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 all the status codes JobLevel, JobType, JobStatus.
19 Testing to do: (painful)
20 - that ALL console command line options work and are always implemented
21 - blocksize recognition code.
22 - Test if rewind at end of tape waits for tape to rewind.
24 - Test not zeroing Autochanger slot when it is wrong.
25 - Test recycling and purging (code changed in db_find_next_volume and
27 - Figure out how to use ssh or stunnel to protect Bacula communications.
30 - Implement job in VerifyToCatalog
32 - Implement List Volume Job=xxx or List scheduled volumes or Status Director
33 - Allow a slot or range of slots on the label barcodes command.
34 - Implement a PostgreSQL driver.
35 - Is a pool specification really needed for a restore?
36 - Bacula needs to propagate SD errors.
37 > > cluster-dir: Start Backup JobId 252, Job=REUTERS.2003-08-11_15.04.12
38 > > prod4-sd: REUTERS.2003-08-11_15.04.12 Error: Write error on device
39 /dev/nst0. ERR=Input/output error.
40 > > prod4-sd: REUTERS.2003-08-11_15.04.12 Error: Re-read of last block failed.
41 Last block=5162 Current block=5164.
42 > > prod4-sd: End of medium on Volume "REU007" Bytes=16,303,521,933
44 - Use autochanger to handle multiple devices.
45 - Fix packet too big problem.
46 - Add SuSE install doc to list.
47 - Check and rechedk "Invalid block number"
48 - Make bextract release the drive properly between tapes
49 so that an autochanger can be made to work.
50 - Fix "restore all" to bypass building the tree.
51 - Fix restore to list errors if Invalid block found, and if # files
52 restored does not match # expected.
53 - User wants to NOT backup up certain big files (email files).
54 - Maybe remove multiple simultaneous devices code in SD.
55 - On Windows with very long path names, it may be impossible to create
56 a file (and thus restore it) because the total length is too long.
57 We must cd into the directory then create the file without the
59 - lstat() is not going to work on Win32 for testing date.
60 - Something is not right in last block of fill command.
61 - Implement a Recycle command
62 - Add FileSet to command line arguments for restore.
63 - Do full check the command line args in update (e.g. VolStatus ...).
64 - Allow multiple Storage specifications (or multiple names on
65 a single Storage specification) in the Job record. Thus a job
66 can be backed up to a number of storage devices.
67 - Add client name to cram-md5 challenge so Director can immediately
68 verify if it is the correct client.
69 - Implement ClientRunBeforeJob and ClientRunAfterJob.
70 - Add JobLevel in FD status (but make sure it is defined).
71 - Audit all UA commands to ensure that we always prompt where possible.
72 - Restrict characters permitted in a Resource name, and don't permit
74 - Prohibit backing up archive device (findlib/find_one.c:128)
75 - Make | and < work on FD side.
76 - Check Jmsg in bnet, may not work, must dup bsock.
77 - Suppress Job Name in Jmsg for console
78 - Create Pools that are referenced in a Run statement at startup if possible.
79 - Use runbeforejob to unload, then reload a volume previously used,
80 then the next job run gets an error reading the drive.
81 - Make bootstrap filename unique.
82 - Test a second language e.g. french.
83 - Start working on Base jobs.
84 - Make "make binary-release" work from any directory.
85 - Implement UnsavedFiles DB record.
86 - Implement argc/argv for daemon command line scanning using table driven
88 - Implement table driven single argc/argv scanner to pickup all arguments.
89 Much like xxx_conf.c scan table.
90 keyword, handler(store_routine), store_address, code, flags, default.
91 - From Phil Stracchino:
92 It would probably be a per-client option, and would be called
93 something like, say, "Automatically purge obsoleted jobs". What it
94 would do is, when you successfully complete a Differential backup of a
95 client, it would automatically purge all Incremental backups for that
96 client that are rendered redundant by that Differential. Likewise,
97 when a Full backup on a client completed, it would automatically purge
98 all Differential and Incremental jobs obsoleted by that Full backup.
99 This would let people minimize the number of tapes they're keeping on
100 hand without having to master the art of retention times.
