5 - Implement SIGHUP in Dir
6 (Implement resources on a single pointer)
7 - Make btape release the drive during the "test" append.
8 - Do tape alerts -- see tapealert.txt
9 - Revisit and revise Disaster Recovery (fix SCSI and RAID
11 - Document a get out of jail procedure if everything breaks if
12 you lost/broke the Catalog -- do the same for "I know my
13 file is there how do I get it back?".
14 - When restore started from console, report jobid.
15 - Win32 inc problem when new directory added.
17 The error I when I installed 1.34.2 clients:
19 linked to missing export KERNEL32.DLL:GetFileAttributesExA.
20 - Add link to doc in message when authentication fails.
21 - Add better error codes to run_program (10000+)
22 - When passwords do not match, print message that points the
24 - Make Verify jobs require exclusive use of Volume as Restore
28 Documentation to do: (any release a little bit at a time)
29 - Document query file format.
30 - Document problems with Verify and pruning.
31 - Document how to use multiple databases.
32 - VXA drives have a "cleaning required"
33 indicator, but Exabyte recommends preventive cleaning after every 75
36 In this context, it should be noted that Exabyte has a command-line
37 vxatool utility available for free download. (The current version is
38 vxatool-3.72.) It can get diagnostic info, read, write and erase tapes,
39 test the drive, unload tapes, change drive settings, flash new firmware,
41 Of particular interest in this context is that vxatool <device> -i will
42 report, among other details, the time since last cleaning in tape motion
43 minutes. This information can be retrieved (and settings changed, for
44 that matter) through the generic-SCSI device even when Bacula has the
45 regular tape device locked. (Needless to say, I don't recommend
46 changing tape settings while a job is running.)
47 - Lookup HP cleaning recommendations.
48 - Lookup HP tape replacement recommendations (see trouble shooting autochanger)
49 - Add more obvious documentation of @ for conf files.
50 - Document doing table repair
52 Testing to do: (painful)
54 - blocksize recognition code.
55 - Test if rewind at end of tape waits for tape to rewind.
58 For 1.33 Testing/Documentation:
59 - Figure out how to use ssh or stunnel to protect Bacula communications.
60 Add Dan's work to manual See ssl.txt
61 - Add db check test to regression. Test each function like delete,
63 - If you use restore replace=never, the directory attributes for
64 non-existent directories will not be restored properly.
67 - Sort Scheduled jobs status listing by start time.
68 - Add priority to Scheduled jobs status listing.
69 - Add multiple-media-types.txt
70 - look at mxt-changer.html
71 - Make ? do a help command (no return needed).
72 - Implement restore directory.
73 - Add All Local Partitions = yes to new style saves.
74 - Document streams and how to implement them.
75 - Possibly implement "Ensure Full Backup = yes" looks for a failed full backup
76 and upgrades the current backup if one exists.
77 - Check that barcode reading and update slots scan works.
78 - Try not to re-backup a file if a new hard link is added.
79 - Add feature to backup hard links only, but not the data.
80 - Add "All Local = yes" option to save to include all local partitions.
81 - Fix stream handling to be simpler.
82 - Add Priority and Bootstrap to Run a Job.
83 - Eliminate Restore "Run Restore Job" prompt by allowing new "run command
85 - Remove View FileSet button from Run a Job dialog.
86 - Don't count higher level non-saved directories in Files expected
88 - Handle prompt for restore job at end of Restore command.
89 - Add display of total selected files to Restore window.
90 - Add tree pane to left of window.
92 - Polling does not work for restore. It tries a number of times,
93 gives up, and crashes the FD.
94 - Lock jcr_chain when doing attach/detach in acquire.c
95 - Add assert in free_jcr if attach/detach chain active.
96 - Max wait time or max run time causes seg fault -- see runtime-bug.txt
97 - Document writing to a CD/DVD with Bacula.
98 - Add check for tape alerts.
99 - Add a "base" package to the window installer for pthreadsVCE.dll
100 which is needed by all packages.
101 - Add message to user to check for fixed block size when the forward
102 space test fails in btape.
103 - When unmarking a directory check if all files below are unmarked and
104 then remove the + flag -- in the restore tree.
105 - Possibly implement: Action = Unmount Device="TapeDrive1" in Admin jobs.
106 - Setup lrrd graphs: (http://www.linpro.no/projects/lrrd/) Mike Acar.
107 - Revisit the question of multiple Volumes (disk) on a single device.
108 - Finish SIGHUP work.
109 - Check that all change in wait status in the SD are
110 signaled to the Director.
111 - Add a block copy option to bcopy.
112 - Investigate adding Mac Resource Forks.
113 - Finish work on Gnome restore GUI.
114 - Fix "llist jobid=xx" where no fileset or client exists.
115 - Check pruning of restore jobs.
117 it seems to be complaining about 12:00pm which should be a valid 12
118 hour time. I changed the time to 11:59am and everything works fine.
119 Also 12:00am works fine. 0:00pm also works (which I don't think
120 should). None of the values 12:00pm - 12:59pm work for that matter.
121 - Add level to estimate command.
122 - For each job type (Admin, Restore, ...) require only the really necessary
124 - Fix option 2 of restore -- list where file is backed up -- require Client,
125 then list last 20 backups.
126 - Add all pools in Dir conf to DB also update them to catch changed
127 LabelFormats and such.
128 - Update volumes FromPool=xxx does all volumes.
129 - Pass Director resource name as an option to the Console.
130 - Add a "batch" mode to the Console (no unsolicited queries, ...).
131 - Add code to check for tape alerts -- tapeinfo.
132 - Make sure list of Volumes needed is in correct order for restore.
