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 recycling algorithm.
22 Testing to do: (painful)
23 - that ALL console command line options work and are always implemented
24 - blocksize recognition code.
25 - multiple simultaneous Volumes
26 - Test if rewind at end of tape waits for tape to rewind.
28 - Test not zeroing Autochanger slot when it is wrong.
29 - Test multiple simultaneous Volumes
30 - Test of last block is correct in JobMedia when splitting file
32 - Test recycling and purging (code changed in db_find_next_volume and
34 - Figure out how to use ssh or stunnel to protect Bacula communications.
37 - Check that Block number in JobMedia are correct.
38 - Complain if record dropped in bnet_recv because too long.
39 - Suppress Job Name in Jmsg for console
40 - Check Jmsg in bnet, may not work, must dup bsock.
41 - Use SD returned values in report -- or add them.
42 - Increment DB version prior to releasing.
43 - Turn off FULL_DEBUG prior to releasing.
44 - Implement forward spacing block/file: position_device(bsr) --
45 just before read_block_from_device();
46 - On Windows with very long path names, it may be impossible to create
47 a file (and thus restore it) because the total length is too long.
48 We must cd into the directory then create the file without the
50 - Move JobFiles and JobBytes to SD rather than FD -- more correct.
51 - Add client name to cram-md5 challenge so Director can immediately
52 verify if it is the correct client.
53 - Use runbeforejob to unload, then reload a volume previously used,
54 then the next job run gets an error reading the drive.
55 - lstat() is not going to work on Win32 for testing date.
56 - Implement a Recycle command
57 - Something is not right in last block of fill command.
58 - Implement List Volume Job=xxx or List scheduled volumes or
60 - Check if Incremental is working correctly when it looks for the previous Job
62 - Add next Volume to be used to status output.
63 - Make bootstrap filename unique.
64 - The bsr for Dan's job has file indexes covering the whole range rather
65 than only the range contained on the volume.
66 Constrain FileIndex to be within range for Volume.
67 - Test a second language e.g. french.
68 - Start working on Base jobs.
69 - Make "make binary-release" work from any directory.
70 - Document c:/working directory better than /working directory.
71 - Unsaved Flag in Job record.
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 - Make | and < work on FD side.
80 - Pass prefix_links to FD.
81 - Implement a M_SECURITY message class.
82 - From Phil Stracchino:
83 It would probably be a per-client option, and would be called
84 something like, say, "Automatically purge obsoleted jobs". What it
85 would do is, when you successfully complete a Differential backup of a
86 client, it would automatically purge all Incremental backups for that
87 client that are rendered redundant by that Differential. Likewise,
88 when a Full backup on a client completed, it would automatically purge
89 all Differential and Incremental jobs obsoleted by that Full backup.
90 This would let people minimize the number of tapes they're keeping on
91 hand without having to master the art of retention times.
92 - Prohibit backing up archive device (findlib/find_one.c:128)
93 - Make Restore report an error if FD or SD term codes are not OK.
94 - Add JobLevel in FD status (but make sure it is defined).
95 - Restrict characters permitted in a Resource name, and don't permit
97 - Implement new serialize subroutines
98 send(socket, "string", &Vol, "uint32", &i, NULL)
99 - Audit all UA commands to ensure that we always prompt where possible.
100 - Scratch Pool where the volumes can be re-assigned to any Pool.
104 - When doing a Backup send all attributes back to the Director, who
105 would then figure out what files have been deleted.
106 - Currently in mount.c:236 the SD simply creates a Volume. It should have
107 explicit permission to do so. It should also mark the tape in error
108 if there is an error.
109 - Make sure all restore counters are working correctly in the FD.
110 - SD Bytes Read is wrong.
111 - Configure mtx-changer to have correct path to mtx.
112 - Look at ALL higher level routines that call block.c to be sure
113 they don't expect something in errmsg.
114 - Investigate doing RAW backup of Win32 partition.
115 - Add JobName= to VerifyToCatalog so that all verifies can be done at the end.
116 - Add thread specific data to hold the jcr -- send error messages from
117 low level routines by accessing it and using Jmsg().
118 - Cancel waiting for Client connect in SD if FD goes away.
119 - Testing Tibs job erred and hung director on Storage resource. This was
120 because there were a whole pile of jobs hanging around in the SD
121 waiting for a connection from the FD that was never coming.
122 - Possibly update all client records at startup.
124 - Add Progress command that periodically reports the progress of
126 - One block was orphaned in the SD probably after cancel.
127 - Add all command line arguments to "update", e.g. slot=nn volStatus=append, ...
130 - Examine Bare Metal restore problem (a FD crash exists somewhere ...).
131 - Implement timeout in response() when it should come quickly.
132 - Implement console @echo command.
133 - Implement a Slot priority (loaded/not loaded).
134 - Implement "vacation" Incremental only saves.
135 - Implement single pane restore (much like the Gftp panes).
136 - Implement Automatic Mount even in operator wait.
