8 Data encryption Meno Abels (see projects)
9 Communication encryption Meno Abels
10 Version 1.35 Kern (see below)
11 ========================================================
13 1.35 Items to do for release:
15 - Fix error handling in spooling both data and attribute.
16 - Perhaps add read/write programs and/or plugins to FileSets.
17 - Add new DCR calling sequences everywhere in SD. This will permit
18 simultaneous use of multiple devices by a single job.
19 - Look at Chris' patch for bscan bug.
21 - Add bscan to four-concurrent-jobs regression.
23 - Add IPv6 to regression
24 - Alternative to static linking "ldd prog" save all binaries listed,
25 restore them and point LD_LIBRARY_PATH to them.
26 - Document a get out of jail procedure if everything breaks if
27 you lost/broke the Catalog -- do the same for "I know my
28 file is there how do I get it back?".
29 - Test/doc Tape Alerts
30 - Doc update AllFromVol
31 - Doc dbcheck eliminate orphaned clients.
32 - Doc new duration time input editing.
33 - Doc -p option in stored
36 Documentation to do: (any release a little bit at a time)
37 - Document query file format.
38 - Add more documentation for bsr files.
39 - Document problems with Verify and pruning.
40 - Document how to use multiple databases.
41 - VXA drives have a "cleaning required"
42 indicator, but Exabyte recommends preventive cleaning after every 75
45 In this context, it should be noted that Exabyte has a command-line
46 vxatool utility available for free download. (The current version is
47 vxatool-3.72.) It can get diagnostic info, read, write and erase tapes,
48 test the drive, unload tapes, change drive settings, flash new firmware,
50 Of particular interest in this context is that vxatool <device> -i will
51 report, among other details, the time since last cleaning in tape motion
52 minutes. This information can be retrieved (and settings changed, for
53 that matter) through the generic-SCSI device even when Bacula has the
54 regular tape device locked. (Needless to say, I don't recommend
55 changing tape settings while a job is running.)
56 - Lookup HP cleaning recommendations.
57 - Lookup HP tape replacement recommendations (see trouble shooting autochanger)
58 - Add more obvious documentation of @ for conf files.
59 - Document doing table repair
61 Testing to do: (painful)
64 For 1.37 Testing/Documentation:
65 - Fix find_device in stored/dircmd.c:462 (see code)
66 - Add db check test to regression. Test each function like delete,
68 - If you use restore replace=never, the directory attributes for
69 non-existent directories will not be restored properly.
72 - see lzma401.zip in others directory for new compression
74 - Minimal autochanger handling in Bacula and in btape.
75 - Look into how tar does not save sockets and the possiblity of
76 not saving them in Bacula (Martin Simmons reported this).
77 The next two lines will show them.
78 localmounts=`awk '/ext/ { print $2 }' /proc/mounts` # or whatever
79 find $localmounts -xdev -type s -ls
80 - Fix restore jobs so that multiple jobs can run if they
81 are not using the same tape(s).
82 - Allow the user to select JobType for manual pruning/purging.
83 - Look at adding Client run command that will use the
84 port opened by the client.
85 - bscan does not put first of two volumes back with all info in
87 - Implement the FreeBSD nodump flag in chflags.
88 - Figure out how to make named console messages go only to that
89 console and to the non-restricted console (new console class?).
90 - Make restricted console prompt for password if *ask* is set or
91 perhaps if password is undefined.
92 - Implement "from ISO-date/time every x hours/days/weeks/months" in
94 ==== from Marc Schoechlin
95 - the help-command should be more verbose
96 (it should explain the paramters of the different
98 -> it´s time-comsuming to consult the manual anytime
99 you need a special parameter
100 -> maybe it´s more easy to maintain this, if the
101 descriptions of that commands are outsourced to
103 - the cd-command should allow complete paths
104 i.e. cd /foo/bar/foo/bar
105 -> if a customer mails me the path to a certain file,
106 it´s faster to enter the specified directory
107 - if the password is not configured in bconsole.conf
108 you should be asked for it.
109 -> sometimes you like to do restore on a customer-machine
110 which shouldn´t know the password for bacula.
111 -> adding the password to the file favours admins
112 to forget to remove the password after usage
114 the protection of that file is less important
115 - long-listed-output of commands should be scrollable
116 like the unix more/less-command does
117 -> if someone runs 200 and more machines, the lists could
118 be a little long and complex
119 - command-output should be shown column by column
120 to reduce scrolling and to increase clarity
122 - lsmark should list the selected files with full
124 - wildcards for selecting and file and directories would be nice
125 - any actions should be interuptable with STRG+C
126 - command-expansion would be pretty cool
128 - Compaction of Disk space by "migrating" Volumes that have pruned
129 Jobs (what criteria? size, #jobs, time).
130 - Add prune all command
131 - Document fact that purge can destroy a part of a restore by purging
132 one volume while others remain valid -- perhaps mark Jobs.