134 - Remove paths (and files that reference them) that have no trailing slash
135 in dbcheck -- or add a trailing slash.
136 - Remove Filenames (and files that reference them) that have a trailing
137 slash in dbcheck -- or remove the trailing slash.
138 - Remove orphaned paths/filenames by copying them to a new table with a
139 reference count, then mark all referenced files/paths and remove unreferenced
141 - Add a .list all files in the restore tree (probably also a list all files)
142 Do both a long and short form.
143 - Allow browsing the catalog to see all versions of a file (with
144 stat data on each file).
145 - Restore attributes of directory if replace=never set but directory
147 - Allow "delete job jobid=xxx,yyy,aaa-bbb" i.e. list + ranges.
148 - Use SHA1 on authentication if possible.
149 - See comtest-xxx.zip for Windows code to talk to USB.
150 - Make btape accept Device Names in addition to Archive names.
151 - Add Events and Perl scripting.
152 - Add John's appended files:
153 Appended = { /files/server/logs/http/*log }
154 and such files would be treated as follows.On a FULL backup, they would
155 be backed up like any other file.On an INCREMENTAL backup, where a
156 previous INCREMENTAL or FULL was already in thecatalogue and the length
157 of the file wasgreater than the length of the last backup, only thedata
158 added since the last backup will be dumped.On an INCREMENTAL backup, if
159 the length of the file is less than thelength of the file with the same
160 name last backed up, the completefile is dumped.On Windows systems, with
161 creation date of files, we can be evensmarter about this and not count
162 entirely upon the length.On a restore, the full and all incrementals
163 since it will beapplied in sequence to restore the file.
164 - Add a regression test for dbcheck.
165 - Add disk seeking on restore. - Allow
166 for optional cancelling of SD and FD in case DIR
167 gets a fatal error. Requested by Jesse Guardiani <jesse@wingnet.net>
168 - Bizarre message: Error: Could not open WriteBootstrap file:
169 - Build console in client only build.
170 - Add "limit=n" for "list jobs"
171 - Check new HAVE_WIN32 open bits.
172 - Check if the tape has moved before writing.
173 - Handling removable disks -- see below:
174 - Multiple drive autochanger support -- see below.
175 - Keep track of tape use time, and report when cleaning is necessary.
176 - Fix FreeBSD mt_count problem.
177 - Add FromClient and ToClient keywords on restore command (or
178 BackupClient RestoreClient).
179 - Automatic "update slots" on user configuration directive when a
181 - Implement a JobSet, which groups any number of jobs. If the
182 JobSet is started, all the jobs are started together.
183 Allow Pool, Level, and Schedule overrides.
184 - Enhance cancel to timeout BSOCK packets after a specific delay.
185 - When I restore to Windows the Created, Accessed and Modifiedtimes are
186 those of the time of the restore, not those of the originalfile.
187 The dates you will find in your restore log seem to be the original
189 - Volume "add"ed to Pool gets recycled in first use. VolBytes=0
190 - If a tape is recycled while it is mounted, Stanislav Tvrudy must do an
191 additional mount to deblock the job.
193 bscan: sql_update.c:65 UPDATE File SET MD5='Ij+5kwN6TFIxK+8l8+/I+A' WHERE FileId=0
194 bscan: bscan.c:1074 Could not add MD5/SHA1 to File record. ERR=sql_update.c:65 Update problem: affected_rows=0
195 - Do scheduling by UTC using gmtime_r() in run_conf, scheduler, and
196 ua_status.!!! Thanks to Alan Brown for this tip.
197 - Look at updating Volume Jobs so that Max Volume Jobs = 1 will work
198 correctly for multiple simultaneous jobs.
199 - Correct code so that FileSet MD5 is calculated for < and | filename
201 - Mark Volume in error on error from WEOF.
202 - Implement the Media record flag that indicates that the Volume does disk
204 - Implement VolAddr, which is used when Volume is addressed like a disk,
205 and form it from VolFile and VolBlock.
206 - Make multiple restore jobs for multiple media types specifying
207 the proper storage type.
208 - Implement MediaType keyword in bsr?
209 - Fix fast block rejection (stored/read_record.c:118). It passes a null
210 pointer (rec) to try_repositioning().
211 - Look at extracting Win data from BackupRead.
212 - Having dashes in filenames apparently creates problems for restore
213 by filename??? hard to believe.
214 - Implement RestoreJobRetention? Maybe better "JobRetention" in a Job,
215 which would take precidence over the Catalog "JobRetention".
216 - Implement Label Format in Add and Label console commands.
217 - Possibly up network buffers to 65K. Put on variable.
218 - Put email tape request delays on one or more variables. User wants
219 to cancel the job after a certain time interval. Maximum Mount Wait?
220 - Job, Client, Device, Pool, or Volume?
221 Is it possible to make this a directive which is *optional* in multiple
222 resources, like Level? If so, I think I'd make it an optional directive
223 in Job, Client, and Pool, with precedence such that Job overrides Client
224 which in turn overrides Pool.
225 - Print a message when a job starts if the conf file is not current.
227 - Spooling ideas taken from Volker Sauer's and other's emails:
228 > IMHO job spooling should be turned on
234 > where and 2) overrides 1) and 3) is independent.
236 Yes, this is the minimum that I think is necessary.
240 > It should be switched by job, because the job that backs up the machine
241 > with the bacula-sd on doesn't need spooling.
242 > It should be switched by schedule, because for full-backups I don't need
243 > spooling, so I can switch it off (because the network faster then the
246 True, with the exception that if you have enough disk spool space,
247 and you want to run concurrent jobs, spooling can eliminate the block
248 interleaving restore inefficiencies.