137 - Implement create "FileSet"?
138 - Implement Release Device in the Job resource to unmount a drive.
139 - Implement Acquire Device in the Job resource to mount a drive,
140 be sure this works with admin jobs so that the user can get
141 prompted to insert the correct tape. Possibly some way to say to
142 run the job but don't save the files.
143 - Implement all command line args on run.
144 - Implement command line "restore" args.
145 - Implement "restore current select=no"
146 - Fix watchdog pthread crash on Win32 (this is pthread_kill() Cygwin bug)
147 - Implement "scratch pool" where tapes are defined and can be
148 taken by any pool that needs them.
149 - Implement restore "current system", but take all files without
150 doing selection tree -- so that jobs without File records can
152 - Implement disk spooling. Two parts: 1. Spool to disk then
153 immediately to tape to speed up tape operations. 2. Spool to
154 disk only when the tape is full, then when a tape is hung move
156 - Implement a relocatable bacula.spec
157 - Allow multiple Storage specifications (or multiple names on
158 a single Storage specification) in the Job record. Thus a job
159 can be backed up to a number of storage devices.
160 - Implement dump/print label to UA
161 - Add prefixlinks to where or not where absolute links to FD.
162 - Issue message to mount a new tape before the rewind.
163 - Simplified client job initiation for portables.
164 - If SD cannot open a drive, make it periodically retry.
165 - Implement LabelTemplate (at least first cut).
166 - Add more of the config info to the tape label.
167 - Implement a Mount Command and an Unmount Command where
168 the users could specify a system command to be performed
169 to do the mount, after which Bacula could attempt to
170 read the device. This is for Removeable media such as a CDROM.
171 - Most likely, this mount command would be invoked explicitly
172 by the user using the current Console "mount" and "unmount"
173 commands -- the Storage Daemon would do the right thing
174 depending on the exact nature of the device.
175 - As with tape drives, when Bacula wanted a new removable
176 disk mounted, it would unmount the old one, and send a message
177 to the user, who would then use "mount" as described above
178 once he had actually inserted the disk.
180 - Make some way so that if a machine is skipped because it is not up
181 that Bacula will continue retrying for a specified period of time --
183 - If tape is marked read-only, then try opening it read-only rather than
184 failing, and remember that it cannot be written.
185 - Refine SD waiting output:
186 Device is being positioned
187 > Device is being positioned for append
188 > Device is being positioned to file x
190 - Figure out some way to estimate output size and to avoid splitting
191 a backup across two Volumes -- this could be useful for writing CDROMs
192 where you really prefer not to have it split -- not serious.
193 - Add RunBeforeJob and RunAfterJob to the Client program.
194 - Have SD compute MD5 or SHA1 and compare to what FD computes.
195 - Make VolumeToCatalog calculate an MD5 or SHA1 from the
196 actual data on the Volume and compare it.
197 - Implement FileOptions (see end of this document)
198 - Implement Bacula plugins -- design API
199 - Make bcopy read through bad tape records.
200 - Fix read_record to handle multiple sessions.
201 - Program files (i.e. execute a program to read/write files).
202 Pass read date of last backup, size of file last time.
203 - Add Signature type to File DB record.
204 - CD into subdirectory when open()ing files for backup to
205 speed up things. Test with testfind().
206 - Priority job to go to top of list.
207 - Why are save/restore of device different sizes (sparse?) Yup! Fix it.
208 - Implement some way for the Console to dynamically create a job.
209 - Restore to a particular time -- e.g. before date, after date.
210 - Solaris -I on tar for include list
211 - Need a verbose mode in restore, perhaps to bsr.
212 - bscan without -v is too quiet -- perhaps show jobs.
213 - Add code to reject whole blocks if not wanted on restore.
214 - Check if we can increase Bacula FD priorty in Win2000
215 - Make sure the MaxVolFiles is fully implemented in SD
216 - Check if both CatalogFiles and UseCatalog are set to SD.
217 - Need return status on read_cb() from read_records(). Need multiple
218 records -- one per Job, maybe a JCR or some other structure with
219 a block and a record.
220 - Figure out how to do a bare metal Windows restore
221 - Possibly add email to Watchdog if drive is unmounted too
222 long and a job is waiting on the drive.
223 - Use read_record.c in SD code.
224 - Restore program that errors in SD due to no tape reports
225 OK incorrectly in output.
226 - After unmount, if restore job started, ask to mount.
227 - Convert all %x substitution variables, which are hard to remember
228 and read to %(variable-name). Idea from TMDA.
229 - Remove NextId for SQLite. Optimize.
230 - Move all SQL statements into a single location.
231 - Add UA rc and history files.
232 - put termcap (used by console) in ./configure and
233 allow -with-termcap-dir.
234 - Enhance time and size scanning routines.
235 - Fix Autoprune for Volumes to respect need for full save.
236 - Fix Win32 config file definition name on /install
237 - Compare tape to Client files (attributes, or attributes and data)
238 - Make all database Ids 64 bit.
