingFiles).
72 - Base Flag in Job record.
73 - Implement UnsavedFiles DB record.
74 - Implement argc/argv for daemon command line scanning using table driven
76 - Implement table driven single argc/argv scanner to pickup all arguments.
77 Much like xxx_conf.c scan table.
78 keyword, handler(store_routine), store_address, code, flags, default.
79 - From Phil Stracchino:
80 It would probably be a per-client option, and would be called
81 something like, say, "Automatically purge obsoleted jobs". What it
82 would do is, when you successfully complete a Differential backup of a
83 client, it would automatically purge all Incremental backups for that
84 client that are rendered redundant by that Differential. Likewise,
85 when a Full backup on a client completed, it would automatically purge
86 all Differential and Incremental jobs obsoleted by that Full backup.
87 This would let people minimize the number of tapes they're keeping on
88 hand without having to master the art of retention times.
89 - Implement new serialize subroutines
90 send(socket, "string", &Vol, "uint32", &i, NULL)
91 - Scratch Pool where the volumes can be re-assigned to any Pool.
92 - Implement a M_SECURITY message class.
93 - Implement forward spacing block/file: position_device(bsr) --
94 just before read_block_from_device();
95 - When doing a Backup send all attributes back to the Director, who
96 would then figure out what files have been deleted.
97 - Currently in mount.c:236 the SD simply creates a Volume. It should have
98 explicit permission to do so. It should also mark the tape in error
100 - Make sure all restore counters are working correctly in the FD.
101 - SD Bytes Read is wrong.
102 - Configure mtx-changer to have correct path to mtx.
103 - Look at ALL higher level routines that call block.c to be sure
104 they don't expect something in errmsg.
105 - Investigate doing RAW backup of Win32 partition.
106 - Add JobName= to VerifyToCatalog so that all verifies can be done at the end.
107 - Add thread specific data to hold the jcr -- send error messages from
108 low level routines by accessing it and using Jmsg().
109 - Cancel waiting for Client connect in SD if FD goes away.
111 - Add Progress command that periodically reports the progress of
113 - One block was orphaned in the SD probably after cancel.
114 - Add all command line arguments to "update", e.g. slot=nn volStatus=append, ...
117 - Examine Bare Metal restore problem (a FD crash exists somewhere ...).
118 - Implement timeout in response() when it should come quickly.
119 - Implement console @echo command.
120 - Implement a Slot priority (loaded/not loaded).
121 - Implement "vacation" Incremental only saves.
122 - Implement single pane restore (much like the Gftp panes).
123 - Implement Automatic Mount even in operator wait.
124 - Implement create "FileSet"?
125 - Implement Release Device in the Job resource to unmount a drive.
126 - Implement Acquire Device in the Job resource to mount a drive,
127 be sure this works with admin jobs so that the user can get
128 prompted to insert the correct tape. Possibly some way to say to
129 run the job but don't save the files.
130 - Implement all command line args on run.
131 - Implement command line "restore" args.
132 - Implement "restore current select=no"
133 - Fix watchdog pthread crash on Win32 (this is pthread_kill() Cygwin bug)
134 - Implement "scratch pool" where tapes are defined and can be
135 taken by any pool that needs them.
136 - Implement restore "current system", but take all files without
137 doing selection tree -- so that jobs without File records can
139 - Implement disk spooling. Two parts: 1. Spool to disk then
140 immediately to tape to speed up tape operations. 2. Spool to
141 disk only when the tape is full, then when a tape is hung move
143 - Implement a relocatable bacula.spec
144 - Implement dump/print label to UA
145 - Add prefixlinks to where or not where absolute links to FD.
146 - Issue message to mount a new tape before the rewind.
147 - Simplified client job initiation for portables.
148 - If SD cannot open a drive, make it periodically retry.
149 - Implement LabelTemplate (at least first cut).
150 - Add more of the config info to the tape label.
151 - Implement a Mount Command and an Unmount Command where
152 the users could specify a system command to be performed
153 to do the mount, after which Bacula could attempt to
154 read the device. This is for Removeable media such as a CDROM.
155 - Most likely, this mount command would be invoked explicitly
156 by the user using the current Console "mount" and "unmount"
157 commands -- the Storage Daemon would do the right thing
158 depending on the exact nature of the device.
159 - As with tape drives, when Bacula wanted a new removable
160 disk mounted, it would unmount the old one, and send a message
161 to the user, who would then use "mount" as described above
162 once he had actually inserted the disk.
164 - Make some way so that if a machine is skipped because it is not up
165 that Bacula will continue retrying for a specified period of time --
167 - If tape is marked read-only, then try opening it read-only rather than
168 failing, and remember that it cannot be written.
169 - Refine SD waiting output:
170 Device is being positioned
171 > Device is being positioned for append
172 > Device is being positioned to file x
174 - Figure out some way to estimate output size and to avoid splitting
175 a backup across two Volumes -- this could be useful for writing CDROMs
176 where you really prefer not to have it split -- not serious.
177 - Add RunBeforeJob and RunAfterJob to the Client program.
178 - Have SD compute MD5 or SHA1 and compare to what FD computes.