250 > And you should be able to turn it of by sd for sd-machines with low disk
251 > capacity or if you just don't need or want this feature.
254 > - definitly the possibility for multipe spool direcories
256 Having multiple directories is no problem -- having different maximum
257 sizes creates specification problems. At some point, I will probably
258 have a common SD pool of spool directories as well as a set of
259 private spool directories for each device. The first implementation
260 will be a set of private spool directories for each device since
261 managing a global pool with a bunch of threads writing into the same
262 directory is *much* more complicated and prone to error.
264 > - the ability to spool parts of a backup (not the whole client)
266 This may change in the future, but for the moment, it will spool
267 either to a job high water mark, or until the directory is full
268 (reaches max spool size or I/O error). It will then write to tape,
269 truncate the spool file, and begin spooling again.
271 > - spooling while writing to tape
273 Not within a job, but yes, if you run concurrent jobs -- each is a
274 different thread. Within a job could be a feature, but *much* later.
276 > - parallel spooling (like parallel jobs/ concurrent jobs) of clients
278 Yes, this is one of my main motivations for doing it (aside from
279 eliminating tape "shoe shine" during incremental backups.
281 > - flushing a backup that only went to disk (like amflush in amanda)
283 This will be a future feature, since spooling is different from backing
284 up to disk. The future feature will be "migration" which will move a job
285 from one backup Volume to another.
287 - New Storage specifications:
288 Passed to SD as a sort of BSR record called Storage Specification
292 MediaType -> Next MediaType
294 Device -> Next Device
295 Write Copy Resource that makes a copy of a resource.
297 Allow multiple Storage specifications
305 Allow Multiple Pool specifications (note, Pool currently
307 Allow Multiple MediaType specifications
308 Allow Multiple Device specifications
309 Perhaps keep this in a single SSR
310 Tie a Volume to a specific device by using a MediaType that
311 is contained in only one device.
312 In SD allow Device to have Multiple MediaTypes
315 - Look at www.nu2.nu/pebuilder as a helper for full windows
317 Ideas from Jerry Scharf:
318 First let's point out some big pluses that bacula has for this
320 more importantly it's active. Thank you so much for that
321 even more important, it's not flaky
322 it has an open access catalog, opening many possibilities
323 it's pushing toward heterogeneous systems capability
325 I don't remember an include file directive for config files
326 (not filesets, actual config directives)
327 can you check the configs without starting the daemon?
328 some warnings about possible common mistakes
330 doing the testing and blessing of concurrent backup writes
331 this is absolutely necessary in the enterprise
332 easy user recovery GUI with full access checking
333 Macintosh file client
334 macs are an interesting niche, but I fear a server is a rathole
335 working bare iron recovery for windows
336 much better handling on running config changes
337 thinking through the logic of what happens to jobs in progress
338 the option for inc/diff backups not reset on fileset revision
339 a) use both change and inode update time against base time
340 b) do the full catalog check (expensive but accurate)
341 sizing guide (how much system is needed to back up N systems/files)
342 consultants on using bacula in building a disaster recovery system
344 or how to get at fancy things that one could do with bacula
345 logwatch code for bacula logs (or similar)
346 linux distro inclusion of bacula (brings good and bad, but necessary)
347 win2k/XP server capability (icky but you asked)
348 support for Oracle database ??
350 - Look at adding SQL server and Exchange support for Windows.
351 - Restore: Enter Filename: 'C:/Documents and Settings/Comercial/My
352 Documents/MOP/formulário de registro BELAS ARTES.doc' causes Bacula to
354 - Each DVD-RAM disk would be a volume, just like each tape is
355 a volume. It's a 4.7GB media with random access, but there's nothing about
356 it that I can see that makes it so different than a tape from bacula's
357 perspective. Why couldn't I back up to a bare floppy as a volume (ignoring
359 - Make dev->file and dev->block_num signed integers so that -1 can
360 be an invalid value which happens with BSR.
361 - Create VolAddr for disk files in place of VolFile and VolBlock. This
362 is needed to properly specify ranges.
363 - Print bsmtp output to job report so that problems will be seen.
364 - Pass the number of files to be restored to the FD for reporting
365 - Add progress of files/bytes to SD and FD.
366 - Don't continue Restore if no files selected.
367 - Print warning message if FileId > 4 billion
368 - do a "messages" before the first prompt in Console
369 - Add a date and time stamp at the beginning of every line in the
370 Job report (Volker Sauer).
371 - Client does not show busy during Estimate command.
372 - Implement Console mtx commands.
373 - Add a default DB password to MySQL.
374 GRANT all privileges ON bacula.* TO bacula@localhost IDENTIFIED BY
377 - Implement a Mount Command and an Unmount Command where
378 the users could specify a system command to be performed
379 to do the mount, after which Bacula could attempt to
380 read the device. This is for Removeable media such as a CDROM.
381 - Most likely, this mount command would be invoked explicitly
382 by the user using the current Console "mount" and "unmount"
383 commands -- the Storage Daemon would do the right thing
384 depending on the exact nature of the device.
385 - As with tape drives, when Bacula wanted a new removable
386 disk mounted, it would unmount the old one, and send a message
387 to the user, who would then use "mount" as described above
388 once he had actually inserted the disk.
389 - Implement dump/print label to UA
390 - Implement disk spooling. Two parts: 1. Spool to disk then
391 immediately to tape to speed up tape operations. 2. Spool to
392 disk only when the tape is full, then when a tape is hung move
394 - Scratch Pool where the volumes can be re-assigned to any Pool.