239 - Write an applet for Linux.
240 - Add estimate to Console commands
241 - Implement new daemon communications protocol.
242 - Allow console commands to detach or run in background.
243 - Fix status delay on storage daemon during rewind.
244 - Add SD message variables to control operator wait time
245 - Maximum Operator Wait
246 - Minimum Message Interval
247 - Maximum Message Interval
248 - Send Operator message when cannot read tape label.
249 - Verify level=Volume (scan only), level=Data (compare of data to file).
250 Verify level=Catalog, level=InitCatalog
252 - Add keyword search to show command in Console.
253 - Events : tape has more than xxx bytes.
254 - Complete code in Bacula Resources -- this will permit
255 reading a new config file at any time.
256 - Handle ctl-c in Console
257 - Implement script driven addition of File daemon to config files.
258 - Think about how to make Bacula work better with File (non-tape) archives.
259 - Write Unix emulator for Windows.
260 - Put memory utilization in Status output of each daemon
261 if full status requested or if some level of debug on.
262 - Make database type selectable by .conf files i.e. at runtime
263 - Set flag for uname -a. Add to Volume label.
264 - Implement throttled work queue.
265 - Restore files modified after date
266 - Restore file modified before date
267 - Restore -- do nothing but show what would happen
268 - SET LD_RUN_PATH=$HOME/mysql/lib/mysql
269 - Implement Restore FileSet=
270 - Create a protocol.h and protocol.c where all protocol messages
272 - Remove duplicate fields from jcr (e.g. jcr.level and jcr.jr.Level, ...).
273 - Timout a job or terminate if link goes down, or reopen link and query.
274 - Concept of precious tapes (cannot be reused).
275 - Make bcopy copy with a single tape drive.
276 - Permit changing ownership during restore.
279 > My suggestion: Add a feature on the systray menu-icon menu to request
280 > an immediate backup now. This would be useful for laptop users who may
281 > not be on the network when the regular scheduled backup is run.
283 > My wife's suggestion: Add a setting to the win32 client to allow it to
284 > shut down the machine after backup is complete (after, of course,
285 > displaying a "System will shut down in one minute, click here to cancel"
286 > warning dialog). This would be useful for sites that want user
287 > woorkstations to be shut down overnight to save power.
290 - Autolabel should be specified by DIR instead of SD.
293 - AutoScan (check checksum of tape)
294 - Format command = "format /dev/nst0"
298 - Seek resolution (usually corresponds to buffer size)
299 - EODErrorCode=ENOSPC or code
300 - Partial Read error code
301 - Partial write error code
302 - Nonformatted read error
303 - Nonformatted write error
304 - WriteProtected error
308 - IgnoreCloseErrors=yes
318 - FD sends unsaved file list to Director at end of job (see
320 - File daemon should build list of files skipped, and then
321 at end of save retry and report any errors.
322 - Write a Storage daemon that uses pipes and
323 standard Unix programs to write to the tape.
325 - Need something that monitors the JCR queue and
326 times out jobs by asking the deamons where they are.
327 - Enhance Jmsg code to permit buffering and saving to disk.
328 - device driver = "xxxx" for drives.
330 - Ensure that /dev/null works
331 - Need report class for messages. Perhaps
332 report resource where report=group of messages
333 - enhance scan_attrib and rename scan_jobtype, and
334 fill in code for "since" option
335 - Director needs a time after which the report status is sent
336 anyway -- or better yet, a retry time for the job.
337 - Don't reschedule a job if previous incarnation is still running.
338 - Some way to automatically backup everything is needed????
339 - Need a structure for pending actions:
341 - termination status (part of buffered msgs?)
343 Read, Write, Clean, Delete
344 - Login to Bacula; Bacula users with different permissions:
345 owner, group, user, quotas
346 - Store info on each file system type (probably in the job header on tape.
347 This could be the output of df; or perhaps some sort of /etc/mtab record.
350 - Design at hierarchial storage for Bacula. Migration and Clone.
351 - Implement FSM (File System Modules).
352 - Audit M_ error codes to ensure they are correct and consistent.
353 - Add variable break characters to lex analyzer.
354 Either a bit mask or a string of chars so that
355 the caller can change the break characters.
356 - Make a single T_BREAK to replace T_COMMA, etc.
357 - Ensure that File daemon and Storage daemon can
358 continue a save if the Director goes down (this
359 is NOT currently the case). Must detect socket error,
360 buffer messages for later.
361 - Enhance time/duration input to allow multiple qualifiers e.g. 3d2h
362 - Add ability to backup to two Storage devices (two SD sessions) at
363 the same time -- e.g. onsite, offsite.
364 - Add the ability to consolidate old backup sets (basically do a restore
365 to tape and appropriately update the catalog). Compress Volume sets.