179 - Make VolumeToCatalog calculate an MD5 or SHA1 from the
180 actual data on the Volume and compare it.
181 - Implement FileOptions (see end of this document)
182 - Implement Bacula plugins -- design API
183 - Make bcopy read through bad tape records.
184 - Fix read_record to handle multiple sessions.
185 - Program files (i.e. execute a program to read/write files).
186 Pass read date of last backup, size of file last time.
187 - Add Signature type to File DB record.
188 - CD into subdirectory when open()ing files for backup to
189 speed up things. Test with testfind().
190 - Priority job to go to top of list.
191 - Why are save/restore of device different sizes (sparse?) Yup! Fix it.
192 - Implement some way for the Console to dynamically create a job.
193 - Restore to a particular time -- e.g. before date, after date.
194 - Solaris -I on tar for include list
195 - Need a verbose mode in restore, perhaps to bsr.
196 - bscan without -v is too quiet -- perhaps show jobs.
197 - Add code to reject whole blocks if not wanted on restore.
198 - Check if we can increase Bacula FD priorty in Win2000
199 - Make sure the MaxVolFiles is fully implemented in SD
200 - Check if both CatalogFiles and UseCatalog are set to SD.
201 - Need return status on read_cb() from read_records(). Need multiple
202 records -- one per Job, maybe a JCR or some other structure with
203 a block and a record.
204 - Figure out how to do a bare metal Windows restore
205 - Possibly add email to Watchdog if drive is unmounted too
206 long and a job is waiting on the drive.
207 - Use read_record.c in SD code.
208 - Restore program that errors in SD due to no tape reports
209 OK incorrectly in output.
210 - After unmount, if restore job started, ask to mount.
211 - Convert all %x substitution variables, which are hard to remember
212 and read to %(variable-name). Idea from TMDA.
213 - Remove NextId for SQLite. Optimize.
214 - Move all SQL statements into a single location.
215 - Add UA rc and history files.
216 - put termcap (used by console) in ./configure and
217 allow -with-termcap-dir.
218 - Enhance time and size scanning routines.
219 - Fix Autoprune for Volumes to respect need for full save.
220 - Fix Win32 config file definition name on /install
221 - Compare tape to Client files (attributes, or attributes and data)
222 - Make all database Ids 64 bit.
223 - Write an applet for Linux.
224 - Add estimate to Console commands
225 - Implement new daemon communications protocol.
226 - Allow console commands to detach or run in background.
227 - Fix status delay on storage daemon during rewind.
228 - Add SD message variables to control operator wait time
229 - Maximum Operator Wait
230 - Minimum Message Interval
231 - Maximum Message Interval
232 - Send Operator message when cannot read tape label.
233 - Verify level=Volume (scan only), level=Data (compare of data to file).
234 Verify level=Catalog, level=InitCatalog
236 - Add keyword search to show command in Console.
237 - Events : tape has more than xxx bytes.
238 - Complete code in Bacula Resources -- this will permit
239 reading a new config file at any time.
240 - Handle ctl-c in Console
241 - Implement script driven addition of File daemon to config files.
242 - Think about how to make Bacula work better with File (non-tape) archives.
243 - Write Unix emulator for Windows.
244 - Put memory utilization in Status output of each daemon
245 if full status requested or if some level of debug on.
246 - Make database type selectable by .conf files i.e. at runtime
247 - Set flag for uname -a. Add to Volume label.
248 - Implement throttled work queue.
249 - Restore files modified after date
250 - Restore file modified before date
251 - Restore -- do nothing but show what would happen
252 - SET LD_RUN_PATH=$HOME/mysql/lib/mysql
253 - Implement Restore FileSet=
254 - Create a protocol.h and protocol.c where all protocol messages
256 - Remove duplicate fields from jcr (e.g. jcr.level and jcr.jr.Level, ...).
257 - Timout a job or terminate if link goes down, or reopen link and query.
258 - Concept of precious tapes (cannot be reused).
259 - Make bcopy copy with a single tape drive.
260 - Permit changing ownership during restore.
263 > My suggestion: Add a feature on the systray menu-icon menu to request
264 > an immediate backup now. This would be useful for laptop users who may
265 > not be on the network when the regular scheduled backup is run.
267 > My wife's suggestion: Add a setting to the win32 client to allow it to
268 > shut down the machine after backup is complete (after, of course,
269 > displaying a "System will shut down in one minute, click here to cancel"
270 > warning dialog). This would be useful for sites that want user
271 > woorkstations to be shut down overnight to save power.
274 - Autolabel should be specified by DIR instead of SD.
277 - AutoScan (check checksum of tape)
278 - Format command = "format /dev/nst0"
282 - Seek resolution (usually corresponds to buffer size)
283 - EODErrorCode=ENOSPC or code
284 - Partial Read error code
285 - Partial write error code
286 - Nonformatted read error
287 - Nonformatted write error
288 - WriteProtected error
292 - IgnoreCloseErrors=yes
302 - FD sends unsaved file list to Director at end of job (see
304 - File daemon should build list of files skipped, and then
305 at end of save retry and report any errors.