101 - Implement new serialize subroutines
102 send(socket, "string", &Vol, "uint32", &i, NULL)
103 - Scratch Pool where the volumes can be re-assigned to any Pool.
104 - Implement a M_SECURITY message class.
105 - Implement forward spacing block/file: position_device(bsr) --
106 just before read_block_from_device();
107 - When doing a Backup send all attributes back to the Director, who
108 would then figure out what files have been deleted.
109 - Currently in mount.c:236 the SD simply creates a Volume. It should have
110 explicit permission to do so. It should also mark the tape in error
111 if there is an error.
112 - Make sure all restore counters are working correctly in the FD.
113 - SD Bytes Read is wrong.
114 - Configure mtx-changer to have correct path to mtx.
115 - Look at ALL higher level routines that call block.c to be sure
116 they don't expect something in errmsg.
117 - Investigate doing RAW backup of Win32 partition.
118 - Add JobName= to VerifyToCatalog so that all verifies can be done at the end.
119 - Add thread specific data to hold the jcr -- send error messages from
120 low level routines by accessing it and using Jmsg().
121 - Cancel waiting for Client connect in SD if FD goes away.
123 - Add Progress command that periodically reports the progress of
125 - One block was orphaned in the SD probably after cancel.
126 - Add all command line arguments to "update", e.g. slot=nn volStatus=append, ...
129 - Examine Bare Metal restore problem (a FD crash exists somewhere ...).
130 - Implement timeout in response() when it should come quickly.
131 - Implement console @echo command.
132 - Implement a Slot priority (loaded/not loaded).
133 - Implement "vacation" Incremental only saves.
134 - Implement single pane restore (much like the Gftp panes).
135 - Implement Automatic Mount even in operator wait.
136 - Implement create "FileSet"?
137 - Implement Release Device in the Job resource to unmount a drive.
138 - Implement Acquire Device in the Job resource to mount a drive,
139 be sure this works with admin jobs so that the user can get
140 prompted to insert the correct tape. Possibly some way to say to
141 run the job but don't save the files.
142 - Implement all command line args on run.
143 - Implement command line "restore" args.
144 - Implement "restore current select=no"
145 - Fix watchdog pthread crash on Win32 (this is pthread_kill() Cygwin bug)
146 - Implement "scratch pool" where tapes are defined and can be
147 taken by any pool that needs them.
148 - Implement restore "current system", but take all files without
149 doing selection tree -- so that jobs without File records can
151 - Implement disk spooling. Two parts: 1. Spool to disk then
152 immediately to tape to speed up tape operations. 2. Spool to
153 disk only when the tape is full, then when a tape is hung move
155 - Implement a relocatable bacula.spec
156 - Implement dump/print label to UA
157 - Add prefixlinks to where or not where absolute links to FD.
158 - Issue message to mount a new tape before the rewind.
159 - Simplified client job initiation for portables.
160 - If SD cannot open a drive, make it periodically retry.
161 - Implement LabelTemplate (at least first cut).
162 - Add more of the config info to the tape label.
163 - Implement a Mount Command and an Unmount Command where
164 the users could specify a system command to be performed
165 to do the mount, after which Bacula could attempt to
166 read the device. This is for Removeable media such as a CDROM.
167 - Most likely, this mount command would be invoked explicitly
168 by the user using the current Console "mount" and "unmount"
169 commands -- the Storage Daemon would do the right thing
170 depending on the exact nature of the device.
171 - As with tape drives, when Bacula wanted a new removable
172 disk mounted, it would unmount the old one, and send a message
173 to the user, who would then use "mount" as described above
174 once he had actually inserted the disk.
176 - Make some way so that if a machine is skipped because it is not up
177 that Bacula will continue retrying for a specified period of time --
179 - If tape is marked read-only, then try opening it read-only rather than
180 failing, and remember that it cannot be written.
181 - Refine SD waiting output:
182 Device is being positioned
183 > Device is being positioned for append
184 > Device is being positioned to file x
186 - Figure out some way to estimate output size and to avoid splitting
187 a backup across two Volumes -- this could be useful for writing CDROMs
188 where you really prefer not to have it split -- not serious.
