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
13 Testing to do: (painful)
14 - that ALL console command line options work and are always implemented
15 - blocksize recognition code.
16 - multiple simultaneous Volumes
17 - Test if rewind at end of tape waits for tape to rewind.
19 - Figure out how to use ssh or stunnel to protect Bacula communications.
22 - Document what characters can go into Volume names.
23 - Implement Volume name checking.
24 - Default duration with no qualifier is sec should be 1 day
25 - Getting the following on all directories on Win32
26 19-May-2003 01:14 tibs-fd: Could not access c:/cygwin/home/kern/rxvt: ERR=Permission denied
27 - Make sure Bacula prunes/purges canceled and failed jobs too!
28 - Find a solution for the multiple FileSet problem (when it is changed).
30 - Cancel waiting for Client connect in SD if FD goes away.
31 - Testing Tibs job erred and hung director on Storage resource.
32 - Make restore more robust in counting error and not immediately bailing
33 out. Also print error message once, but try to continue.
34 - Make SD keep track of Files, Bytes during restore.
35 - Add code to check that blocks are sequential on restore.
36 - File the Automatically selected: xxx
37 to say Automatically selected Pool: xxx
38 - Should Bacula make an Append tape as Purged when purging?
39 - Shell expansion failes for working_directory in SD from time to tim.
40 - Possibly update all client records at startup.
41 - Volume names with spaces get jammed into the catalog with 0x1
42 i.e. the SD bashes the Volume but they are not unbased by Dir.
43 jerom-dir: MonthlySave.2003-05-10_17.12.01 Error: Unable to get Media
44 record for Volume Tape^A1: ERR=sql_get.c:788 Media record for Volume
46 jerom-sd: MonthlySave.2003-05-10_17.12.01 Error: Error updating Volume
47 Info: 1991 Catalog Request failed: sql_get.c:788 Media record for Volume
50 - Cancellation caused JobMedia error:
51 babylon5-dir: Last FULL backup time not found. Doing FULL backup.
52 babylon5-dir: Start Backup JobId 416, Job=Zocalo_Save.2003-05-19_02.15.06
53 babylon5-sd: End of media on Volume VXA-V17-Inc-001 Bytes=31,982,900,672 Blocks=495,781.
54 babylon5-sd: Job Zocalo_Save.2003-05-19_02.15.06 waiting. Cannot find any appendable volumes.
55 babylon5-sd: Someone woke me up, but I cannot find any appendable volumes
56 for Job=Zocalo_Save.2003-05-19_02.15.06.
57 babylon5-sd: Zocalo_Save.2003-05-19_02.15.06 Fatal error: Job
58 Zocalo_Save.2003-05-19_02.15.06 canceled while waiting for mount on
59 Storage Device "Ecrix_VXA-1".
60 babylon5-sd: Zocalo_Save.2003-05-19_02.15.06 Fatal error: Cannot fixup device
61 error. Job Zocalo_Save.2003-05-19_02.15.06 canceled while waiting for
62 mount on Storage Device "Ecrix_VXA-1".
64 babylon5-dir: Zocalo_Save.2003-05-19_02.15.06 Error: Catalog error creating
65 JobMedia record. sql_create.c:125 Create JobMedia failed. Record already
67 babylon5-sd: Zocalo_Save.2003-05-19_02.15.06 Error: Error creating JobMedia
68 record: 1991 Update JobMedia error
70 babylon5-sd: Zocalo_Save.2003-05-19_02.15.06 Error: askdir.c:158 NULL Volume
71 name. This shouldn't happen!!!
72 zocalo-fd: Zocalo_Save.2003-05-19_02.15.06 Error: bnet.c:310 Write error
73 sending to Storage daemon:babylon5:9103: ERR=Broken pipe
76 - ChangeServiceConfig2A does not exist on WinNT (ADVAPI32.DLL).
78 - Implement MTIOCERRSTAT on FreeBSD to clear tape error conditions.
80 - Add Progress command that periodically reports the progress of
82 - Implement "Reschedule OnError=yes interval=nnn times=xxx"
83 - One block was orphaned in the SD probably after cancel.
84 - Add all command line arguments to "update", e.g. slot=nn volStatus=append, ...