306 - Write a Storage daemon that uses pipes and
307 standard Unix programs to write to the tape.
309 - Need something that monitors the JCR queue and
310 times out jobs by asking the deamons where they are.
311 - Enhance Jmsg code to permit buffering and saving to disk.
312 - device driver = "xxxx" for drives.
314 - Ensure that /dev/null works
315 - Need report class for messages. Perhaps
316 report resource where report=group of messages
317 - enhance scan_attrib and rename scan_jobtype, and
318 fill in code for "since" option
319 - Director needs a time after which the report status is sent
320 anyway -- or better yet, a retry time for the job.
321 - Don't reschedule a job if previous incarnation is still running.
322 - Some way to automatically backup everything is needed????
323 - Need a structure for pending actions:
325 - termination status (part of buffered msgs?)
327 Read, Write, Clean, Delete
328 - Login to Bacula; Bacula users with different permissions:
329 owner, group, user, quotas
330 - Store info on each file system type (probably in the job header on tape.
331 This could be the output of df; or perhaps some sort of /etc/mtab record.
334 - Design at hierarchial storage for Bacula. Migration and Clone.
335 - Implement FSM (File System Modules).
336 - Audit M_ error codes to ensure they are correct and consistent.
337 - Add variable break characters to lex analyzer.
338 Either a bit mask or a string of chars so that
339 the caller can change the break characters.
340 - Make a single T_BREAK to replace T_COMMA, etc.
341 - Ensure that File daemon and Storage daemon can
342 continue a save if the Director goes down (this
343 is NOT currently the case). Must detect socket error,
344 buffer messages for later.
345 - Enhance time/duration input to allow multiple qualifiers e.g. 3d2h
346 - Add ability to backup to two Storage devices (two SD sessions) at
347 the same time -- e.g. onsite, offsite.
348 - Add the ability to consolidate old backup sets (basically do a restore
349 to tape and appropriately update the catalog). Compress Volume sets.
350 Might need to spool via file is only one drive is available.
351 - Compress or consolidate Volumes of old possibly deleted files. Perhaps
352 someway to do so with every volume that has less than x% valid
356 Migration: Move a backup from one Volume to another
357 Clone: Copy a backup -- two Volumes
359 Bacula Migration is based on Jobs (apparently Networker is file by file).
361 Migration triggered by:
365 Highwater mark (keep total size)
369 Bacula Projects Roadmap
371 last update 8 May 2003
373 Item 1: Multiple simultaneous Jobs. (done)
374 Done -- Restore part needs better implementation to work correctly
375 Also, it needs considerable testing
377 What: Permit multiple simultaneous jobs in Bacula.
379 Why: An enterprise level solution needs to go fast without the
380 need for the system administrator to carefully tweak
381 timing. Based on the benchmarks, during a full
382 backup, NetWorker typically hit 10 times the bandwidth to
383 the tape compared to Bacula--largely. This is probably due to
384 running parallel jobs and multi-threaded filling of buffers
385 and writing them to tape. This should also make things work
386 better when you have a mix of fast and slow machines backing
389 Notes: Bacula was designed to run multiple simultaneous jobs. Thus
390 implementing this is a matter of some small cleanups and
394 Item 2: Make the Storage daemon use intermediate file storage to buffer data.
395 Deferred -- not necessary yet -- possibly implement with Migration.
397 What: If data is coming into the SD too fast, buffer it to
398 disk if the user has configured this option.
400 Why: This would be nice, especially if it more or less falls out
401 when implementing (1) above. If not, it probably should not
402 be given a high priority because fundamentally the backup time
403 is limited by the tape bandwidth. Even though you may finish a
404 client job quicker by spilling to disk, you still have to
405 eventually get it onto tape. If intermediate disk buffering
406 allows us to improve write bandwidth to tape, it may make
409 Notes: Whether or not this is implemented will depend upon performance
410 testing after item 1 is implemented.
413 Item 3: Write the bscan program -- also write a bcopy program.
416 What: Write a program that reads a Bacula tape and puts all the
417 appropriate data into the catalog. This allows recovery
418 from a tape that is no longer in the database, or it allows
419 re-creation of a database if lost.
421 Why: This is a fundamental robustness and disaster recovery tool
422 which will increase the comfort level of a sysadmin
423 considering adopting Bacula.
425 Notes: A skeleton of this program already exists, but much work
426 needs to be done. Implementing this will also make apparent
427 any deficiencies in the current Bacula tape format.
430 Item 4: Implement Base jobs.
432 What: A base job is sort of like a Full save except that you
433 will want the FileSet to contain only files that are unlikely
434 to change in the future (i.e. a snapshot of most of your
435 system after installing it). After the base job has been run,
436 when you are doing a Full save, you can specify to exclude
437 all files saved by the base job that have not been modified.
439 Why: This is something none of the competition does, as far as we know
440 (except BackupPC, which is a Perl program that saves to disk
441 only). It is big win for the user, it makes Bacula stand out
442 as offering a unique optimization that immediately saves time
445 Notes: Big savings in tape usage. Will require more resources because
446 the e. DIR must send FD a list of files/attribs, and the FD must
447 search the list and compare it for each file to be saved.