366 Might need to spool via file is only one drive is available.
367 - Compress or consolidate Volumes of old possibly deleted files. Perhaps
368 someway to do so with every volume that has less than x% valid
372 Migration: Move a backup from one Volume to another
373 Clone: Copy a backup -- two Volumes
375 Bacula Migration is based on Jobs (apparently Networker is file by file).
377 Migration triggered by:
381 Highwater mark (keep total size)
385 Bacula Projects Roadmap
387 last update 8 May 2003
389 Item 1: Multiple simultaneous Jobs. (done)
390 Done -- Restore part needs better implementation to work correctly
391 Also, it needs considerable testing
393 What: Permit multiple simultaneous jobs in Bacula.
395 Why: An enterprise level solution needs to go fast without the
396 need for the system administrator to carefully tweak
397 timing. Based on the benchmarks, during a full
398 backup, NetWorker typically hit 10 times the bandwidth to
399 the tape compared to Bacula--largely. This is probably due to
400 running parallel jobs and multi-threaded filling of buffers
401 and writing them to tape. This should also make things work
402 better when you have a mix of fast and slow machines backing
405 Notes: Bacula was designed to run multiple simultaneous jobs. Thus
406 implementing this is a matter of some small cleanups and
410 Item 2: Make the Storage daemon use intermediate file storage to buffer data.
411 Deferred -- not necessary yet -- possibly implement with Migration.
413 What: If data is coming into the SD too fast, buffer it to
414 disk if the user has configured this option.
416 Why: This would be nice, especially if it more or less falls out
417 when implementing (1) above. If not, it probably should not
418 be given a high priority because fundamentally the backup time
419 is limited by the tape bandwidth. Even though you may finish a
420 client job quicker by spilling to disk, you still have to
421 eventually get it onto tape. If intermediate disk buffering
422 allows us to improve write bandwidth to tape, it may make
425 Notes: Whether or not this is implemented will depend upon performance
426 testing after item 1 is implemented.
429 Item 3: Write the bscan program -- also write a bcopy program.
432 What: Write a program that reads a Bacula tape and puts all the
433 appropriate data into the catalog. This allows recovery
434 from a tape that is no longer in the database, or it allows
435 re-creation of a database if lost.
437 Why: This is a fundamental robustness and disaster recovery tool
438 which will increase the comfort level of a sysadmin
439 considering adopting Bacula.
441 Notes: A skeleton of this program already exists, but much work
442 needs to be done. Implementing this will also make apparent
443 any deficiencies in the current Bacula tape format.
446 Item 4: Implement Base jobs.
448 What: A base job is sort of like a Full save except that you
449 will want the FileSet to contain only files that are unlikely
450 to change in the future (i.e. a snapshot of most of your
451 system after installing it). After the base job has been run,
452 when you are doing a Full save, you can specify to exclude
453 all files saved by the base job that have not been modified.
455 Why: This is something none of the competition does, as far as we know
456 (except BackupPC, which is a Perl program that saves to disk
457 only). It is big win for the user, it makes Bacula stand out
458 as offering a unique optimization that immediately saves time
461 Notes: Big savings in tape usage. Will require more resources because
462 the e. DIR must send FD a list of files/attribs, and the FD must
463 search the list and compare it for each file to be saved.
466 Item 5: Implement Label templates
469 What: This is a mechanism whereby Bacula can automatically create
470 a tape label for new tapes according to a detailed specification
471 provided by the user.
473 Why: It is a major convenience item for folks who use automated label
476 Notes: Bacula already has a working form of automatic tape label
477 creation, but it is very crude. The design for the complete
478 tape labeling project is already documented in the manual.
481 Item 6: Write a regression script.
482 Done -- Continue to expand its testing.
484 What: This is an automatic script that runs and tests as many features
485 of Bacula as possible. The output is compared to previous
486 versions of Bacula and any differences are reported.
488 Why: This is an enormous help in preventing introduction of new
489 errors in parts of the program that already work correctly.
491 Notes: This probably should be ranked higher, it's something the typical
492 user doesn't see. Depending on how it's implemented, it may
493 make sense to defer it until the archival tape format and
494 user interface mature.
497 Item 7: GUI for interactive restore
498 Item 8: GUI for interactive backup
500 What: The current interactive restore is implemented with a tty
501 interface. It would be much nicer to be able to "see" the
502 list of files backed up in typical GUI tree format.
503 The same mechanism could also be used for creating
504 ad-hoc backup FileSets (item 8).
506 Why: Ease of use -- especially for the end user.
508 Notes: Rather than implementing in Gtk, we probably should go directly
509 for a Browser implementation, even if doing so meant the
510 capability wouldn't be available until much later. Not only
511 is there the question of Windows sites, most
512 Solaris/HP/IRIX, etc, shops can't currently run Gtk programs
513 without installing lots of stuff admins are very wary about.
514 Real sysadmins will always use the command line anyway, and
515 the user who's doing an interactive restore or backup of his
516 own files will in most cases be on a Windows machine running
520 Item 9: Add SSL to daemon communications.
522 What: This provides for secure communications between the daemons.
524 Why: This would allow doing backup across the Internet without
525 privacy concerns (or with much less concern).