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.
92 - Fix "access not allowed" for backup of files on WinXP.
93 - Check for existence of all new Win32 API's. See LoadLibrary in
95 - Examine Bare Metal restore problem (a FD crash exists somewhere ...).
96 - Test multiple simultaneous Volumes
97 - Document FInclude ...
98 - Implement timeout in response() when it should come quickly.
99 - Implement console @echo command.
100 - Bug: fix access problems on files restored on WinXP.
101 - Implement a Slot priority (loaded/not loaded).
102 - Implement "vacation" Incremental only saves.
103 - Implement single pane restore (much like the Gftp panes).
104 - Implement Automatic Mount even in operator wait.
105 - Implement create "FileSet"?
106 - Implement Release Device in the Job resource to unmount a drive.
107 - Implement Acquire Device in the Job resource to mount a drive,
108 be sure this works with admin jobs so that the user can get
109 prompted to insert the correct tape. Possibly some way to say to
110 run the job but don't save the files.
111 - Implement all command line args on run.
112 - Implement command line "restore" args.
113 - Implement "restore current select=no"
114 - Fix watchdog pthread crash on Win32 (this is pthread_kill() Cygwin bug)
115 - Implement "scratch pool" where tapes are defined and can be
116 taken by any pool that needs them.
117 - Implement restore "current system", but take all files without
118 doing selection tree -- so that jobs without File records can
120 - Make | and < work on FD side.
121 - Pass prefix_links to FD.
122 - Implement a M_SECURITY message class.
123 - Implement disk spooling. Two parts: 1. Spool to disk then
124 immediately to tape to speed up tape operations. 2. Spool to
125 disk only when the tape is full, then when a tape is hung move
127 - From Phil Stracchino:
128 It would probably be a per-client option, and would be called
129 something like, say, "Automatically purge obsoleted jobs". What it
130 would do is, when you successfully complete a Differential backup of a
131 client, it would automatically purge all Incremental backups for that
132 client that are rendered redundant by that Differential. Likewise,
133 when a Full backup on a client completed, it would automatically purge
134 all Differential and Incremental jobs obsoleted by that Full backup.
135 This would let people minimize the number of tapes they're keeping on
136 hand without having to master the art of retention times.
137 - Implement a relocatable bacula.spec
138 - Allow multiple Storage specifications (or multiple names on
139 a single Storage specification) in the Job record. Thus a job
140 can be backed up to a number of storage devices.
141 - Implement dump/print label to UA
142 - Add prefixlinks to where or not where absolute links to FD.
143 - Look at Python for a Bacula scripting language -- www.python.org
144 - Issue message to mount a new tape before the rewind.
145 - Simplified client job initiation for portables.
146 - If SD cannot open a drive, make it periodically retry.
147 - Implement LabelTemplate (at least first cut).
148 - Add more of the config info to the tape label.
149 - Implement a Mount Command and an Unmount Command where
150 the users could specify a system command to be performed
151 to do the mount, after which Bacula could attempt to
152 read the device. This is for Removeable media such as a CDROM.
153 - Most likely, this mount command would be invoked explicitly
154 by the user using the current Console "mount" and "unmount"
155 commands -- the Storage Daemon would do the right thing
156 depending on the exact nature of the device.
157 - As with tape drives, when Bacula wanted a new removable
158 disk mounted, it would unmount the old one, and send a message
159 to the user, who would then use "mount" as described above
160 once he had actually inserted the disk.
162 - Make some way so that if a machine is skipped because it is not up
163 that Bacula will continue retrying for a specified period of time --
165 - If tape is marked read-only, then try opening it read-only rather than
166 failing, and remember that it cannot be written.
167 - Refine SD waiting output:
168 Device is being positioned
169 > Device is being positioned for append
170 > Device is being positioned to file x
172 - Figure out some way to estimate output size and to avoid splitting
173 a backup across two Volumes -- this could be useful for writing CDROMs
174 where you really prefer not to have it split -- not serious.
175 - Add RunBeforeJob and RunAfterJob to the Client program.
176 - Have SD compute MD5 or SHA1 and compare to what FD computes.