189 - Add RunBeforeJob and RunAfterJob to the Client program.
190 - Have SD compute MD5 or SHA1 and compare to what FD computes.
191 - Make VolumeToCatalog calculate an MD5 or SHA1 from the
192 actual data on the Volume and compare it.
193 - Implement FileOptions (see end of this document)
194 - Implement Bacula plugins -- design API
195 - Make bcopy read through bad tape records.
196 - Fix read_record to handle multiple sessions.
197 - Program files (i.e. execute a program to read/write files).
198 Pass read date of last backup, size of file last time.
199 - Add Signature type to File DB record.
200 - CD into subdirectory when open()ing files for backup to
201 speed up things. Test with testfind().
202 - Priority job to go to top of list.
203 - Why are save/restore of device different sizes (sparse?) Yup! Fix it.
204 - Implement some way for the Console to dynamically create a job.
205 - Restore to a particular time -- e.g. before date, after date.
206 - Solaris -I on tar for include list
207 - Need a verbose mode in restore, perhaps to bsr.
208 - bscan without -v is too quiet -- perhaps show jobs.
209 - Add code to reject whole blocks if not wanted on restore.
210 - Check if we can increase Bacula FD priorty in Win2000
211 - Make sure the MaxVolFiles is fully implemented in SD
212 - Check if both CatalogFiles and UseCatalog are set to SD.
213 - Need return status on read_cb() from read_records(). Need multiple
214 records -- one per Job, maybe a JCR or some other structure with
215 a block and a record.
216 - Figure out how to do a bare metal Windows restore
217 - Possibly add email to Watchdog if drive is unmounted too
218 long and a job is waiting on the drive.
219 - Use read_record.c in SD code.
220 - Restore program that errors in SD due to no tape reports
221 OK incorrectly in output.
222 - After unmount, if restore job started, ask to mount.
223 - Convert all %x substitution variables, which are hard to remember
224 and read to %(variable-name). Idea from TMDA.
225 - Remove NextId for SQLite. Optimize.
226 - Move all SQL statements into a single location.
227 - Add UA rc and history files.
228 - put termcap (used by console) in ./configure and
229 allow -with-termcap-dir.
230 - Enhance time and size scanning routines.
231 - Fix Autoprune for Volumes to respect need for full save.
232 - Fix Win32 config file definition name on /install
233 - Compare tape to Client files (attributes, or attributes and data)
234 - Make all database Ids 64 bit.
235 - Write an applet for Linux.
236 - Add estimate to Console commands
237 - Implement new daemon communications protocol.
238 - Allow console commands to detach or run in background.
239 - Fix status delay on storage daemon during rewind.
240 - Add SD message variables to control operator wait time
241 - Maximum Operator Wait
242 - Minimum Message Interval
243 - Maximum Message Interval
244 - Send Operator message when cannot read tape label.
245 - Verify level=Volume (scan only), level=Data (compare of data to file).
246 Verify level=Catalog, level=InitCatalog
248 - Add keyword search to show command in Console.
249 - Events : tape has more than xxx bytes.
250 - Complete code in Bacula Resources -- this will permit
251 reading a new config file at any time.
252 - Handle ctl-c in Console
253 - Implement script driven addition of File daemon to config files.
254 - Think about how to make Bacula work better with File (non-tape) archives.
255 - Write Unix emulator for Windows.
256 - Put memory utilization in Status output of each daemon
257 if full status requested or if some level of debug on.
258 - Make database type selectable by .conf files i.e. at runtime
259 - Set flag for uname -a. Add to Volume label.
260 - Implement throttled work queue.
261 - Restore files modified after date
262 - Restore file modified before date
263 - Restore -- do nothing but show what would happen
264 - SET LD_RUN_PATH=$HOME/mysql/lib/mysql
265 - Implement Restore FileSet=
266 - Create a protocol.h and protocol.c where all protocol messages
268 - Remove duplicate fields from jcr (e.g. jcr.level and jcr.jr.Level, ...).
269 - Timout a job or terminate if link goes down, or reopen link and query.
270 - Concept of precious tapes (cannot be reused).