527 Notes: The vast majority of near term potential users will be backing up
528 a single site over a LAN and, correctly or not, they probably
529 won't be concerned with security, at least not enough to go to
530 the trouble to set up keys, etc. to screw things down. We suspect
531 that many users genuinely interested in multi-site backup
532 already run some form of VPN software in their internetwork
533 connections, and are willing to delegate security to that layer.
536 Item 10: Define definitive tape format.
539 What: Define that definitive tape format that will not change
540 for the next millennium.
542 Why: Stability, security.
544 Notes: See notes for item 11 below.
547 Item 11: New daemon communication protocol.
549 What: The current daemon to daemon protocol is basically an ASCII
550 printf() and sending the buffer. On the receiving end, the
551 buffer is sscanf()ed to unpack it. The new scheme would
552 be a binary format that allows quick packing and unpacking
553 of any data type with named fields.
555 Why: Using binary packing would be faster. Named fields will permit
556 error checking to ensure that what is sent is what the
557 receiver really wants.
559 Notes: These are internal improvements in the interest of the
560 long-term stability and evolution of the program. On the one
561 hand, the sooner they're done, the less code we have to rip
562 up when the time comes to install them. On the other hand, they
563 don't bring an immediately perceptible benefit to potential
564 users. Item 10 and possibly item 11 should be deferred until Bacula
565 is well established with a growing user community more or
566 less happy with the feature set. At that time, it will make a
567 good "next generation" upgrade in the interest of data
573 ======================================================
575 It is somewhat like a Full save becomes an incremental since
576 the Base job (or jobs) plus other non-base files.
578 - New BaseFile table that contains:
579 JobId, BaseJobId, FileId (from Base).
580 i.e. for each base file that exists but is not saved because
581 it has not changed, the File daemon sends the JobId, BaseId,
582 and FileId back to the Director who creates the DB entry.
583 - To initiate a Base save, the Director sends the FD
584 the FileId, and full filename for each file in the Base.
585 - When the FD finds a Base file, he requests the Director to
586 send him the full File entry (stat packet plus MD5), or
587 conversely, the FD sends it to the Director and the Director
588 says yes or no. This can be quite rapid if the FileId is kept
589 by the FD for each Base Filename.
590 - It is probably better to have the comparison done by the FD
591 despite the fact that the File entry must be sent across the
593 - An alternative would be to send the FD the whole File entry
594 from the start. The disadvantage is that it requires a lot of
595 space. The advantage is that it requires less communications
597 - The Job record must be updated to indicate that one or more
599 - At end of Job, FD returns:
600 1. Count of base files/bytes not written to tape (i.e. matches)
601 2. Count of base file that were saved i.e. had changed.
602 - No tape record would be written for a Base file that matches, in the
603 same way that no tape record is written for Incremental jobs where
604 the file is not saved because it is unchanged.
605 - On a restore, all the Base file records must explicitly be
606 found from the BaseFile tape. I.e. for each Full save that is marked
607 to have one or more Base Jobs, search the BaseFile for all occurrences
609 - An optimization might be to make the BaseFile have:
615 This would avoid the need to explicitly fetch each File record for
616 the Base job. The Base Job record will be fetched to get the
617 VolSessionId and VolSessionTime.
618 =========================================================
622 =============================================================
624 Request For Comments For File Backup Options
627 Subject: File Backup Options
630 A few days ago, a Bacula user who is backing up to file volumes and
631 using compression asked if it was possible to suppress compressing
632 all .gz files since it was a waste of CPU time. Although Bacula
633 currently permits using different options (compression, ...) on
634 a directory by directory basis, it cannot do it on a file by
635 file basis, which is clearly what was desired.
637 Proposed Implementation:
638 To solve this problem, I propose the following:
640 - Add a new Director resource type called FileOptions.
642 - The FileOptions resource will have records for all
643 options that can currently be specified on the Include record
644 (in a FileSet). Examples below.
646 - The FileOptions resource will permit an exclude option as well
647 as a number of additional options.
649 - The heart of the FileOptions resource is the ability to
650 supply any number of ApplyTo records which specify POSIX
651 regular expressions. These ApplyTo regular expressions are
652 applied to the fully qualified filename (path and all). If
653 one matches, then the FileOptions will be used.
655 - When an ApplyTo specification matches an included file, the
656 options specified in the FileOptions resource will override
657 the default options specified on the Include record.
659 - Include records will be modified to permit referencing one or
660 more FileOptions resources. The FileOptions will be used
661 in the order listed on the Include record and the first
662 one that matches will be applied.
664 - Options (or specifications) currently supplied on the Include
665 record will be deprecated (i.e. removed in a later version a
666 year or so from now).
668 - The Exclude record will be deprecated as the same functionality
669 can be obtained by using an Exclude = yes in the FileOptions.
672 The following records can appear in the FileOptions resource. An
673 asterisk preceding the name indicates a feature not currently
677 - Compression= (GZIP, ...)