395 - bextract is sending everything to the log file ****FIXME****
396 - Add Progress command that periodically reports the progress of
398 - Restrict characters permitted in a Resource name, and don't permit
400 - Allow multiple Storage specifications (or multiple names on
401 a single Storage specification) in the Job record. Thus a job
402 can be backed up to a number of storage devices.
403 - Implement some way for the File daemon to contact the Director
404 to start a job or pass its DHCP obtained IP number.
405 - Implement multiple Consoles.
406 - Implement a query tape prompt/replace feature for a console
408 Two jobs ready to go, first one blocked waiting for media
409 Cancel 2nd job ("waiting execution" one)
412 - Copy console @ code to gnome2-console
413 - Make AES the only encryption algorithm see
414 http://csrc.nist.gov/CryptoToolkit/aes/). It's
415 an officially adopted standard, has survived peer
416 review, and provides keys up to 256 bits.
417 - Think about how space could be freed up on a tape -- perhaps this
418 is a Merge or Compact feature that is needed.
419 - Modify FileSet, did not upgrade the current Increment job, but
420 waited for the next job to be upgraded.
421 - Take a careful look at SetACL http://setacl.sourceforge.net
422 - Implement a where command for the tree telling where a file
424 - Take a careful look at Level for the estimate command, maybe make
425 it a command line option.
426 - Add Volume name to "I cannot write on this volume because"
427 - Make tree walk routines like cd, ls, ... more user friendly
428 by handling spaces better.
429 - Write your PID file and chown root:wheel before drop.
430 - Make sure there is no symlink in a file before creating a
432 - Look at mktemp or mkstemp(3).
433 mktemp and mkstemp create files with predictable names too. That's
434 not the vulnerability. The vulnerability is in creating files without
435 using the O_EXCL flag, which means "only create this file if it doesn't
436 exist, including if the file is a dangling symlink."
438 It is *NOT* enough to do the equivalent of
440 if doesn't exist $filename
441 then create $filename
443 because between the test and the create another process could have
444 gotten the CPU and created the file. You must use atomic functions
445 (those that don't get interrupted by other processes) and O_EXCL is
446 the only way for this particular example.
447 - Automatically create pools, but instead of looking for what
448 in in Job records, walk through the pool resources.
449 - Check and double check tree code, why does it take so long?
450 - Add device name to "Current Volume not acceptable because ..."
451 - Make sure that Bacula rechecks the tape after the 20 min wait.
452 - Set IO_NOWAIT on Bacula TCP/IP packets.
453 - Try doing a raw partition backup and restore by mounting a
456 Yes, it would allow to highly automatic the request for new tapes. If a
457 tape is empty, bacula reads the barcodes (native or simulated), and if
458 an unused tape is found, it runs the label command with all the
459 necessary parameters.
461 By the way can bacula automatically "move" an empty/purged volume say
462 in the "short" pool to the "long" pool if this pool runs out of volume
464 - Eliminate orphaned jobs: dbcheck, normal pruning, delete job command.
465 Hm. Well, there are the remaining orphaned job records:
467 | 105 | Llioness Save | 0000-00-00 00:00:00 | B | D | 0 | 0 | f |
468 | 110 | Llioness Save | 0000-00-00 00:00:00 | B | I | 0 | 0 | f |
469 | 115 | Llioness Save | 2003-09-10 02:22:03 | B | I | 0 | 0 | A |
470 | 128 | Catalog Save | 2003-09-11 03:53:32 | B | I | 0 | 0 | C |
471 | 131 | Catalog Save | 0000-00-00 00:00:00 | B | I | 0 | 0 | f |
473 As you can see, three of the five are failures. I already deleted the
474 one restore and one other failure using the by-client option. Deciding
475 what is an orphaned job is a tricky problem though, I agree. All these
476 records have or had 0 files/ 0 bytes, except for the restore. With no
477 files, of course, I don't know of the job ever actually becomes
478 associated with a Volume.
480 (I'm not sure if this is documented anywhere -- what are the meanings of
481 all the possible JobStatus codes?)
483 Looking at my database, it appears to me as though all the "orphaned"
484 jobs fit into one of two categories:
486 1) The Job record has a StartTime but no EndTime, and the job is not
489 2) The Job record has an EndTime, indicating that it completed, but
490 it has no associated JobMedia record.
493 This does suggest an approach. If failed jobs (or jobs that, for some
494 other reason, write no files) are associated with a volume via a
495 JobMedia record, then they should be purged when the associated volume
496 is purged. I see two ways to handle jobs that are NOT associated with a
499 1) purge them automatically whenever any volume is manually purged;
501 2) add an option to the purge command to manually purge all jobs with
502 no associated volume.
504 I think Restore jobs also fall into category 2 above .... so one might
505 want to make that "The Job record has an EndTime,, but no associated
506 JobMedia record, and is not a Restore job."
507 - make "btape /tmp" work.
508 - Make sure a rescheduled job is properly reported by status.
509 - Walk through the Pool records rather than the Job records
510 in dird.c to create/update pools.
511 - What to do about "list files job=xxx".
512 - When job rescheduled, status gives is waiting for Client Rufus
513 to connect to Storage File. Dir needs to inform SD that job
515 - Make Dmsg look at global before calling subroutine.
516 - Enable trace output at runtime for Win32
517 - Available volumes for autochangers (see patrick@baanboard.com 3 Sep 03
518 and 4 Sep) scan slots.
519 - Get and test MySQL 4.0
520 - Do a complete audit of all pthreads_mutex, cond, ... to ensure that
521 any that are dynamically initialized are destroyed when no longer used.
522 - Look at how fuser works and /proc/PID/fd that is how Nic found the
523 file descriptor leak in Bacula.