177 - Make VolumeToCatalog calculate an MD5 or SHA1 from the
178 actual data on the Volume and compare it.
179 - Implement FileOptions (see end of this document)
180 - Implement Bacula plugins -- design API
181 - Make bcopy read through bad tape records.
182 - Fix read_record to handle multiple sessions.
183 - Program files (i.e. execute a program to read/write files).
184 Pass read date of last backup, size of file last time.
185 - Add Signature type to File DB record.
186 - Make Restore report an error if FD or SD term codes are not OK.
187 - CD into subdirectory when open()ing files for backup to
188 speed up things. Test with testfind().
189 - Priority job to go to top of list.
190 - Find out why Full saves run slower and slower (hashing?)
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 - Prohibit backing up archive device (findlib/find_one.c:128)
196 - Need a verbose mode in restore, perhaps to bsr.
197 - bscan without -v is too quiet -- perhaps show jobs.
198 - Add code to reject whole blocks if not wanted on restore.
199 - Start working on Base jobs.
200 - Check if we can increase Bacula FD priorty in Win2000
201 - Make sure the MaxVolFiles is fully implemented in SD
202 - Check if both CatalogFiles and UseCatalog are set to SD.
203 - Need return status on read_cb() from read_records(). Need multiple
204 records -- one per Job, maybe a JCR or some other structure with
205 a block and a record.
206 - Figure out how to do a bare metal Windows restore
207 - Put system type returned by FD into catalog.
208 - Possibly add email to Watchdog if drive is unmounted too
209 long and a job is waiting on the drive.
210 - Use read_record.c in SD code.
211 - Why don't we get an error message from Win32 FD when bootstrap
212 file cannot be created for restore command?
213 - When Marking a file in Restore that is a hard link, also
214 mark the link so that the data will be reloaded.
215 - Restore program that errors in SD due to no tape reports
216 OK incorrectly in output.
217 - After unmount, if restore job started, ask to mount.
218 - Convert all %x substitution variables, which are hard to remember
219 and read to %(variable-name). Idea from TMDA.
220 - Add JobLevel in FD status (but make sure it is defined).
221 - Make Pool resource handle Counter resources.
222 - Remove NextId for SQLite. Optimize.
223 - Move all SQL statements into a single location.
224 - Add UA rc and history files.
225 - put termcap (used by console) in ./configure and
226 allow -with-termcap-dir.
227 - Enhance time and size scanning routines.
228 - Fix Autoprune for Volumes to respect need for full save.
229 - Fix Win32 config file definition name on /install
230 - No READLINE_SRC if found in alternate directory.
231 - Test a second language e.g. french.
232 - Compare tape to Client files (attributes, or attributes and data)
233 - Make all database Ids 64 bit.
234 - Write an applet for Linux.
235 - Add estimate to Console commands
236 - Find solution to blank filename (i.e. path only) problem.
237 - Implement new daemon communications protocol.
238 - Remove PoolId from Job table, it exists in Media.
239 - Allow console commands to detach or run in background.
240 - Fix status delay on storage daemon during rewind.
241 - Add SD message variables to control operator wait time
242 - Maximum Operator Wait
243 - Minimum Message Interval
244 - Maximum Message Interval
245 - Send Operator message when cannot read tape label.
246 - Verify level=Volume (scan only), level=Data (compare of data to file).
247 Verify level=Catalog, level=InitCatalog
249 - Add keyword search to show command in Console.
250 - Fix Win2000 error with no messages during startup.
251 - Events : tape has more than xxx bytes.
252 - Restrict characters permitted in a Resource name.
253 - Complete code in Bacula Resources -- this will permit
254 reading a new config file at any time.
255 - Handle ctl-c in Console
256 - Implement script driven addition of File daemon to config files.
257 - Think about how to make Bacula work better with File (non-tape) archives.
258 - Write Unix emulator for Windows.
259 - Implement new serialize subroutines
260 send(socket, "string", &Vol, "uint32", &i, NULL)
261 - Audit all UA commands to ensure that we always prompt where possible.
262 - If ./btape is called without /dev, assume argument is a Storage resource name.
263 - Put memory utilization in Status output of each daemon
264 if full status requested or if some level of debug on.