678 - Signature= (MD5, SHA1, ...)
680 - OneFs= (yes/no) - remain on one filesystem
681 - Recurse= (yes/no) - recurse into subdirectories
682 - Sparse= (yes/no) - do sparse file backup
683 - *Exclude= (yes/no) - exclude file from being saved
684 - *Reader= (filename) - external read (backup) program
685 - *Plugin= (filename) - read/write plugin module
688 - verify= (ipnougsamc5) - verify options
691 - replace= (always/ifnewer/ifolder/never) - replace options currently
693 - *Writer= (filename) - external write (restore) program
697 Currently options specifying compression, MD5 signatures, recursion,
698 ... of a FileSet are supplied on the Include record. These will now
699 all be collected into a FileOptions resource, which will be
700 specified on the Include in place of the options. Multiple FileOptions
701 may be specified. Since the FileOptions contain regular expressions
702 that are applied to the full filename, this will give the ability
703 to specify backup options on a file by file basis to whatever level
712 Include = compression=GZIP signature=MD5 {
721 Include = FileOptions=Opts {
732 That's a lot more to do the same thing, but it gives the ability to
733 apply options on a file by file basis. For example, suppose you
734 want to compress all files but not any file with extensions .gz or .Z.
735 You could do so as follows:
739 Include = FileOptions=NoCompress FileOptions=Opts {
747 ApplyTo = /*.?*/ # matches all files
752 # Note multiple ApplyTos are ORed
753 ApplyTo = /*.gz/ # matches .gz files */
754 ApplyTo = /*.Z/ # matches .Z files */
757 Now, since the NoCompress FileOptions is specified first on the
758 Include line, any *.gz or *.Z file will have an MD5 signature computed,
759 but will not be compressed. For all other files, the NoCompress will not
760 match, so the Opts options will be used which will include GZIP
764 - Is it necessary to provide some means of ANDing regular expressions
765 and negation? (not currently planned)
767 e.g. ApplyTo = /*.gz/ && !/big.gz/
769 - I see that Networker has a "null" module which, if specified, does not
770 backup the file, but does make an record of the file in the catalog
771 so that the catalog will reflect an exact picture of the filesystem.
772 The result is that the file can be "seen" when "browsing" the save
773 sets, but it cannot be restored.
775 Is this really useful? Should it be implemented in Bacula?
778 After implementing the above, the user will be able to specify
779 on a file by file basis (using regular expressions) what options are
780 applied for the backup.
783 =============================================
785 ==========================================================
787 For each Incremental job that is run, there may be files that
788 were found but not saved because they were locked (this applies
789 only to Windows). Such a system could send back to the Director
790 a list of Unsaved files.
792 - New UnSavedFiles table that contains:
796 - Then in the next Incremental job, the list of Unsaved Files will be
797 feed to the FD, who will ensure that they are explicitly chosen even
798 if standard date/time check would not have selected them.
799 =============================================================
802 Done: (see kernsdone for more)
803 - Heartbeat between daemons.
804 - Fix Dir heartbeat in restore and verify vol. Be sure to make
805 bnet_recv() ignore BNET_HEARTBEAT.
806 - Implement HEART_BEAT while SD waiting for tapes.
807 - Include RunBeforeJob and RunAfterJob output in the message
809 - Change M_INFO to M_RESTORED for all restored files.
810 - Fix command prompt in gnome-console by checking on Ready.
811 - Merge SQLite, MySQL, and Rel spec into a single file.
812 - Fix config of "console"
813 - Check if cancel works with FD (fixed).
814 - Properly configure console and gconsole (currently for source not
815 configured for installation).
816 - Error labeling tape from console gets Jmsg error because of no Job.
817 - Test and implement get_pint and get_yesno.
818 - Implement global with DB name and add to btraceback.gdb
819 - Remove subsysdir from conf files (used only in autostart scripts).
821 rufus-dir: Max configured use duration exceeded. Marking Volume "MatouBackup" as Used.
822 rufus-sd: Volume "" previously written, moving to end of data.
823 rufus-sd: Matou.2003-05-10_10.39.18 Error: I canot write on this volume because:
824 The number of files mismatch! Volume=1 Catalog=0
825 rufus-sd: Matou.2003-05-10_10.39.18 Error: askdir.c:155 NULL Volume name. This shouldn't happen!!!
826 - Shell character expansion is failing occassionally.
827 - Add a section to the doc on Manual cycling of Volumes.
828 - Check if Job/File retentions apply to multivolume jobs.
829 - Fix missing casette in autoloader during read:
830 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
831 14-May-2003 14:41 undef-sd: 3301 Issuing autochanger "loaded" command.
832 14-May-2003 14:41 undef-sd: 3302 Issuing autochanger "unload" command.
833 14-May-2003 14:42 undef-sd: 3303 Issuing autochanger "load slot 1" command.
834 14-May-2003 14:42 undef-sd: 3304 Autochanger "load slot 1" status is OK.