450 Item 5: Implement Label templates
453 What: This is a mechanism whereby Bacula can automatically create
454 a tape label for new tapes according to a detailed specification
455 provided by the user.
457 Why: It is a major convenience item for folks who use automated label
460 Notes: Bacula already has a working form of automatic tape label
461 creation, but it is very crude. The design for the complete
462 tape labeling project is already documented in the manual.
465 Item 6: Write a regression script.
466 Done -- Continue to expand its testing.
468 What: This is an automatic script that runs and tests as many features
469 of Bacula as possible. The output is compared to previous
470 versions of Bacula and any differences are reported.
472 Why: This is an enormous help in preventing introduction of new
473 errors in parts of the program that already work correctly.
475 Notes: This probably should be ranked higher, it's something the typical
476 user doesn't see. Depending on how it's implemented, it may
477 make sense to defer it until the archival tape format and
478 user interface mature.
481 Item 7: GUI for interactive restore
482 Item 8: GUI for interactive backup
484 What: The current interactive restore is implemented with a tty
485 interface. It would be much nicer to be able to "see" the
486 list of files backed up in typical GUI tree format.
487 The same mechanism could also be used for creating
488 ad-hoc backup FileSets (item 8).
490 Why: Ease of use -- especially for the end user.
492 Notes: Rather than implementing in Gtk, we probably should go directly
493 for a Browser implementation, even if doing so meant the
494 capability wouldn't be available until much later. Not only
495 is there the question of Windows sites, most
496 Solaris/HP/IRIX, etc, shops can't currently run Gtk programs
497 without installing lots of stuff admins are very wary about.
498 Real sysadmins will always use the command line anyway, and
499 the user who's doing an interactive restore or backup of his
500 own files will in most cases be on a Windows machine running
504 Item 9: Add SSL to daemon communications.
506 What: This provides for secure communications between the daemons.
508 Why: This would allow doing backup across the Internet without
509 privacy concerns (or with much less concern).
511 Notes: The vast majority of near term potential users will be backing up
512 a single site over a LAN and, correctly or not, they probably
513 won't be concerned with security, at least not enough to go to
514 the trouble to set up keys, etc. to screw things down. We suspect
515 that many users genuinely interested in multi-site backup
516 already run some form of VPN software in their internetwork
517 connections, and are willing to delegate security to that layer.
520 Item 10: Define definitive tape format.
523 What: Define that definitive tape format that will not change
524 for the next millennium.
526 Why: Stability, security.
528 Notes: See notes for item 11 below.
531 Item 11: New daemon communication protocol.
533 What: The current daemon to daemon protocol is basically an ASCII
534 printf() and sending the buffer. On the receiving end, the
535 buffer is sscanf()ed to unpack it. The new scheme would
536 be a binary format that allows quick packing and unpacking
537 of any data type with named fields.
539 Why: Using binary packing would be faster. Named fields will permit
540 error checking to ensure that what is sent is what the
541 receiver really wants.
543 Notes: These are internal improvements in the interest of the
544 long-term stability and evolution of the program. On the one
545 hand, the sooner they're done, the less code we have to rip
546 up when the time comes to install them. On the other hand, they
547 don't bring an immediately perceptible benefit to potential
548 users. Item 10 and possibly item 11 should be deferred until Bacula
549 is well established with a growing user community more or
550 less happy with the feature set. At that time, it will make a
551 good "next generation" upgrade in the interest of data
557 ======================================================
559 It is somewhat like a Full save becomes an incremental since
560 the Base job (or jobs) plus other non-base files.
562 - New BaseFile table that contains:
563 JobId, BaseJobId, FileId (from Base).
564 i.e. for each base file that exists but is not saved because
565 it has not changed, the File daemon sends the JobId, BaseId,
566 and FileId back to the Director who creates the DB entry.
567 - To initiate a Base save, the Director sends the FD
568 the FileId, and full filename for each file in the Base.
569 - When the FD finds a Base file, he requests the Director to
570 send him the full File entry (stat packet plus MD5), or
571 conversely, the FD sends it to the Director and the Director
572 says yes or no. This can be quite rapid if the FileId is kept
573 by the FD for each Base Filename.
574 - It is probably better to have the comparison done by the FD
575 despite the fact that the File entry must be sent across the
577 - An alternative would be to send the FD the whole File entry
578 from the start. The disadvantage is that it requires a lot of
579 space. The advantage is that it requires less communications
581 - The Job record must be updated to indicate that one or more
583 - At end of Job, FD returns:
584 1. Count of base files/bytes not written to tape (i.e. matches)
585 2. Count of base file that were saved i.e. had changed.
586 - No tape record would be written for a Base file that matches, in the
587 same way that no tape record is written for Incremental jobs where
588 the file is not saved because it is unchanged.