265 - Make database type selectable by .conf files i.e. at runtime
266 - Set flag for uname -a. Add to Volume label.
267 - Implement throttled work queue.
268 - Check for EOT at ENOSPC or EIO or ENXIO (unix Pc)
269 - Restore files modified after date
270 - Restore file modified before date
271 - Emergency restore info:
273 - Backup working directory
275 - Restore -- do nothing but show what would happen
276 - SET LD_RUN_PATH=$HOME/mysql/lib/mysql
277 - Implement Restore FileSet=
278 - Create a protocol.h and protocol.c where all protocol messages
280 - Remove duplicate fields from jcr (e.g. jcr.level and jcr.jr.Level, ...).
281 - Timout a job or terminate if link goes down, or reopen link and query.
282 - Find general solution for sscanf size problems (as well
283 as sprintf. Do at run time?
284 - Concept of precious tapes (cannot be reused).
285 - Make bcopy copy with a single tape drive.
286 - Permit changing ownership during restore.
288 - Autolabel should be specified by DIR instead of SD.
289 - Find out how to get the system tape block limits, e.g.:
290 Apr 22 21:22:10 polymatou kernel: st1: Block limits 1 - 245760 bytes.
291 Apr 22 21:22:10 polymatou kernel: st0: Block limits 2 - 16777214 bytes.
294 - AutoScan (check checksum of tape)
295 - Format command = "format /dev/nst0"
299 - Seek resolution (usually corresponds to buffer size)
300 - EODErrorCode=ENOSPC or code
301 - Partial Read error code
302 - Partial write error code
303 - Nonformatted read error
304 - Nonformatted write error
305 - WriteProtected error
309 - IgnoreCloseErrors=yes
319 - FD sends unsaved file list to Director at end of job (see
321 - File daemon should build list of files skipped, and then
322 at end of save retry and report any errors.
323 - Write a Storage daemon that uses pipes and
324 standard Unix programs to write to the tape.
326 - Need something that monitors the JCR queue and
327 times out jobs by asking the deamons where they are.
328 - Enhance Jmsg code to permit buffering and saving to disk.
329 - device driver = "xxxx" for drives.
330 - restart: paranoid: read label fsf to
331 eom read append block, and go
332 super-paranoid: read label, read all files
333 in between, read append block, and go
334 verify: backspace, read append block, and go
335 permissive: same as above but frees drive
336 if tape is not valid.
338 - Ensure that /dev/null works
339 - Need report class for messages. Perhaps
340 report resource where report=group of messages
341 - enhance scan_attrib and rename scan_jobtype, and
342 fill in code for "since" option
343 - Director needs a time after which the report status is sent
344 anyway -- or better yet, a retry time for the job.
345 Don't reschedule a job if previous incarnation is still running.
346 - Some way to automatically backup everything is needed????
347 - Need a structure for pending actions:
349 - termination status (part of buffered msgs?)
351 Read, Write, Clean, Delete
352 - Login to Bacula; Bacula users with different permissions:
353 owner, group, user, quotas
354 - Store info on each file system type (probably in the job header on tape.
355 This could be the output of df; or perhaps some sort of /etc/mtab record.
358 - Design at hierarchial storage for Bacula. Migration and Clone.
359 - Implement FSM (File System Modules).
360 - Identify unchanged or "system" files and save them to a
361 special tape thus removing them from the standard
362 backup FileSet -- BASE backup.
363 - Audit M_ error codes to ensure they are correct and consistent.
364 - Add variable break characters to lex analyzer.
365 Either a bit mask or a string of chars so that
366 the caller can change the break characters.
367 - Make a single T_BREAK to replace T_COMMA, etc.
368 - Ensure that File daemon and Storage daemon can
369 continue a save if the Director goes down (this
370 is NOT currently the case). Must detect socket error,
371 buffer messages for later.
372 - Enhance time/duration input to allow multiple qualifiers e.g. 3d2h
373 - Add ability to backup to two Storage devices (two SD sessions) at
374 the same time -- e.g. onsite, offsite.
375 - Add the ability to consolidate old backup sets (basically do a restore
376 to tape and appropriately update the catalog). Compress Volume sets.