835 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
836 14-May-2003 14:42 undef-sd: 3301 Issuing autochanger "loaded" command.
837 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
838 14-May-2003 14:42 undef-sd: 3301 Issuing autochanger "loaded" command.
839 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
840 14-May-2003 14:42 undef-sd: 3301 Issuing autochanger "loaded" command.
841 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
842 14-May-2003 14:42 undef-sd: 3301 Issuing autochanger "loaded" command.
843 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".
844 14-May-2003 14:42 undef-dir: Bacula 1.31 (12May03): 14-May-2003 14:42
845 - Fix problem reported by Christopher McCurdy <cmccurdy@eecis.udel.edu>
846 xeon-fd: Could not stat c:/Documents and Settings/All
847 Users/Application Data/Humc:\Documents and Settings\All User98_AIX.kbf:
848 ERR=No such file or directory
850 - The following Re-read last block at EOT failed. ERR=block.c:523 Read zero bytes on device /dev/nrsa0.
851 undef-sd: block.c:523 Read zero bytes on device /dev/nrsa0.
852 apparently masks the standard EOM message.
853 - BSD (probably) does not have strtoll()
854 - BSD does not have ioctl() MTEOM
855 - BSD defines a number of MT_xxx variables which conflict
856 with those defined by Bacula.
857 - Make default duration days if no qualifier (e.g. s) is specified.
858 - BSDI fix finding gcc version
859 - When the FD errs (e.g. disk full) have a more graceful shutdown.
860 - Make sure Bacula prunes/purges canceled and failed jobs too and all
861 jobs with zero JobFiles.
862 - Implement Volume name checking.
863 - Document what characters can go into Volume names.
864 - Getting the following on all directories on Win32
865 19-May-2003 01:14 tibs-fd: Could not access c:/cygwin/home/kern/rxvt: ERR=Permission denied
866 - Cancellation caused JobMedia error:
867 babylon5-dir: Last FULL backup time not found. Doing FULL backup.
868 babylon5-dir: Start Backup JobId 416, Job=Zocalo_Save.2003-05-19_02.15.06
869 babylon5-sd: End of media on Volume VXA-V17-Inc-001 Bytes=31,982,900,672 Blocks=495,781.
870 babylon5-sd: Job Zocalo_Save.2003-05-19_02.15.06 waiting. Cannot find any appendable volumes.
871 babylon5-sd: Someone woke me up, but I cannot find any appendable volumes
872 for Job=Zocalo_Save.2003-05-19_02.15.06.
873 babylon5-sd: Zocalo_Save.2003-05-19_02.15.06 Fatal error: Job
874 Zocalo_Save.2003-05-19_02.15.06 canceled while waiting for mount on
875 Storage Device "Ecrix_VXA-1".
876 babylon5-sd: Zocalo_Save.2003-05-19_02.15.06 Fatal error: Cannot fixup device
877 error. Job Zocalo_Save.2003-05-19_02.15.06 canceled while waiting for
878 mount on Storage Device "Ecrix_VXA-1".
880 babylon5-dir: Zocalo_Save.2003-05-19_02.15.06 Error: Catalog error creating
881 JobMedia record. sql_create.c:125 Create JobMedia failed. Record already
883 babylon5-sd: Zocalo_Save.2003-05-19_02.15.06 Error: Error creating JobMedia
884 record: 1991 Update JobMedia error
886 babylon5-sd: Zocalo_Save.2003-05-19_02.15.06 Error: askdir.c:158 NULL Volume
887 name. This shouldn't happen!!!
888 zocalo-fd: Zocalo_Save.2003-05-19_02.15.06 Error: bnet.c:310 Write error
889 sending to Storage daemon:babylon5:9103: ERR=Broken pipe
890 - Volume names with spaces get jammed into the catalog with 0x1
891 i.e. the SD bashes the Volume but they are not unbased by Dir.
892 jerom-dir: MonthlySave.2003-05-10_17.12.01 Error: Unable to get Media
893 record for Volume Tape^A1: ERR=sql_get.c:788 Media record for Volume
895 jerom-sd: MonthlySave.2003-05-10_17.12.01 Error: Error updating Volume
896 Info: 1991 Catalog Request failed: sql_get.c:788 Media record for Volume
898 - ChangeServiceConfig2A does not exist on WinNT (ADVAPI32.DLL).
899 - Fix "access not allowed" for backup of files on WinXP.
900 - Check for existence of all new Win32 API's. See LoadLibrary in
902 - Count errors during restore and print them in the Job report.
903 - Bug: fix access problems on files restored on WinXP.
904 - Put system type returned by FD into catalog.
905 - Finish WIN32_DATA stream code (bextract, check if can handle stream)
906 - Make SD keep track of Files, Bytes during restore.
907 - If you enter the userid by hand for restore, you get:
908 Enter JobId(s), comma separated, to restore: 74
909 You have selected the following JobId: 74
910 Building directory tree for JobId 74 ...
911 134645140 items inserted into the tree and marked for extraction.