589 - On a restore, all the Base file records must explicitly be
590 found from the BaseFile tape. I.e. for each Full save that is marked
591 to have one or more Base Jobs, search the BaseFile for all occurrences
593 - An optimization might be to make the BaseFile have:
599 This would avoid the need to explicitly fetch each File record for
600 the Base job. The Base Job record will be fetched to get the
601 VolSessionId and VolSessionTime.
602 =========================================================
606 =============================================================
608 Request For Comments For File Backup Options
611 Subject: File Backup Options
614 A few days ago, a Bacula user who is backing up to file volumes and
615 using compression asked if it was possible to suppress compressing
616 all .gz files since it was a waste of CPU time. Although Bacula
617 currently permits using different options (compression, ...) on
618 a directory by directory basis, it cannot do it on a file by
619 file basis, which is clearly what was desired.
621 Proposed Implementation:
622 To solve this problem, I propose the following:
624 - Add a new Director resource type called FileOptions.
626 - The FileOptions resource will have records for all
627 options that can currently be specified on the Include record
628 (in a FileSet). Examples below.
630 - The FileOptions resource will permit an exclude option as well
631 as a number of additional options.
633 - The heart of the FileOptions resource is the ability to
634 supply any number of ApplyTo records which specify POSIX
635 regular expressions. These ApplyTo regular expressions are
636 applied to the fully qualified filename (path and all). If
637 one matches, then the FileOptions will be used.
639 - When an ApplyTo specification matches an included file, the
640 options specified in the FileOptions resource will override
641 the default options specified on the Include record.
643 - Include records will be modified to permit referencing one or
644 more FileOptions resources. The FileOptions will be used
645 in the order listed on the Include record and the first
646 one that matches will be applied.
648 - Options (or specifications) currently supplied on the Include
649 record will be deprecated (i.e. removed in a later version a
650 year or so from now).
652 - The Exclude record will be deprecated as the same functionality
653 can be obtained by using an Exclude = yes in the FileOptions.
656 The following records can appear in the FileOptions resource. An
657 asterisk preceding the name indicates a feature not currently
661 - Compression= (GZIP, ...)
662 - Signature= (MD5, SHA1, ...)
664 - OneFs= (yes/no) - remain on one filesystem
665 - Recurse= (yes/no) - recurse into subdirectories
666 - Sparse= (yes/no) - do sparse file backup
667 - *Exclude= (yes/no) - exclude file from being saved
668 - *Reader= (filename) - external read (backup) program
669 - *Plugin= (filename) - read/write plugin module
672 - verify= (ipnougsamc5) - verify options
675 - replace= (always/ifnewer/ifolder/never) - replace options currently
677 - *Writer= (filename) - external write (restore) program
681 Currently options specifying compression, MD5 signatures, recursion,
682 ... of a FileSet are supplied on the Include record. These will now
683 all be collected into a FileOptions resource, which will be
684 specified on the Include in place of the options. Multiple FileOptions
685 may be specified. Since the FileOptions contain regular expressions
686 that are applied to the full filename, this will give the ability
687 to specify backup options on a file by file basis to whatever level
696 Include = compression=GZIP signature=MD5 {
705 Include = FileOptions=Opts {
716 That's a lot more to do the same thing, but it gives the ability to
717 apply options on a file by file basis. For example, suppose you
718 want to compress all files but not any file with extensions .gz or .Z.
719 You could do so as follows:
723 Include = FileOptions=NoCompress FileOptions=Opts {
731 ApplyTo = /*.?*/ # matches all files
736 # Note multiple ApplyTos are ORed
737 ApplyTo = /*.gz/ # matches .gz files */
738 ApplyTo = /*.Z/ # matches .Z files */
741 Now, since the NoCompress FileOptions is specified first on the
742 Include line, any *.gz or *.Z file will have an MD5 signature computed,
743 but will not be compressed. For all other files, the NoCompress will not
744 match, so the Opts options will be used which will include GZIP
748 - Is it necessary to provide some means of ANDing regular expressions
749 and negation? (not currently planned)
751 e.g. ApplyTo = /*.gz/ && !/big.gz/
753 - I see that Networker has a "null" module which, if specified, does not
754 backup the file, but does make an record of the file in the catalog
755 so that the catalog will reflect an exact picture of the filesystem.
756 The result is that the file can be "seen" when "browsing" the save
757 sets, but it cannot be restored.
759 Is this really useful? Should it be implemented in Bacula?
762 After implementing the above, the user will be able to specify
763 on a file by file basis (using regular expressions) what options are
764 applied for the backup.
767 =============================================
769 ==========================================================
771 For each Incremental job that is run, there may be files that
772 were found but not saved because they were locked (this applies
773 only to Windows). Such a system could send back to the Director
774 a list of Unsaved files.
776 - New UnSavedFiles table that contains:
780 - Then in the next Incremental job, the list of Unsaved Files will be
781 feed to the FD, who will ensure that they are explicitly chosen even
782 if standard date/time check would not have selected them.
783 =============================================================
786 Done: (see kernsdone for more)
787 - Heartbeat between daemons.
788 - Fix Dir heartbeat in restore and verify vol. Be sure to make
789 bnet_recv() ignore BNET_HEARTBEAT.
790 - Implement HEART_BEAT while SD waiting for tapes.
791 - Include RunBeforeJob and RunAfterJob output in the message
793 - Change M_INFO to M_RESTORED for all restored files.