377 Might need to spool via file is only one drive is available.
378 - Compress or consolidate Volumes of old possibly deleted files. Perhaps
379 someway to do so with every volume that has less than x% valid
383 Migration: Move a backup from one Volume to another
384 Clone: Copy a backup -- two Volumes
386 Bacula Migration is based on Jobs (apparently Networker is file by file).
388 Migration triggered by:
392 Highwater mark (keep total size)
396 Bacula Projects Roadmap
398 last update 8 May 2003
400 Item 1: Multiple simultaneous Jobs. (done)
401 Done -- Restore part needs better implementation to work correctly
402 Also, it needs considerable testing
404 What: Permit multiple simultaneous jobs in Bacula.
406 Why: An enterprise level solution needs to go fast without the
407 need for the system administrator to carefully tweak
408 timing. Based on the benchmarks, during a full
409 backup, NetWorker typically hit 10 times the bandwidth to
410 the tape compared to Bacula--largely. This is probably due to
411 running parallel jobs and multi-threaded filling of buffers
412 and writing them to tape. This should also make things work
413 better when you have a mix of fast and slow machines backing
416 Notes: Bacula was designed to run multiple simultaneous jobs. Thus
417 implementing this is a matter of some small cleanups and
421 Item 2: Make the Storage daemon use intermediate file storage to buffer data.
422 Deferred -- not necessary yet.
424 What: If data is coming into the SD too fast, buffer it to
425 disk if the user has configured this option.
427 Why: This would be nice, especially if it more or less falls out
428 when implementing (1) above. If not, it probably should not
429 be given a high priority because fundamentally the backup time
430 is limited by the tape bandwidth. Even though you may finish a
431 client job quicker by spilling to disk, you still have to
432 eventually get it onto tape. If intermediate disk buffering
433 allows us to improve write bandwidth to tape, it may make
436 Notes: Whether or not this is implemented will depend upon performance
437 testing after item 1 is implemented.
440 Item 3: Write the bscan program -- also write a bcopy program.
443 What: Write a program that reads a Bacula tape and puts all the
444 appropriate data into the catalog. This allows recovery
445 from a tape that is no longer in the database, or it allows
446 re-creation of a database if lost.
448 Why: This is a fundamental robustness and disaster recovery tool
449 which will increase the comfort level of a sysadmin
450 considering adopting Bacula.
452 Notes: A skeleton of this program already exists, but much work
453 needs to be done. Implementing this will also make apparent
454 any deficiencies in the current Bacula tape format.
457 Item 4: Implement Base jobs.
459 What: A base job is sort of like a Full save except that you
460 will want the FileSet to contain only files that are unlikely
461 to change in the future (i.e. a snapshot of most of your
462 system after installing it). After the base job has been run,
463 when you are doing a Full save, you can specify to exclude
464 all files saved by the base job that have not been modified.
466 Why: This is something none of the competition does, as far as we know
467 (except BackupPC, which is a Perl program that saves to disk
468 only). It is big win for the user, it makes Bacula stand out
469 as offering a unique optimization that immediately saves time
472 Notes: Big savings in tape usage. Will require more resources because
473 the e. DIR must send FD a list of files/attribs, and the FD must
474 search the list and compare it for each file to be saved.
477 Item 5: Implement Label templates
479 What: This is a mechanism whereby Bacula can automatically create
480 a tape label for new tapes according to a detailed specification
481 provided by the user.
483 Why: It is a major convenience item for folks who use automated label
486 Notes: Bacula already has a working form of automatic tape label
487 creation, but it is very crude. The design for the complete
488 tape labeling project is already documented in the manual.
491 Item 6: Write a regression script.
492 Done -- Continue to expand its testing.
494 What: This is an automatic script that runs and tests as many features
495 of Bacula as possible. The output is compared to previous
496 versions of Bacula and any differences are reported.
498 Why: This is an enormous help in preventing introduction of new
499 errors in parts of the program that already work correctly.
501 Notes: This probably should be ranked higher, it's something the typical
502 user doesn't see. Depending on how it's implemented, it may
503 make sense to defer it until the archival tape format and
504 user interface mature.