133 - Add multiple-media-types.txt
134 - look at mxt-changer.html
135 - Make ? do a help command (no return needed).
136 - Implement restore directory.
137 - Add All Local Partitions = yes to new style saves.
138 - Document streams and how to implement them.
139 - Possibly implement "Ensure Full Backup = yes" looks for a failed full backup
140 and upgrades the current backup if one exists.
141 - Check that barcode reading and update slots scan works.
142 - Try not to re-backup a file if a new hard link is added.
143 - Add feature to backup hard links only, but not the data.
144 - Add "All Local = yes" option to save to include all local partitions.
145 - Fix stream handling to be simpler.
146 - Add Priority and Bootstrap to Run a Job.
147 - Eliminate Restore "Run Restore Job" prompt by allowing new "run command
149 - Remove View FileSet button from Run a Job dialog.
150 - Don't count higher level non-saved directories in Files expected
152 - Handle prompt for restore job at end of Restore command.
153 - Add display of total selected files to Restore window.
154 - Add tree pane to left of window.
155 - Add progress meter.
156 - Polling does not work for restore. It tries a number of times,
157 gives up, and crashes the SD.
158 - Lock jcr_chain when doing attach/detach in acquire.c
159 - Add assert in free_jcr if attach/detach chain active.
160 - Max wait time or max run time causes seg fault -- see runtime-bug.txt
161 - Document writing to a CD/DVD with Bacula.
162 - Add check for tape alerts.
163 - Add a "base" package to the window installer for pthreadsVCE.dll
164 which is needed by all packages.
165 - Add message to user to check for fixed block size when the forward
166 space test fails in btape.
167 - When unmarking a directory check if all files below are unmarked and
168 then remove the + flag -- in the restore tree.
169 - Possibly implement: Action = Unmount Device="TapeDrive1" in Admin jobs.
170 - Setup lrrd graphs: (http://www.linpro.no/projects/lrrd/) Mike Acar.
171 - Revisit the question of multiple Volumes (disk) on a single device.
172 - Finish SIGHUP work.
173 - Check that all change in wait status in the SD are
174 signaled to the Director.
175 - Add a block copy option to bcopy.
176 - Investigate adding Mac Resource Forks.
177 - Finish work on Gnome restore GUI.
178 - Fix "llist jobid=xx" where no fileset or client exists.
179 - Check pruning of restore jobs.
181 it seems to be complaining about 12:00pm which should be a valid 12
182 hour time. I changed the time to 11:59am and everything works fine.
183 Also 12:00am works fine. 0:00pm also works (which I don't think
184 should). None of the values 12:00pm - 12:59pm work for that matter.
185 - Add level to estimate command.
186 - For each job type (Admin, Restore, ...) require only the really necessary
188 - Fix option 2 of restore -- list where file is backed up -- require Client,
189 then list last 20 backups.
190 - Add all pools in Dir conf to DB also update them to catch changed
191 LabelFormats and such.
192 - Update volumes FromPool=xxx does all volumes.
193 - Pass Director resource name as an option to the Console.
194 - Add a "batch" mode to the Console (no unsolicited queries, ...).
195 - Add code to check for tape alerts -- tapeinfo.
196 - Make sure list of Volumes needed is in correct order for restore.
198 - Remove paths (and files that reference them) that have no trailing slash
199 in dbcheck -- or add a trailing slash.
200 - Remove Filenames (and files that reference them) that have a trailing
201 slash in dbcheck -- or remove the trailing slash.
202 - Remove orphaned paths/filenames by copying them to a new table with a
203 reference count, then mark all referenced files/paths and remove unreferenced
205 - Add a .list all files in the restore tree (probably also a list all files)
206 Do both a long and short form.
207 - Allow browsing the catalog to see all versions of a file (with
208 stat data on each file).
209 - Restore attributes of directory if replace=never set but directory
211 - Allow "delete job jobid=xxx,yyy,aaa-bbb" i.e. list + ranges.
212 - Use SHA1 on authentication if possible.
213 - See comtest-xxx.zip for Windows code to talk to USB.
214 - Make btape accept Device Names in addition to Archive names.
215 - Add Events and Perl scripting.
216 - Add John's appended files:
217 Appended = { /files/server/logs/http/*log }
218 and such files would be treated as follows.On a FULL backup, they would
219 be backed up like any other file.On an INCREMENTAL backup, where a
220 previous INCREMENTAL or FULL was already in thecatalogue and the length
221 of the file wasgreater than the length of the last backup, only thedata
222 added since the last backup will be dumped.On an INCREMENTAL backup, if
223 the length of the file is less than thelength of the file with the same
224 name last backed up, the completefile is dumped.On Windows systems, with
225 creation date of files, we can be evensmarter about this and not count
226 entirely upon the length.On a restore, the full and all incrementals
227 since it will beapplied in sequence to restore the file.