794 - Fix command prompt in gnome-console by checking on Ready.
795 - Merge SQLite, MySQL, and Rel spec into a single file.
796 - Fix config of "console"
797 - Check if cancel works with FD (fixed).
798 - Properly configure console and gconsole (currently for source not
799 configured for installation).
800 - Error labeling tape from console gets Jmsg error because of no Job.
801 - Test and implement get_pint and get_yesno.
802 - Implement global with DB name and add to btraceback.gdb
803 - Remove subsysdir from conf files (used only in autostart scripts).
805 rufus-dir: Max configured use duration exceeded. Marking Volume "MatouBackup" as Used.
806 rufus-sd: Volume "" previously written, moving to end of data.
807 rufus-sd: Matou.2003-05-10_10.39.18 Error: I canot write on this volume because:
808 The number of files mismatch! Volume=1 Catalog=0
809 rufus-sd: Matou.2003-05-10_10.39.18 Error: askdir.c:155 NULL Volume name. This shouldn't happen!!!
810 - Shell character expansion is failing occassionally.
811 - Add a section to the doc on Manual cycling of Volumes.
812 - Check if Job/File retentions apply to multivolume jobs.
813 - Fix missing casette in autoloader during read:
814 14-May-2003 14:41 undef-sd: RestoreFiles.2003-05-14_14.41.00 Warning: acquire.c:106 Volume name mismatch. Wanted TestVolume0005 got TestVolume0010
815 14-May-2003 14:41 undef-sd: 3301 Issuing autochanger "loaded" command.
816 14-May-2003 14:41 undef-sd: 3302 Issuing autochanger "unload" command.
817 14-May-2003 14:42 undef-sd: 3303 Issuing autochanger "load slot 1" command.
818 14-May-2003 14:42 undef-sd: 3304 Autochanger "load slot 1" status is OK.
819 14-May-2003 14:42 undef-sd: RestoreFiles.2003-05-14_14.41.00 Warning: acquire.c:106 Volume name mismatch. Wanted TestVolume0005 got TestVolume0009
820 14-May-2003 14:42 undef-sd: 3301 Issuing autochanger "loaded" command.
821 14-May-2003 14:42 undef-sd: RestoreFiles.2003-05-14_14.41.00 Warning: acquire.c:106 Volume name mismatch. Wanted TestVolume0005 got TestVolume0009
822 14-May-2003 14:42 undef-sd: 3301 Issuing autochanger "loaded" command.
823 14-May-2003 14:42 undef-sd: RestoreFiles.2003-05-14_14.41.00 Warning: acquire.c:106 Volume name mismatch. Wanted TestVolume0005 got TestVolume0009
824 14-May-2003 14:42 undef-sd: 3301 Issuing autochanger "loaded" command.
825 14-May-2003 14:42 undef-sd: RestoreFiles.2003-05-14_14.41.00 Warning: acquire.c:106 Volume name mismatch. Wanted TestVolume0005 got TestVolume0009
826 14-May-2003 14:42 undef-sd: 3301 Issuing autochanger "loaded" command.
827 14-May-2003 14:42 undef-sd: RestoreFiles.2003-05-14_14.41.00 Fatal error: acquire.c:129 Too many errors trying to mount device "/dev/nrsa0".
828 14-May-2003 14:42 undef-dir: Bacula 1.31 (12May03): 14-May-2003 14:42
829 - Fix problem reported by Christopher McCurdy <cmccurdy@eecis.udel.edu>
830 xeon-fd: Could not stat c:/Documents and Settings/All
831 Users/Application Data/Humc:\Documents and Settings\All User98_AIX.kbf:
832 ERR=No such file or directory
834 - The following Re-read last block at EOT failed. ERR=block.c:523 Read zero bytes on device /dev/nrsa0.
835 undef-sd: block.c:523 Read zero bytes on device /dev/nrsa0.
836 apparently masks the standard EOM message.
837 - BSD (probably) does not have strtoll()
838 - BSD does not have ioctl() MTEOM
839 - BSD defines a number of MT_xxx variables which conflict
840 with those defined by Bacula.
841 - Make default duration days if no qualifier (e.g. s) is specified.
842 - BSDI fix finding gcc version
843 - When the FD errs (e.g. disk full) have a more graceful shutdown.
844 - Make sure Bacula prunes/purges canceled and failed jobs too and all
845 jobs with zero JobFiles.
846 - Implement Volume name checking.
847 - Document what characters can go into Volume names.
848 - Getting the following on all directories on Win32
849 19-May-2003 01:14 tibs-fd: Could not access c:/cygwin/home/kern/rxvt: ERR=Permission denied
850 - Cancellation caused JobMedia error:
851 babylon5-dir: Last FULL backup time not found. Doing FULL backup.
852 babylon5-dir: Start Backup JobId 416, Job=Zocalo_Save.2003-05-19_02.15.06
853 babylon5-sd: End of media on Volume VXA-V17-Inc-001 Bytes=31,982,900,672 Blocks=495,781.
854 babylon5-sd: Job Zocalo_Save.2003-05-19_02.15.06 waiting. Cannot find any appendable volumes.