507 Item 7: GUI for interactive restore
508 Item 8: GUI for interactive backup
510 What: The current interactive restore is implemented with a tty
511 interface. It would be much nicer to be able to "see" the
512 list of files backed up in typical GUI tree format.
513 The same mechanism could also be used for creating
514 ad-hoc backup FileSets (item 8).
516 Why: Ease of use -- especially for the end user.
518 Notes: Rather than implementing in Gtk, we probably should go directly
519 for a Browser implementation, even if doing so meant the
520 capability wouldn't be available until much later. Not only
521 is there the question of Windows sites, most
522 Solaris/HP/IRIX, etc, shops can't currently run Gtk programs
523 without installing lots of stuff admins are very wary about.
524 Real sysadmins will always use the command line anyway, and
525 the user who's doing an interactive restore or backup of his
526 own files will in most cases be on a Windows machine running
530 Item 9: Add SSL to daemon communications.
531 Inprogress as of version 1.31.
533 What: This provides for secure communications between the daemons.
535 Why: This would allow doing backup across the Internet without
536 privacy concerns (or with much less concern).
538 Notes: The vast majority of near term potential users will be backing up
539 a single site over a LAN and, correctly or not, they probably
540 won't be concerned with security, at least not enough to go to
541 the trouble to set up keys, etc. to screw things down. We suspect
542 that many users genuinely interested in multi-site backup
543 already run some form of VPN software in their internetwork
544 connections, and are willing to delegate security to that layer.
547 Item 10: Define definitive tape format.
550 What: Define that definitive tape format that will not change
551 for the next millennium.
553 Why: Stability, security.
555 Notes: See notes for item 11 below.
558 Item 11: New daemon communication protocol.
560 What: The current daemon to daemon protocol is basically an ASCII
561 printf() and sending the buffer. On the receiving end, the
562 buffer is sscanf()ed to unpack it. The new scheme would
563 be a binary format that allows quick packing and unpacking
564 of any data type with named fields.
566 Why: Using binary packing would be faster. Named fields will permit
567 error checking to ensure that what is sent is what the
568 receiver really wants.
570 Notes: These are internal improvements in the interest of the
571 long-term stability and evolution of the program. On the one
572 hand, the sooner they're done, the less code we have to rip
573 up when the time comes to install them. On the other hand, they
574 don't bring an immediately perceptible benefit to potential
575 users. Item 10 and possibly item 11 should be deferred until Bacula
576 is well established with a growing user community more or
577 less happy with the feature set. At that time, it will make a
578 good "next generation" upgrade in the interest of data
584 ======================================================
586 It is somewhat like a Full save becomes an incremental since
587 the Base job (or jobs) plus other non-base files.
589 - New BaseFile table that contains:
590 JobId, BaseJobId, FileId (from Base).
591 i.e. for each base file that exists but is not saved because
592 it has not changed, the File daemon sends the JobId, BaseId,
593 and FileId back to the Director who creates the DB entry.
594 - To initiate a Base save, the Director sends the FD
595 the FileId, and full filename for each file in the Base.
596 - When the FD finds a Base file, he requests the Director to
597 send him the full File entry (stat packet plus MD5), or
598 conversely, the FD sends it to the Director and the Director
599 says yes or no. This can be quite rapid if the FileId is kept
600 by the FD for each Base Filename.
601 - It is probably better to have the comparison done by the FD
602 despite the fact that the File entry must be sent across the
604 - An alternative would be to send the FD the whole File entry
605 from the start. The disadvantage is that it requires a lot of
606 space. The advantage is that it requires less communications
608 - The Job record must be updated to indicate that one or more
610 - At end of Job, FD returns:
611 1. Count of base files/bytes not written to tape (i.e. matches)
612 2. Count of base file that were saved i.e. had changed.
613 - No tape record would be written for a Base file that matches, in the
614 same way that no tape record is written for Incremental jobs where
615 the file is not saved because it is unchanged.