524 - Implement WrapCounters in Counters.
525 - Turn on SIGHUP in dird.c and test.
526 - Use system dependent calls to get more precise info on tape errors.
527 - Add heartbeat from FD to SD if hb interval expires.
528 - Suppress read error on blank tape when doing a label.
529 - Can we dynamically change FileSets?
530 - If pool specified to label command and Label Format is specified,
531 automatically generate the Volume name.
532 - Why can't SQL do the filename sort for restore?
533 - Look at libkse (man kse) for FreeBSD threading.
534 - Look into Microsoft Volume Shadowcopy Service VSS for backing
535 up system state components (Active Directory, System Volume, ...)
536 - Add ExhautiveRestoreSearch
537 - Look at the possibility of loading only the necessary
538 data into the restore tree (i.e. do it one directory at a
539 time as the user walks through the tree).
540 - Possibly use the hash code if the user selects all for a restore command.
541 - Orphaned Dir buffer at parse_conf.c:373 => store_dir
542 - Fix "restore all" to bypass building the tree.
543 - Prohibit backing up archive device (findlib/find_one.c:128)
544 - Implement Release Device in the Job resource to unmount a drive.
545 - Implement Acquire Device in the Job resource to mount a drive,
546 be sure this works with admin jobs so that the user can get
547 prompted to insert the correct tape. Possibly some way to say to
548 run the job but don't save the files.
549 - Implement FileOptions (see end of this document)
550 - Make things like list where a file is saved case independent for
553 - Bacula needs to propagate SD errors.
554 > > cluster-dir: Start Backup JobId 252, Job=REUTERS.2003-08-11_15.04.12
555 > > prod4-sd: REUTERS.2003-08-11_15.04.12 Error: Write error on device
556 /dev/nst0. ERR=Input/output error.
557 > > prod4-sd: REUTERS.2003-08-11_15.04.12 Error: Re-read of last block failed.
558 Last block=5162 Current block=5164.
559 > > prod4-sd: End of medium on Volume "REU007" Bytes=16,303,521,933
561 - Use autochanger to handle multiple devices.
562 - Add SuSE install doc to list.
563 - Check and rechedk "Invalid block number"
564 - Make bextract release the drive properly between tapes
565 so that an autochanger can be made to work.
566 - User wants to NOT backup up certain big files (email files).
567 - Maybe remove multiple simultaneous devices code in SD.
568 - On Windows with very long path names, it may be impossible to create
569 a file (and thus restore it) because the total length is too long.
570 We must cd into the directory then create the file without the
572 - lstat() is not going to work on Win32 for testing date.
573 - Implement a Recycle command
574 - Add client name to cram-md5 challenge so Director can immediately
575 verify if it is the correct client.
576 - Add JobLevel in FD status (but make sure it is defined).
577 - Audit all UA commands to ensure that we always prompt where possible.
578 - Check Jmsg in bnet, may not work, must dup bsock.
579 - Suppress Job Name in Jmsg for console
580 - Create Pools that are referenced in a Run statement at startup if possible.
581 - Use runbeforejob to unload, then reload a volume previously used,
582 then the next job run gets an error reading the drive.
583 - Make bootstrap filename unique.
584 - Test a second language e.g. french.
585 - Start working on Base jobs.
586 - Make "make binary-release" work from any directory.
587 - Implement UnsavedFiles DB record.
588 - Implement argc/argv for daemon command line scanning using table driven
590 - Implement table driven single argc/argv scanner to pickup all arguments.
591 Much like xxx_conf.c scan table.
592 keyword, handler(store_routine), store_address, code, flags, default.
593 - From Phil Stracchino:
594 It would probably be a per-client option, and would be called
595 something like, say, "Automatically purge obsoleted jobs". What it
596 would do is, when you successfully complete a Differential backup of a
597 client, it would automatically purge all Incremental backups for that
598 client that are rendered redundant by that Differential. Likewise,
599 when a Full backup on a client completed, it would automatically purge
600 all Differential and Incremental jobs obsoleted by that Full backup.
601 This would let people minimize the number of tapes they're keeping on
602 hand without having to master the art of retention times.
603 - Implement a M_SECURITY message class.
604 - When doing a Backup send all attributes back to the Director, who
605 would then figure out what files have been deleted.
606 - Currently in mount.c:236 the SD simply creates a Volume. It should have
607 explicit permission to do so. It should also mark the tape in error
608 if there is an error.
609 - Make sure all restore counters are working correctly in the FD.
610 - SD Bytes Read is wrong.
611 - Look at ALL higher level routines that call block.c to be sure
612 they don't expect something in errmsg.
613 - Investigate doing RAW backup of Win32 partition.
614 - Add thread specific data to hold the jcr -- send error messages from
615 low level routines by accessing it and using Jmsg().
616 - Cancel waiting for Client connect in SD if FD goes away.
618 - Examine Bare Metal restore problem (a FD crash exists somewhere ...).
619 - Implement timeout in response() when it should come quickly.
620 - Implement console @echo command.
621 - Implement a Slot priority (loaded/not loaded).
622 - Implement "vacation" Incremental only saves.
623 - Implement single pane restore (much like the Gftp panes).
624 - Implement Automatic Mount even in operator wait.
625 - Implement create "FileSet"?
626 - Fix watchdog pthread crash on Win32 (this is pthread_kill() Cygwin bug)
627 - Implement "scratch pool" where tapes are defined and can be
628 taken by any pool that needs them.
629 - Implement restore "current system", but take all files without
630 doing selection tree -- so that jobs without File records can
632 - Add prefixlinks to where or not where absolute links to FD.