271 - Make bcopy copy with a single tape drive.
272 - Permit changing ownership during restore.
275 > My suggestion: Add a feature on the systray menu-icon menu to request
276 > an immediate backup now. This would be useful for laptop users who may
277 > not be on the network when the regular scheduled backup is run.
279 > My wife's suggestion: Add a setting to the win32 client to allow it to
280 > shut down the machine after backup is complete (after, of course,
281 > displaying a "System will shut down in one minute, click here to cancel"
282 > warning dialog). This would be useful for sites that want user
283 > woorkstations to be shut down overnight to save power.
286 - Autolabel should be specified by DIR instead of SD.
289 - AutoScan (check checksum of tape)
290 - Format command = "format /dev/nst0"
294 - Seek resolution (usually corresponds to buffer size)
295 - EODErrorCode=ENOSPC or code
296 - Partial Read error code
297 - Partial write error code
298 - Nonformatted read error
299 - Nonformatted write error
300 - WriteProtected error
304 - IgnoreCloseErrors=yes
314 - FD sends unsaved file list to Director at end of job (see
316 - File daemon should build list of files skipped, and then
317 at end of save retry and report any errors.
318 - Write a Storage daemon that uses pipes and
319 standard Unix programs to write to the tape.
321 - Need something that monitors the JCR queue and
322 times out jobs by asking the deamons where they are.
323 - Enhance Jmsg code to permit buffering and saving to disk.
324 - device driver = "xxxx" for drives.
326 - Ensure that /dev/null works
327 - Need report class for messages. Perhaps
328 report resource where report=group of messages
329 - enhance scan_attrib and rename scan_jobtype, and
330 fill in code for "since" option
331 - Director needs a time after which the report status is sent
332 anyway -- or better yet, a retry time for the job.
333 - Don't reschedule a job if previous incarnation is still running.
334 - Some way to automatically backup everything is needed????
335 - Need a structure for pending actions:
337 - termination status (part of buffered msgs?)
339 Read, Write, Clean, Delete
340 - Login to Bacula; Bacula users with different permissions:
341 owner, group, user, quotas
342 - Store info on each file system type (probably in the job header on tape.
343 This could be the output of df; or perhaps some sort of /etc/mtab record.
346 - Design at hierarchial storage for Bacula. Migration and Clone.
347 - Implement FSM (File System Modules).
348 - Audit M_ error codes to ensure they are correct and consistent.
349 - Add variable break characters to lex analyzer.
350 Either a bit mask or a string of chars so that
351 the caller can change the break characters.
352 - Make a single T_BREAK to replace T_COMMA, etc.
353 - Ensure that File daemon and Storage daemon can
354 continue a save if the Director goes down (this
355 is NOT currently the case). Must detect socket error,
356 buffer messages for later.
357 - Enhance time/duration input to allow multiple qualifiers e.g. 3d2h
358 - Add ability to backup to two Storage devices (two SD sessions) at
359 the same time -- e.g. onsite, offsite.
360 - Add the ability to consolidate old backup sets (basically do a restore
361 to tape and appropriately update the catalog). Compress Volume sets.
362 Might need to spool via file is only one drive is available.
363 - Compress or consolidate Volumes of old possibly deleted files. Perhaps
364 someway to do so with every volume that has less than x% valid
368 Migration: Move a backup from one Volume to another
369 Clone: Copy a backup -- two Volumes
371 Bacula Migration is based on Jobs (apparently Networker is file by file).
373 Migration triggered by:
377 Highwater mark (keep total size)
381 Bacula Projects Roadmap
383 last update 8 May 2003
385 Item 1: Multiple simultaneous Jobs. (done)
386 Done -- Restore part needs better implementation to work correctly
387 Also, it needs considerable testing
389 What: Permit multiple simultaneous jobs in Bacula.
391 Why: An enterprise level solution needs to go fast without the
392 need for the system administrator to carefully tweak
393 timing. Based on the benchmarks, during a full
394 backup, NetWorker typically hit 10 times the bandwidth to
395 the tape compared to Bacula--largely. This is probably due to
396 running parallel jobs and multi-threaded filling of buffers
397 and writing them to tape. This should also make things work
398 better when you have a mix of fast and slow machines backing
401 Notes: Bacula was designed to run multiple simultaneous jobs. Thus
402 implementing this is a matter of some small cleanups and
406 Item 2: Make the Storage daemon use intermediate file storage to buffer data.
407 Deferred -- not necessary yet -- possibly implement with Migration.
409 What: If data is coming into the SD too fast, buffer it to
410 disk if the user has configured this option.