616 - On a restore, all the Base file records must explicitly be
617 found from the BaseFile tape. I.e. for each Full save that is marked
618 to have one or more Base Jobs, search the BaseFile for all occurrences
620 - An optimization might be to make the BaseFile have:
626 This would avoid the need to explicitly fetch each File record for
627 the Base job. The Base Job record will be fetched to get the
628 VolSessionId and VolSessionTime.
629 =========================================================
633 =============================================================
635 Request For Comments For File Backup Options
638 Subject: File Backup Options
641 A few days ago, a Bacula user who is backing up to file volumes and
642 using compression asked if it was possible to suppress compressing
643 all .gz files since it was a waste of CPU time. Although Bacula
644 currently permits using different options (compression, ...) on
645 a directory by directory basis, it cannot do it on a file by
646 file basis, which is clearly what was desired.
648 Proposed Implementation:
649 To solve this problem, I propose the following:
651 - Add a new Director resource type called FileOptions.
653 - The FileOptions resource will have records for all
654 options that can currently be specified on the Include record
655 (in a FileSet). Examples below.
657 - The FileOptions resource will permit an exclude option as well
658 as a number of additional options.
660 - The heart of the FileOptions resource is the ability to
661 supply any number of ApplyTo records which specify POSIX
662 regular expressions. These ApplyTo regular expressions are
663 applied to the fully qualified filename (path and all). If
664 one matches, then the FileOptions will be used.
666 - When an ApplyTo specification matches an included file, the
667 options specified in the FileOptions resource will override
668 the default options specified on the Include record.
670 - Include records will be modified to permit referencing one or
671 more FileOptions resources. The FileOptions will be used
672 in the order listed on the Include record and the first
673 one that matches will be applied.
675 - Options (or specifications) currently supplied on the Include
676 record will be deprecated (i.e. removed in a later version a
677 year or so from now).
679 - The Exclude record will be deprecated as the same functionality
680 can be obtained by using an Exclude = yes in the FileOptions.
683 The following records can appear in the FileOptions resource. An
684 asterisk preceding the name indicates a feature not currently
688 - Compression= (GZIP, ...)
689 - Signature= (MD5, SHA1, ...)
691 - OneFs= (yes/no) - remain on one filesystem
692 - Recurse= (yes/no) - recurse into subdirectories
693 - Sparse= (yes/no) - do sparse file backup
694 - *Exclude= (yes/no) - exclude file from being saved
695 - *Reader= (filename) - external read (backup) program
696 - *Plugin= (filename) - read/write plugin module
699 - verify= (ipnougsamc5) - verify options
702 - replace= (always/ifnewer/ifolder/never) - replace options currently
704 - *Writer= (filename) - external write (restore) program
708 Currently options specifying compression, MD5 signatures, recursion,
709 ... of a FileSet are supplied on the Include record. These will now
710 all be collected into a FileOptions resource, which will be
711 specified on the Include in place of the options. Multiple FileOptions
712 may be specified. Since the FileOptions contain regular expressions
713 that are applied to the full filename, this will give the ability
714 to specify backup options on a file by file basis to whatever level
723 Include = compression=GZIP signature=MD5 {
732 Include = FileOptions=Opts {
743 That's a lot more to do the same thing, but it gives the ability to
744 apply options on a file by file basis. For example, suppose you
745 want to compress all files but not any file with extensions .gz or .Z.
746 You could do so as follows:
750 Include = FileOptions=NoCompress FileOptions=Opts {
758 ApplyTo = /*.?*/ # matches all files
763 # Note multiple ApplyTos are ORed
764 ApplyTo = /*.gz/ # matches .gz files */
765 ApplyTo = /*.Z/ # matches .Z files */
768 Now, since the NoCompress FileOptions is specified first on the
769 Include line, any *.gz or *.Z file will have an MD5 signature computed,
770 but will not be compressed. For all other files, the NoCompress will not
771 match, so the Opts options will be used which will include GZIP
775 - Is it necessary to provide some means of ANDing regular expressions
776 and negation? (not currently planned)
778 e.g. ApplyTo = /*.gz/ && !/big.gz/
780 - I see that Networker has a "null" module which, if specified, does not
781 backup the file, but does make an record of the file in the catalog
782 so that the catalog will reflect an exact picture of the filesystem.
783 The result is that the file can be "seen" when "browsing" the save
784 sets, but it cannot be restored.