855 babylon5-sd: Someone woke me up, but I cannot find any appendable volumes
856 for Job=Zocalo_Save.2003-05-19_02.15.06.
857 babylon5-sd: Zocalo_Save.2003-05-19_02.15.06 Fatal error: Job
858 Zocalo_Save.2003-05-19_02.15.06 canceled while waiting for mount on
859 Storage Device "Ecrix_VXA-1".
860 babylon5-sd: Zocalo_Save.2003-05-19_02.15.06 Fatal error: Cannot fixup device
861 error. Job Zocalo_Save.2003-05-19_02.15.06 canceled while waiting for
862 mount on Storage Device "Ecrix_VXA-1".
864 babylon5-dir: Zocalo_Save.2003-05-19_02.15.06 Error: Catalog error creating
865 JobMedia record. sql_create.c:125 Create JobMedia failed. Record already
867 babylon5-sd: Zocalo_Save.2003-05-19_02.15.06 Error: Error creating JobMedia
868 record: 1991 Update JobMedia error
870 babylon5-sd: Zocalo_Save.2003-05-19_02.15.06 Error: askdir.c:158 NULL Volume
871 name. This shouldn't happen!!!
872 zocalo-fd: Zocalo_Save.2003-05-19_02.15.06 Error: bnet.c:310 Write error
873 sending to Storage daemon:babylon5:9103: ERR=Broken pipe
874 - Volume names with spaces get jammed into the catalog with 0x1
875 i.e. the SD bashes the Volume but they are not unbased by Dir.
876 jerom-dir: MonthlySave.2003-05-10_17.12.01 Error: Unable to get Media
877 record for Volume Tape^A1: ERR=sql_get.c:788 Media record for Volume
879 jerom-sd: MonthlySave.2003-05-10_17.12.01 Error: Error updating Volume
880 Info: 1991 Catalog Request failed: sql_get.c:788 Media record for Volume
882 - ChangeServiceConfig2A does not exist on WinNT (ADVAPI32.DLL).
883 - Fix "access not allowed" for backup of files on WinXP.
884 - Check for existence of all new Win32 API's. See LoadLibrary in
886 - Count errors during restore and print them in the Job report.
887 - Bug: fix access problems on files restored on WinXP.
888 - Put system type returned by FD into catalog.
889 - Finish WIN32_DATA stream code (bextract, check if can handle stream)
890 - Make SD keep track of Files, Bytes during restore.
891 - If you enter the userid by hand for restore, you get:
892 Enter JobId(s), comma separated, to restore: 74
893 You have selected the following JobId: 74
894 Building directory tree for JobId 74 ...
895 134645140 items inserted into the tree and marked for extraction.
896 - Add SDWriteSeqNo to SD, and probably Read on FD side.
897 - If bootstrap is non-zero for restore, do not show JobId in the
898 OK to run? (yes/mod/no): list.
899 - When all cassettes in magazine are used, got:
900 22-May-2003 18:24 undef-sd: 3304 Autochanger "load slot 1" status is OK.
901 22-May-2003 18:24 undef-sd: NightlySave.2003-05-22_14.08.16 Warning: mount.c:245 Director wanted Volume "TestVolume0009".
902 Current Volume "TestVolume0005" not acceptable because:
903 1998 Volume "TestVolume0005" not Append or Recycle.
904 22-May-2003 18:24 undef-sd: NightlySave.2003-05-22_14.08.16 Error: Autochanger Volume "TestVolume0009" not found in slot 1.
905 Setting slot to zero in catalog.
906 22-May-2003 18:24 undef-sd: Please mount Volume "TestVolume0009" on Storage Device "ARCHIVE 4586" for Job NightlySave.2003-05-22_14
908 Use "mount" command to release Job.
909 22-May-2003 19:24 undef-sd: Please mount Volume "TestVolume0009" on Storage Device "ARCHIVE 4586" for Job NightlySave.2003-05-22_14
911 Use "mount" command to release Job.
912 - Don't zero the Slot when the wrong volume is found -- simply ask
914 - Implement MTIOCERRSTAT on FreeBSD to clear tape error conditions.
915 - Shell expansion fails for working_directory in SD from time to time.
916 - File the Automatically selected: xxx
917 to say Automatically selected Pool: xxx
918 - Default duration with no qualifier is sec should be 1 day
919 - zap sd_auth_key in SD after FD connection.
920 - Find a solution for the multiple FileSet problem (when it is changed). Add date?
921 - Look at Python for a Bacula scripting language -- www.python.org
922 - When Marking a file in Restore that is a hard link, also
923 mark the link so that the data will be reloaded.
924 - Emergency restore info:
926 - Backup working directory
928 - Why don't we get an error message from Win32 FD when bootstrap
929 file cannot be created for restore command?
930 - Fix Win2000 error with no messages during startup.
931 - Make restore more robust in counting error and not immediately bailing
932 out. Also print error message once, but try to continue.
933 - Add code to check that blocks are sequential on restore.
934 - Remove "rufus" and such references from regress.
935 - No READLINE_SRC if found in alternate directory.