412 Why: This would be nice, especially if it more or less falls out
413 when implementing (1) above. If not, it probably should not
414 be given a high priority because fundamentally the backup time
415 is limited by the tape bandwidth. Even though you may finish a
416 client job quicker by spilling to disk, you still have to
417 eventually get it onto tape. If intermediate disk buffering
418 allows us to improve write bandwidth to tape, it may make
421 Notes: Whether or not this is implemented will depend upon performance
422 testing after item 1 is implemented.
425 Item 3: Write the bscan program -- also write a bcopy program.
428 What: Write a program that reads a Bacula tape and puts all the
429 appropriate data into the catalog. This allows recovery
430 from a tape that is no longer in the database, or it allows
431 re-creation of a database if lost.
433 Why: This is a fundamental robustness and disaster recovery tool
434 which will increase the comfort level of a sysadmin
435 considering adopting Bacula.
437 Notes: A skeleton of this program already exists, but much work
438 needs to be done. Implementing this will also make apparent
439 any deficiencies in the current Bacula tape format.
442 Item 4: Implement Base jobs.
444 What: A base job is sort of like a Full save except that you
445 will want the FileSet to contain only files that are unlikely
446 to change in the future (i.e. a snapshot of most of your
447 system after installing it). After the base job has been run,
448 when you are doing a Full save, you can specify to exclude
449 all files saved by the base job that have not been modified.
451 Why: This is something none of the competition does, as far as we know
452 (except BackupPC, which is a Perl program that saves to disk
453 only). It is big win for the user, it makes Bacula stand out
454 as offering a unique optimization that immediately saves time
457 Notes: Big savings in tape usage. Will require more resources because
458 the e. DIR must send FD a list of files/attribs, and the FD must
459 search the list and compare it for each file to be saved.
462 Item 5: Implement Label templates
465 What: This is a mechanism whereby Bacula can automatically create
466 a tape label for new tapes according to a detailed specification
467 provided by the user.
469 Why: It is a major convenience item for folks who use automated label
472 Notes: Bacula already has a working form of automatic tape label
473 creation, but it is very crude. The design for the complete
474 tape labeling project is already documented in the manual.
477 Item 6: Write a regression script.
478 Done -- Continue to expand its testing.
480 What: This is an automatic script that runs and tests as many features
481 of Bacula as possible. The output is compared to previous
482 versions of Bacula and any differences are reported.
484 Why: This is an enormous help in preventing introduction of new
485 errors in parts of the program that already work correctly.
487 Notes: This probably should be ranked higher, it's something the typical
488 user doesn't see. Depending on how it's implemented, it may
489 make sense to defer it until the archival tape format and
490 user interface mature.
493 Item 7: GUI for interactive restore
494 Item 8: GUI for interactive backup
496 What: The current interactive restore is implemented with a tty
497 interface. It would be much nicer to be able to "see" the
498 list of files backed up in typical GUI tree format.
499 The same mechanism could also be used for creating
500 ad-hoc backup FileSets (item 8).
502 Why: Ease of use -- especially for the end user.
504 Notes: Rather than implementing in Gtk, we probably should go directly
505 for a Browser implementation, even if doing so meant the
506 capability wouldn't be available until much later. Not only
507 is there the question of Windows sites, most
508 Solaris/HP/IRIX, etc, shops can't currently run Gtk programs
509 without installing lots of stuff admins are very wary about.
510 Real sysadmins will always use the command line anyway, and
511 the user who's doing an interactive restore or backup of his
512 own files will in most cases be on a Windows machine running
516 Item 9: Add SSL to daemon communications.
518 What: This provides for secure communications between the daemons.
520 Why: This would allow doing backup across the Internet without
521 privacy concerns (or with much less concern).