786 Is this really useful? Should it be implemented in Bacula?
789 After implementing the above, the user will be able to specify
790 on a file by file basis (using regular expressions) what options are
791 applied for the backup.
794 =============================================
796 ==========================================================
798 For each Incremental job that is run, there may be files that
799 were found but not saved because they were locked (this applies
800 only to Windows). Such a system could send back to the Director
801 a list of Unsaved files.
803 - New UnSavedFiles table that contains:
807 - Then in the next Incremental job, the list of Unsaved Files will be
808 feed to the FD, who will ensure that they are explicitly chosen even
809 if standard date/time check would not have selected them.
810 =============================================================
813 Done: (see kernsdone for more)
814 - Heartbeat between daemons.
815 - Fix Dir heartbeat in restore and verify vol. Be sure to make
816 bnet_recv() ignore BNET_HEARTBEAT.
817 - Implement HEART_BEAT while SD waiting for tapes.
818 - Include RunBeforeJob and RunAfterJob output in the message
820 - Change M_INFO to M_RESTORED for all restored files.
821 - Fix command prompt in gnome-console by checking on Ready.
822 - Merge SQLite, MySQL, and Rel spec into a single file.
823 - Fix config of "console"
824 - Check if cancel works with FD (fixed).
825 - Properly configure console and gconsole (currently for source not
826 configured for installation).
827 - Error labeling tape from console gets Jmsg error because of no Job.
828 - Test and implement get_pint and get_yesno.
829 - Implement global with DB name and add to btraceback.gdb
830 - Remove subsysdir from conf files (used only in autostart scripts).
832 rufus-dir: Max configured use duration exceeded. Marking Volume "MatouBackup" as Used.
833 rufus-sd: Volume "" previously written, moving to end of data.
834 rufus-sd: Matou.2003-05-10_10.39.18 Error: I canot write on this volume because:
835 The number of files mismatch! Volume=1 Catalog=0
836 rufus-sd: Matou.2003-05-10_10.39.18 Error: askdir.c:155 NULL Volume name. This shouldn't happen!!!
837 - Shell character expansion is failing occassionally.
838 - Add a section to the doc on Manual cycling of Volumes.
839 - Check if Job/File retentions apply to multivolume jobs.
840 - Fix missing casette in autoloader during read:
841 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
842 14-May-2003 14:41 undef-sd: 3301 Issuing autochanger "loaded" command.
843 14-May-2003 14:41 undef-sd: 3302 Issuing autochanger "unload" command.
844 14-May-2003 14:42 undef-sd: 3303 Issuing autochanger "load slot 1" command.
845 14-May-2003 14:42 undef-sd: 3304 Autochanger "load slot 1" status is OK.
846 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
847 14-May-2003 14:42 undef-sd: 3301 Issuing autochanger "loaded" command.
848 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
849 14-May-2003 14:42 undef-sd: 3301 Issuing autochanger "loaded" command.
850 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
851 14-May-2003 14:42 undef-sd: 3301 Issuing autochanger "loaded" command.
852 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
853 14-May-2003 14:42 undef-sd: 3301 Issuing autochanger "loaded" command.
854 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".
855 14-May-2003 14:42 undef-dir: Bacula 1.31 (12May03): 14-May-2003 14:42
856 - Fix problem reported by Christopher McCurdy <cmccurdy@eecis.udel.edu>
857 xeon-fd: Could not stat c:/Documents and Settings/All
858 Users/Application Data/Humc:\Documents and Settings\All User98_AIX.kbf:
859 ERR=No such file or directory
861 - The following Re-read last block at EOT failed. ERR=block.c:523 Read zero bytes on device /dev/nrsa0.
862 undef-sd: block.c:523 Read zero bytes on device /dev/nrsa0.
863 apparently masks the standard EOM message.
864 - BSD (probably) does not have strtoll()
865 - BSD does not have ioctl() MTEOM
866 - BSD defines a number of MT_xxx variables which conflict
867 with those defined by Bacula.
868 - Make default duration days if no qualifier (e.g. s) is specified.
869 - BSDI fix finding gcc version
870 - When the FD errs (e.g. disk full) have a more graceful shutdown.