936 - If ./btape is called without /dev, assume argument is a Storage resource name.
937 - Find general solution for sscanf size problems (as well as sprintf. Do at run time?
938 - Bytes restored is wrong.
939 - The "List last 20 Jobs run" doesnt work correctly in restore.
940 It doesnt show the last 20 jobs , but some older ones.
941 - Fix Verify VolumeToCatalog to use BSRs -- it is broken.
942 - Implement Release Storage=xxx
943 - Fix restore on Win95/98
944 - Remove the Jmsg() in sql_find.c:102 or only print on hard error.
945 - Implement FileSet VolIndex -- done, but must update old records.
946 - Check this below from Phil.
947 This was SD reported data rather than FD data!
948 > When the job was done, Bacula reported 11084 files restored:
951 > Job: Zocalo_Restore.2003-06-05_16.42.01
953 > Start time: 05-Jun-2003 16:42
954 > End time: 06-Jun-2003 01:21
955 > Files Restored: 11,084
956 > Bytes Restored: 65,474,772
958 > FD termination status: OK
959 > Termination: Restore OK
961 > when it should probably have reported 11084 files scanned, 250 restored.
962 > The bytes restored count looks about right.
964 - Should Bacula make an Append tape as Purged when purging?
965 - Use switch() in backup.c and restore.c in FD instead of giant if statement.
966 - If during a restore, a hard linked file already exists (on option), delete
967 the file and re-link it. This is to avoid the possibility that the
968 user had re-linked the file between the backup and the restore.
969 Do lstat() to see if it is already properly linked.
970 Same for symlinked file.
971 Make sure ifnewer, ifolder, never, ... apply correctly.
972 - Flag so that no connect does not error, and Reschedule a job.
973 - Implement "Reschedule OnError=yes interval=nnn times=xxx"
974 - That restoring a hard link that already exists works correctly.
976 - Make Pool resource handle Counter resources.
977 - Fix first block number after label to be zero instead of 1 (reset after label).
978 - Grep for Backup OK in regression script.
979 - Do NOT reuse same JobId if tape written.
980 - Implement non-blocking writes and bsock->terminate in heartbeat
981 thread, or set it in status.c cancel (used pthread_kill() instead of
983 - Add restore to specific date.
984 - Instrument use_count on DEVICE packets and ensure that the device is
985 being close()ed at the appropriate time.
986 - Test long path names (>64 chars) in Windows -- crashes FD?
987 - Implement fast block rejection: match_bsr_block().
988 - Complain if record dropped in bnet_recv because too long.
989 - Test multiple simultaneous Volumes
990 - Document recycling algorithm.
991 - Make Restore report an error if FD or SD term codes are not OK.
992 - To link with mysqlclient_r may require -lssl -lcrypto
993 - Document Heart beat code
994 - Non-fatal errors are not correct counting attribs.c:277
995 - Check that Block number in JobMedia are correct.
996 - The bsr for Dan's job has file indexes covering the whole range rather
997 than only the range contained on the volume.
998 Constrain FileIndex to be within range for Volume.
999 - Pass prefix_links to FD.
1000 - Fix restore list of volumes if Volume not selected.
1001 - Complete (or turn off) the command line code in update.
1002 - Testing Tibs job erred and hung director on Storage resource. This was
1003 because there were a whole pile of jobs hanging around in the SD
1004 waiting for a connection from the FD that was never coming.
1005 - Possibly update all client records at startup.
1006 - Listing of Win32 restored files has ??????? for user.
1007 - In Win portable restore the directory is not create
1008 27-Jun-2003 16:52 tibs-fd: kernsrestore.2003-06-27_16.52.20 Error:
1009 create_file.c:175 Could not create
1010 /tmp/bacula-restores/cygwin/home/kern/bacula/k/src/dird/dird_conf.o: 0
1011 ERR=The system cannot find the path specified.
1012 - Finish Windows implementation (add setting of correct type on restore,
1013 add Portable Data Format flag).
1014 - Document portable=yes
1015 - Document c:/working directory better than /working directory.
1016 - Document run "yes".
1017 - Document update volume: jobid, current, before, all
1018 - Document that bscan does not work with multiple simultaneous jobs.
1019 - Update Automatic Volume Labeling in disk.wml
1020 - Figure out how to handle DHCP IP addresses -- use dynamic DNS.
1021 - Add OS, GCC version to traceback output.
1022 - Implement a Recycle or Prune Current Volume (as from Nic Bellamy),
1023 possibly implement Prune Oldest Volume.
1024 - Backup on havana gave:
1025 Total Bytes Read=4,629,249,084,728,272,628 Blocks Read=0 Bytes/block=134,735,604
1026 - Move JobFiles and JobBytes to SD rather than FD -- more correct.
1027 - Test of last block is correct in JobMedia when splitting file
1029 - Doc win32 table from Marc Brueckner der Megauser
1030 - Fix default time in seconds without qualifier -- bacula-dir.conf
1031 - Add Nic Bellamy's backup scheme.
1032 - Document using multiple Pools for daily rotating tapes.
1033 - Check if Incremental is working correctly when it looks for the previous Job