523 Notes: The vast majority of near term potential users will be backing up
524 a single site over a LAN and, correctly or not, they probably
525 won't be concerned with security, at least not enough to go to
526 the trouble to set up keys, etc. to screw things down. We suspect
527 that many users genuinely interested in multi-site backup
528 already run some form of VPN software in their internetwork
529 connections, and are willing to delegate security to that layer.
532 Item 10: Define definitive tape format.
535 What: Define that definitive tape format that will not change
536 for the next millennium.
538 Why: Stability, security.
540 Notes: See notes for item 11 below.
543 Item 11: New daemon communication protocol.
545 What: The current daemon to daemon protocol is basically an ASCII
546 printf() and sending the buffer. On the receiving end, the
547 buffer is sscanf()ed to unpack it. The new scheme would
548 be a binary format that allows quick packing and unpacking
549 of any data type with named fields.
551 Why: Using binary packing would be faster. Named fields will permit
552 error checking to ensure that what is sent is what the
553 receiver really wants.
555 Notes: These are internal improvements in the interest of the
556 long-term stability and evolution of the program. On the one
557 hand, the sooner they're done, the less code we have to rip
558 up when the time comes to install them. On the other hand, they
559 don't bring an immediately perceptible benefit to potential
560 users. Item 10 and possibly item 11 should be deferred until Bacula
561 is well established with a growing user community more or
562 less happy with the feature set. At that time, it will make a
563 good "next generation" upgrade in the interest of data
569 ======================================================
571 It is somewhat like a Full save becomes an incremental since
572 the Base job (or jobs) plus other non-base files.
574 - New BaseFile table that contains:
575 JobId, BaseJobId, FileId (from Base).
576 i.e. for each base file that exists but is not saved because
577 it has not changed, the File daemon sends the JobId, BaseId,
578 and FileId back to the Director who creates the DB entry.
579 - To initiate a Base save, the Director sends the FD
580 the FileId, and full filename for each file in the Base.
581 - When the FD finds a Base file, he requests the Director to
582 send him the full File entry (stat packet plus MD5), or
583 conversely, the FD sends it to the Director and the Director
584 says yes or no. This can be quite rapid if the FileId is kept
585 by the FD for each Base Filename.
586 - It is probably better to have the comparison done by the FD
587 despite the fact that the File entry must be sent across the
589 - An alternative would be to send the FD the whole File entry
590 from the start. The disadvantage is that it requires a lot of
591 space. The advantage is that it requires less communications
593 - The Job record must be updated to indicate that one or more
595 - At end of Job, FD returns:
596 1. Count of base files/bytes not written to tape (i.e. matches)
597 2. Count of base file that were saved i.e. had changed.
598 - No tape record would be written for a Base file that matches, in the
599 same way that no tape record is written for Incremental jobs where
600 the file is not saved because it is unchanged.
601 - On a restore, all the Base file records must explicitly be
602 found from the BaseFile tape. I.e. for each Full save that is marked
603 to have one or more Base Jobs, search the BaseFile for all occurrences
605 - An optimization might be to make the BaseFile have:
611 This would avoid the need to explicitly fetch each File record for
612 the Base job. The Base Job record will be fetched to get the
613 VolSessionId and VolSessionTime.
614 =========================================================
618 =============================================================
620 Request For Comments For File Backup Options
623 Subject: File Backup Options
626 A few days ago, a Bacula user who is backing up to file volumes and
627 using compression asked if it was possible to suppress compressing
628 all .gz files since it was a waste of CPU time. Although Bacula
629 currently permits using different options (compression, ...) on
630 a directory by directory basis, it cannot do it on a file by
631 file basis, which is clearly what was desired.
633 Proposed Implementation:
634 To solve this problem, I propose the following:
636 - Add a new Director resource type called FileOptions.
638 - The FileOptions resource will have records for all
639 options that can currently be specified on the Include record
640 (in a FileSet). Examples below.
642 - The FileOptions resource will permit an exclude option as well
643 as a number of additional options.