633 - Issue message to mount a new tape before the rewind.
634 - Simplified client job initiation for portables.
635 - If SD cannot open a drive, make it periodically retry.
636 - Add more of the config info to the tape label.
638 - If tape is marked read-only, then try opening it read-only rather than
639 failing, and remember that it cannot be written.
640 - Refine SD waiting output:
641 Device is being positioned
642 > Device is being positioned for append
643 > Device is being positioned to file x
645 - Figure out some way to estimate output size and to avoid splitting
646 a backup across two Volumes -- this could be useful for writing CDROMs
647 where you really prefer not to have it split -- not serious.
648 - Have SD compute MD5 or SHA1 and compare to what FD computes.
649 - Make VolumeToCatalog calculate an MD5 or SHA1 from the
650 actual data on the Volume and compare it.
651 - Implement Bacula plugins -- design API
652 - Make bcopy read through bad tape records.
653 - Program files (i.e. execute a program to read/write files).
654 Pass read date of last backup, size of file last time.
655 - Add Signature type to File DB record.
656 - CD into subdirectory when open()ing files for backup to
657 speed up things. Test with testfind().
658 - Priority job to go to top of list.
659 - Why are save/restore of device different sizes (sparse?) Yup! Fix it.
660 - Implement some way for the Console to dynamically create a job.
661 - Restore to a particular time -- e.g. before date, after date.
662 - Solaris -I on tar for include list
663 - Need a verbose mode in restore, perhaps to bsr.
664 - bscan without -v is too quiet -- perhaps show jobs.
665 - Add code to reject whole blocks if not wanted on restore.
666 - Check if we can increase Bacula FD priorty in Win2000
667 - Make sure the MaxVolFiles is fully implemented in SD
668 - Check if both CatalogFiles and UseCatalog are set to SD.
669 - Figure out how to do a bare metal Windows restore
670 - Possibly add email to Watchdog if drive is unmounted too
671 long and a job is waiting on the drive.
672 - Restore program that errs in SD due to no tape, reports
673 OK incorrectly in output.
674 - After unmount, if restore job started, ask to mount.
675 - Convert all %x substitution variables, which are hard to remember
676 and read to %(variable-name). Idea from TMDA.
677 - Remove NextId for SQLite. Optimize.
678 - Move all SQL statements into a single location.
679 - Add UA rc and history files.
680 - put termcap (used by console) in ./configure and
681 allow -with-termcap-dir.
682 - Fix Autoprune for Volumes to respect need for full save.
683 - Fix Win32 config file definition name on /install
684 - Compare tape to Client files (attributes, or attributes and data)
685 - Make all database Ids 64 bit.
686 - Write an applet for Linux.
687 - Allow console commands to detach or run in background.
688 - Fix status delay on storage daemon during rewind.
689 - Add SD message variables to control operator wait time
690 - Maximum Operator Wait
691 - Minimum Message Interval
692 - Maximum Message Interval
693 - Send Operator message when cannot read tape label.
694 - Verify level=Volume (scan only), level=Data (compare of data to file).
695 Verify level=Catalog, level=InitCatalog
697 - Add keyword search to show command in Console.
698 - Events : tape has more than xxx bytes.
699 - Complete code in Bacula Resources -- this will permit
700 reading a new config file at any time.
701 - Handle ctl-c in Console
702 - Implement script driven addition of File daemon to config files.
703 - Think about how to make Bacula work better with File (non-tape) archives.
704 - Write Unix emulator for Windows.
705 - Put memory utilization in Status output of each daemon
706 if full status requested or if some level of debug on.
707 - Make database type selectable by .conf files i.e. at runtime
708 - Set flag for uname -a. Add to Volume label.
709 - Implement throttled work queue.
710 - Restore files modified after date
711 - SET LD_RUN_PATH=$HOME/mysql/lib/mysql
712 - Implement Restore FileSet=
713 - Create a protocol.h and protocol.c where all protocol messages
715 - Remove duplicate fields from jcr (e.g. jcr.level and jcr.jr.Level, ...).
716 - Timout a job or terminate if link goes down, or reopen link and query.
717 - Concept of precious tapes (cannot be reused).
718 - Make bcopy copy with a single tape drive.
719 - Permit changing ownership during restore.
722 > My suggestion: Add a feature on the systray menu-icon menu to request
723 > an immediate backup now. This would be useful for laptop users who may
724 > not be on the network when the regular scheduled backup is run.
726 > My wife's suggestion: Add a setting to the win32 client to allow it to
727 > shut down the machine after backup is complete (after, of course,
728 > displaying a "System will shut down in one minute, click here to cancel"
729 > warning dialog). This would be useful for sites that want user
730 > woorkstations to be shut down overnight to save power.
733 - Autolabel should be specified by DIR instead of SD.
736 - AutoScan (check checksum of tape)
737 - Format command = "format /dev/nst0"
741 - Seek resolution (usually corresponds to buffer size)
742 - EODErrorCode=ENOSPC or code
743 - Partial Read error code
744 - Partial write error code
745 - Nonformatted read error
746 - Nonformatted write error
747 - WriteProtected error
751 - IgnoreCloseErrors=yes
761 - FD sends unsaved file list to Director at end of job (see
763 - File daemon should build list of files skipped, and then
764 at end of save retry and report any errors.
765 - Write a Storage daemon that uses pipes and
766 standard Unix programs to write to the tape.
768 - Need something that monitors the JCR queue and
769 times out jobs by asking the deamons where they are.
770 - Enhance Jmsg code to permit buffering and saving to disk.
771 - device driver = "xxxx" for drives.