912 - Add SDWriteSeqNo to SD, and probably Read on FD side.
913 - If bootstrap is non-zero for restore, do not show JobId in the
914 OK to run? (yes/mod/no): list.
915 - When all cassettes in magazine are used, got:
916 22-May-2003 18:24 undef-sd: 3304 Autochanger "load slot 1" status is OK.
917 22-May-2003 18:24 undef-sd: NightlySave.2003-05-22_14.08.16 Warning: mount.c:245 Director wanted Volume "TestVolume0009".
918 Current Volume "TestVolume0005" not acceptable because:
919 1998 Volume "TestVolume0005" not Append or Recycle.
920 22-May-2003 18:24 undef-sd: NightlySave.2003-05-22_14.08.16 Error: Autochanger Volume "TestVolume0009" not found in slot 1.
921 Setting slot to zero in catalog.
922 22-May-2003 18:24 undef-sd: Please mount Volume "TestVolume0009" on Storage Device "ARCHIVE 4586" for Job NightlySave.2003-05-22_14
924 Use "mount" command to release Job.
925 22-May-2003 19:24 undef-sd: Please mount Volume "TestVolume0009" on Storage Device "ARCHIVE 4586" for Job NightlySave.2003-05-22_14
927 Use "mount" command to release Job.
928 - Don't zero the Slot when the wrong volume is found -- simply ask
930 - Implement MTIOCERRSTAT on FreeBSD to clear tape error conditions.
931 - Shell expansion fails for working_directory in SD from time to time.
932 - File the Automatically selected: xxx
933 to say Automatically selected Pool: xxx
934 - Default duration with no qualifier is sec should be 1 day
935 - zap sd_auth_key in SD after FD connection.
936 - Find a solution for the multiple FileSet problem (when it is changed). Add date?
937 - Look at Python for a Bacula scripting language -- www.python.org
938 - When Marking a file in Restore that is a hard link, also
939 mark the link so that the data will be reloaded.
940 - Emergency restore info:
942 - Backup working directory
944 - Why don't we get an error message from Win32 FD when bootstrap
945 file cannot be created for restore command?
946 - Fix Win2000 error with no messages during startup.
947 - Make restore more robust in counting error and not immediately bailing
948 out. Also print error message once, but try to continue.
949 - Add code to check that blocks are sequential on restore.
950 - Remove "rufus" and such references from regress.
951 - No READLINE_SRC if found in alternate directory.
952 - If ./btape is called without /dev, assume argument is a Storage resource name.
953 - Find general solution for sscanf size problems (as well as sprintf. Do at run time?
954 - Bytes restored is wrong.
955 - The "List last 20 Jobs run" doesnt work correctly in restore.
956 It doesnt show the last 20 jobs , but some older ones.
957 - Fix Verify VolumeToCatalog to use BSRs -- it is broken.
958 - Implement Release Storage=xxx
959 - Fix restore on Win95/98
960 - Remove the Jmsg() in sql_find.c:102 or only print on hard error.
961 - Implement FileSet VolIndex -- done, but must update old records.
962 - Check this below from Phil.
963 This was SD reported data rather than FD data!
964 > When the job was done, Bacula reported 11084 files restored:
967 > Job: Zocalo_Restore.2003-06-05_16.42.01
969 > Start time: 05-Jun-2003 16:42
970 > End time: 06-Jun-2003 01:21
971 > Files Restored: 11,084
972 > Bytes Restored: 65,474,772
974 > FD termination status: OK
975 > Termination: Restore OK
977 > when it should probably have reported 11084 files scanned, 250 restored.
978 > The bytes restored count looks about right.
980 - Should Bacula make an Append tape as Purged when purging?
981 - Use switch() in backup.c and restore.c in FD instead of giant if statement.
982 - If during a restore, a hard linked file already exists (on option), delete
983 the file and re-link it. This is to avoid the possibility that the
984 user had re-linked the file between the backup and the restore.
985 Do lstat() to see if it is already properly linked.
986 Same for symlinked file.
987 Make sure ifnewer, ifolder, never, ... apply correctly.
988 - Flag so that no connect does not error, and Reschedule a job.
989 - Implement "Reschedule OnError=yes interval=nnn times=xxx"
990 - That restoring a hard link that already exists works correctly.
992 - Make Pool resource handle Counter resources.
993 - Fix first block number after label to be zero instead of 1 (reset after label).
994 - Grep for Backup OK in regression script.
995 - Do NOT reuse same JobId if tape written.
996 - Implement non-blocking writes and bsock->terminate in heartbeat
997 thread, or set it in status.c cancel (used pthread_kill() instead of
999 - Add restore to specific date.
1000 - Instrument use_count on DEVICE packets and ensure that the device is
1001 being close()ed at the appropriate time.
1002 - Test long path names (>64 chars) in Windows -- crashes FD?
1003 - Implement fast block rejection: match_bsr_block().