645 - The heart of the FileOptions resource is the ability to
646 supply any number of ApplyTo records which specify POSIX
647 regular expressions. These ApplyTo regular expressions are
648 applied to the fully qualified filename (path and all). If
649 one matches, then the FileOptions will be used.
651 - When an ApplyTo specification matches an included file, the
652 options specified in the FileOptions resource will override
653 the default options specified on the Include record.
655 - Include records will be modified to permit referencing one or
656 more FileOptions resources. The FileOptions will be used
657 in the order listed on the Include record and the first
658 one that matches will be applied.
660 - Options (or specifications) currently supplied on the Include
661 record will be deprecated (i.e. removed in a later version a
662 year or so from now).
664 - The Exclude record will be deprecated as the same functionality
665 can be obtained by using an Exclude = yes in the FileOptions.
668 The following records can appear in the FileOptions resource. An
669 asterisk preceding the name indicates a feature not currently
673 - Compression= (GZIP, ...)
674 - Signature= (MD5, SHA1, ...)
676 - OneFs= (yes/no) - remain on one filesystem
677 - Recurse= (yes/no) - recurse into subdirectories
678 - Sparse= (yes/no) - do sparse file backup
679 - *Exclude= (yes/no) - exclude file from being saved
680 - *Reader= (filename) - external read (backup) program
681 - *Plugin= (filename) - read/write plugin module
684 - verify= (ipnougsamc5) - verify options
687 - replace= (always/ifnewer/ifolder/never) - replace options currently
689 - *Writer= (filename) - external write (restore) program
693 Currently options specifying compression, MD5 signatures, recursion,
694 ... of a FileSet are supplied on the Include record. These will now
695 all be collected into a FileOptions resource, which will be
696 specified on the Include in place of the options. Multiple FileOptions
697 may be specified. Since the FileOptions contain regular expressions
698 that are applied to the full filename, this will give the ability
699 to specify backup options on a file by file basis to whatever level
708 Include = compression=GZIP signature=MD5 {
720 Match = /*.?*/ # matches all files.
725 That's a lot more to do the same thing, but it gives the ability to
726 apply options on a file by file basis. For example, suppose you
727 want to compress all files but not any file with extensions .gz or .Z.
728 You could do so as follows:
735 # Note multiple Matches are ORed
736 Match = /*.gz/ # matches .gz files */
737 Match = /*.Z/ # matches .Z files */
742 Match = /*.?*/ # matches all files
748 Now, since the NoCompress FileOptions is specified first on the
749 Include line, any *.gz or *.Z file will have an MD5 signature computed,
750 but will not be compressed. For all other files, the NoCompress will not
751 match, so the Opts options will be used which will include GZIP
755 - Is it necessary to provide some means of ANDing regular expressions
756 and negation? (not currently planned)
758 e.g. Match = /*.gz/ && !/big.gz/
760 - I see that Networker has a "null" module which, if specified, does not
761 backup the file, but does make an record of the file in the catalog
762 so that the catalog will reflect an exact picture of the filesystem.
763 The result is that the file can be "seen" when "browsing" the save
764 sets, but it cannot be restored.
766 Is this really useful? Should it be implemented in Bacula?
769 After implementing the above, the user will be able to specify
770 on a file by file basis (using regular expressions) what options are
771 applied for the backup.
774 =============================================
776 ==========================================================
778 For each Incremental job that is run, there may be files that
779 were found but not saved because they were locked (this applies
780 only to Windows). Such a system could send back to the Director
781 a list of Unsaved files.
783 - New UnSavedFiles table that contains:
787 - Then in the next Incremental job, the list of Unsaved Files will be
788 feed to the FD, who will ensure that they are explicitly chosen even
789 if standard date/time check would not have selected them.
790 =============================================================
793 Done: (see kernsdone for more)
794 - Implement new alist in FileSet scanning.
795 - bls should continue reading even if it finds Win32 data on the tape.
796 The error should be Warning rather the Error.
797 - Add user configurable timeout for connecting to SD.
798 - Unsaved Flag in Job record (use JobMissingFiles).
799 - Base Flag in Job record.