773 - Ensure that /dev/null works
774 - Need report class for messages. Perhaps
775 report resource where report=group of messages
776 - enhance scan_attrib and rename scan_jobtype, and
777 fill in code for "since" option
778 - Director needs a time after which the report status is sent
779 anyway -- or better yet, a retry time for the job.
780 - Don't reschedule a job if previous incarnation is still running.
781 - Some way to automatically backup everything is needed????
782 - Need a structure for pending actions:
784 - termination status (part of buffered msgs?)
786 Read, Write, Clean, Delete
787 - Login to Bacula; Bacula users with different permissions:
788 owner, group, user, quotas
789 - Store info on each file system type (probably in the job header on tape.
790 This could be the output of df; or perhaps some sort of /etc/mtab record.
793 - Design at hierarchial storage for Bacula. Migration and Clone.
794 - Implement FSM (File System Modules).
795 - Audit M_ error codes to ensure they are correct and consistent.
796 - Add variable break characters to lex analyzer.
797 Either a bit mask or a string of chars so that
798 the caller can change the break characters.
799 - Make a single T_BREAK to replace T_COMMA, etc.
800 - Ensure that File daemon and Storage daemon can
801 continue a save if the Director goes down (this
802 is NOT currently the case). Must detect socket error,
803 buffer messages for later.
804 - Enhance time/duration input to allow multiple qualifiers e.g. 3d2h
805 - Add ability to backup to two Storage devices (two SD sessions) at
806 the same time -- e.g. onsite, offsite.
807 - Add the ability to consolidate old backup sets (basically do a restore
808 to tape and appropriately update the catalog). Compress Volume sets.
809 Might need to spool via file is only one drive is available.
810 - Compress or consolidate Volumes of old possibly deleted files. Perhaps
811 someway to do so with every volume that has less than x% valid
815 Migration: Move a backup from one Volume to another
816 Clone: Copy a backup -- two Volumes
818 Bacula Migration is based on Jobs (apparently Networker is file by file).
820 Migration triggered by:
824 Highwater mark (keep total size)
829 ======================================================
831 It is somewhat like a Full save becomes an incremental since
832 the Base job (or jobs) plus other non-base files.
834 - A Base backup is same as Full backup, just different type.
835 - New BaseFiles table that contains:
837 BaseJobId - Base JobId referenced for this FileId (needed ???)
838 JobId - JobId currently running
839 FileId - File not backed up, exists in Base Job
840 FileIndex - FileIndex from Base Job.
841 i.e. for each base file that exists but is not saved because
842 it has not changed, the File daemon sends the JobId, BaseId,
843 FileId, FileIndex back to the Director who creates the DB entry.
844 - To initiate a Base save, the Director sends the FD
845 the FileId, and full filename for each file in the Base.
846 - When the FD finds a Base file, he requests the Director to
847 send him the full File entry (stat packet plus MD5), or
848 conversely, the FD sends it to the Director and the Director
849 says yes or no. This can be quite rapid if the FileId is kept
850 by the FD for each Base Filename.
851 - It is probably better to have the comparison done by the FD
852 despite the fact that the File entry must be sent across the
854 - An alternative would be to send the FD the whole File entry
855 from the start. The disadvantage is that it requires a lot of
856 space. The advantage is that it requires less communications
858 - The Job record must be updated to indicate that one or more
860 - At end of Job, FD returns:
861 1. Count of base files/bytes not written to tape (i.e. matches)
862 2. Count of base file that were saved i.e. had changed.
863 - No tape record would be written for a Base file that matches, in the
864 same way that no tape record is written for Incremental jobs where
865 the file is not saved because it is unchanged.
866 - On a restore, all the Base file records must explicitly be
867 found from the BaseFile tape. I.e. for each Full save that is marked
868 to have one or more Base Jobs, search the BaseFile for all occurrences
870 - An optimization might be to make the BaseFile have:
876 This would avoid the need to explicitly fetch each File record for
877 the Base job. The Base Job record will be fetched to get the
878 VolSessionId and VolSessionTime.
879 =========================================================
882 ==========================================================
884 For each Incremental job that is run, there may be files that
885 were found but not saved because they were locked (this applies
886 only to Windows). Such a system could send back to the Director
887 a list of Unsaved files.
889 - New UnSavedFiles table that contains:
893 - Then in the next Incremental job, the list of Unsaved Files will be
894 feed to the FD, who will ensure that they are explicitly chosen even
895 if standard date/time check would not have selected them.
896 =============================================================
899 Done: (see kernsdone for more)
902 - John's Full save failed with 1.32c FD and 1.31 Dir no FD status,
903 and no error message.
904 - Add fd and st as Console keywords.
905 - Recycling volume with a Slot requires an operator intervention:
906 rufus-dir: Start Backup JobId 18, Job=kernsave.2003-11-01_21.23.52
907 rufus-dir: Pruned 1 Job on Volume Vol01 from catalog.
908 rufus-dir: There are no Jobs associated with Volume Vol01. Marking it purged.
909 rufus-dir: Recycled volume "Vol01"
910 rufus-sd: Please mount Volume "Vol01" on Storage Device "DDS-4" for Job kernsave.2003-11-01_21.23.52
911 Use "mount" command to release Job.
912 - Implement Dan's bacula script (email of 26 Oct).
913 - Add JobName= to VerifyToCatalog so that all verifies can be done at the end.
914 - Edit the Client/Storage name into authentication failure messages.
915 - Fix packet too big problem. This is most likely a Windows TCP stack
917 - Implement ClientRunBeforeJob and ClientRunAfterJob.
918 - Implement forward spacing block/file: position_device(bsr) --
919 just before read_block_from_device();
922 Multiple drive autochanger data: see Alan Brown
923 mtx -f xxx unloadStorage Element 1 is Already Full(drive 0 was empty)
924 Unloading Data Transfer Element into Storage Element 1...source Element
927 (drive 0 was empty and so was slot 1)
928 > mtx -f xxx load 15 0
929 no response, just returns to the command prompt when complete.
930 > mtx -f xxx status Storage Changer /dev/changer:2 Drives, 60 Slots ( 2 Import/Export )
931 Data Transfer Element 0:Full (Storage Element 15 Loaded):VolumeTag = HX001
932 Data Transfer Element 1:Empty
933 Storage Element 1:Empty
934 Storage Element 2:Full :VolumeTag=HX002
935 Storage Element 3:Full :VolumeTag=HX003
936 Storage Element 4:Full :VolumeTag=HX004
937 Storage Element 5:Full :VolumeTag=HX005
938 Storage Element 6:Full :VolumeTag=HX006
939 Storage Element 7:Full :VolumeTag=HX007
940 Storage Element 8:Full :VolumeTag=HX008
941 Storage Element 9:Full :VolumeTag=HX009
942 Storage Element 10:Full :VolumeTag=HX010
943 Storage Element 11:Empty
944 Storage Element 12:Empty
945 Storage Element 13:Empty
946 Storage Element 14:Empty
947 Storage Element 15:Empty
948 Storage Element 16:Empty....
949 Storage Element 28:Empty
950 Storage Element 29:Full :VolumeTag=CLNU01L1
951 Storage Element 30:Empty....
952 Storage Element 57:Empty
953 Storage Element 58:Full :VolumeTag=NEX261L2
954 Storage Element 59 IMPORT/EXPORT:Empty
955 Storage Element 60 IMPORT/EXPORT:Empty
957 Unloading Data Transfer Element into Storage Element 15...done
959 (just to verify it remembers where it came from, however it can be
960 overrriden with mtx unload {slotnumber} to go to any storage slot.)
962 There needs to be a table of drive # to devices somewhere - If there are
963 multiple changers or drives there may not be a 1:1 correspondance between
964 changer drive number and system device name - and depending on the way the
965 drives are hooked up to scsi busses, they may not be linearly numbered
966 from an offset point either.something like
968 Autochanger drives = 2
969 Autochanger drive 0 = /dev/nst1
970 Autochanger drive 1 = /dev/nst2
971 IMHO, it would be _safest_ to use explicit mtx unload commands at all
972 times, not just for multidrive changers. For a 1 drive changer, that's
978 MTX's manpage (1.2.15):
979 unload [<slotnum>] [ <drivenum> ]
980 Unloads media from drive <drivenum> into slot
981 <slotnum>. If <drivenum> is omitted, defaults to
982 drive 0 (as do all commands). If <slotnum> is
983 omitted, defaults to the slot that the drive was
984 loaded from. Note that there's currently no way
985 to say 'unload drive 1's media to the slot it
986 came from', other than to explicitly use that
987 slot number as the destination.AB
993 undef# camcontrol devlist
994 <WANGTEK 51000 SCSI M74H 12B3> at scbus0 target 2 lun 0 (pass0,sa0)
995 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 0 (pass1,sa1)
996 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 1 (pass2)
998 tapeinfo -f /dev/sg0 with a bad tape in drive 1:
999 [kern@rufus mtx-1.2.17kes]$ ./tapeinfo -f /dev/sg0
1000 Product Type: Tape Drive
1002 Product ID: 'C5713A '
1004 Attached Changer: No
1005 TapeAlert[3]: Hard Error: Uncorrectable read/write error.
1006 TapeAlert[20]: Clean Now: The tape drive neads cleaning NOW.
1013 Medium Type: Not Loaded
1016 DataCompEnabled: yes
1017 DataCompCapable: yes
1018 DataDeCompEnabled: yes
1025 Handling removable disks
1027 From: Karl Cunningham <karlc@keckec.com>
1029 My backups are only to hard disk these days, in removable bays. This is my
1030 idea of how a backup to hard disk would work more smoothly. Some of these
1031 things Bacula does already, but I mention them for completeness. If others
1032 have better ways to do this, I'd like to hear about it.
1034 1. Accommodate several disks, rotated similar to how tapes are. Identified
1035 by partition volume ID or perhaps by the name of a subdirectory.
1036 2. Abort & notify the admin if the wrong disk is in the bay.
1037 3. Write backups to different subdirectories for each machine to be backed
1039 4. Volumes (files) get created as needed in the proper subdirectory, one
1041 5. When a disk is recycled, remove or zero all old backup files. This is
1042 important as the disk being recycled may be close to full. This may be
1043 better done manually since the backup files for many machines may be
1044 scattered in many subdirectories.
1049 - To pass Include 1 or two letter commands
1050 I Name Include name - first record
1051 B Name Base name - repeat
1054 E zzz Exclude expression (wild card)
1056 D "reader" Reader program
1057 T "writer" Writer program
1058 O Options In current commpressed format
1059 (compression, signature, onefs, recurse, sparse,
1060 replace, verify options, ...)
1062 B BaseName Start second option set any letter
1065 F Number Number of filenames to follow
1070 F Number Number of filenames to follow
1073 - Add regression of btape "fill"
1075 - Implement "label slots=1,2... pool=yyy barcodes"
1076 - Mark a non-removable disk Volume in error rather than asking sys op to
1078 - Implement Fixed storage LabelFormat test.
1079 - Add reporting in attr despooling.
1082 - Implement fast tree insert (doubly linked list?)
1083 - Disallow using Internal database
1084 - Feedback while the tree is being built.