228 - Add a regression test for dbcheck.
229 - Add disk seeking on restore. - Allow
230 for optional cancelling of SD and FD in case DIR
231 gets a fatal error. Requested by Jesse Guardiani <jesse@wingnet.net>
232 - Bizarre message: Error: Could not open WriteBootstrap file:
233 - Build console in client only build.
234 - Add "limit=n" for "list jobs"
235 - Check new HAVE_WIN32 open bits.
236 - Check if the tape has moved before writing.
237 - Handling removable disks -- see below:
238 - Multiple drive autochanger support -- see below.
239 - Keep track of tape use time, and report when cleaning is necessary.
240 - Fix FreeBSD mt_count problem.
241 - Add FromClient and ToClient keywords on restore command (or
242 BackupClient RestoreClient).
243 - Automatic "update slots" on user configuration directive when a
245 - Implement a JobSet, which groups any number of jobs. If the
246 JobSet is started, all the jobs are started together.
247 Allow Pool, Level, and Schedule overrides.
248 - Enhance cancel to timeout BSOCK packets after a specific delay.
249 - When I restore to Windows the Created, Accessed and Modifiedtimes are
250 those of the time of the restore, not those of the originalfile.
251 The dates you will find in your restore log seem to be the original
253 - Volume "add"ed to Pool gets recycled in first use. VolBytes=0
254 - If a tape is recycled while it is mounted, Stanislav Tvrudy must do an
255 additional mount to deblock the job.
257 bscan: sql_update.c:65 UPDATE File SET MD5='Ij+5kwN6TFIxK+8l8+/I+A' WHERE FileId=0
258 bscan: bscan.c:1074 Could not add MD5/SHA1 to File record. ERR=sql_update.c:65 Update problem: affected_rows=0
259 - Do scheduling by UTC using gmtime_r() in run_conf, scheduler, and
260 ua_status.!!! Thanks to Alan Brown for this tip.
261 - Look at updating Volume Jobs so that Max Volume Jobs = 1 will work
262 correctly for multiple simultaneous jobs.
263 - Correct code so that FileSet MD5 is calculated for < and | filename
265 - Mark Volume in error on error from WEOF.
266 - Implement the Media record flag that indicates that the Volume does disk
268 - Implement VolAddr, which is used when Volume is addressed like a disk,
269 and form it from VolFile and VolBlock.
270 - Make multiple restore jobs for multiple media types specifying
271 the proper storage type.
272 - Implement MediaType keyword in bsr?
273 - Fix fast block rejection (stored/read_record.c:118). It passes a null
274 pointer (rec) to try_repositioning().
275 - Look at extracting Win data from BackupRead.
276 - Having dashes in filenames apparently creates problems for restore
277 by filename??? hard to believe.
278 - Implement RestoreJobRetention? Maybe better "JobRetention" in a Job,
279 which would take precidence over the Catalog "JobRetention".
280 - Implement Label Format in Add and Label console commands.
281 - Possibly up network buffers to 65K. Put on variable.
282 - Put email tape request delays on one or more variables. User wants
283 to cancel the job after a certain time interval. Maximum Mount Wait?
284 - Job, Client, Device, Pool, or Volume?
285 Is it possible to make this a directive which is *optional* in multiple
286 resources, like Level? If so, I think I'd make it an optional directive
287 in Job, Client, and Pool, with precedence such that Job overrides Client
288 which in turn overrides Pool.
289 - Print a message when a job starts if the conf file is not current.
291 - Spooling ideas taken from Volker Sauer's and other's emails:
292 > IMHO job spooling should be turned on
298 > where and 2) overrides 1) and 3) is independent.
300 Yes, this is the minimum that I think is necessary.
304 > It should be switched by job, because the job that backs up the machine
305 > with the bacula-sd on doesn't need spooling.
306 > It should be switched by schedule, because for full-backups I don't need
307 > spooling, so I can switch it off (because the network faster then the
310 True, with the exception that if you have enough disk spool space,
311 and you want to run concurrent jobs, spooling can eliminate the block
312 interleaving restore inefficiencies.
314 > And you should be able to turn it of by sd for sd-machines with low disk
315 > capacity or if you just don't need or want this feature.
318 > - definitly the possibility for multipe spool direcories
320 Having multiple directories is no problem -- having different maximum
321 sizes creates specification problems. At some point, I will probably
322 have a common SD pool of spool directories as well as a set of
323 private spool directories for each device. The first implementation
324 will be a set of private spool directories for each device since
325 managing a global pool with a bunch of threads writing into the same
326 directory is *much* more complicated and prone to error.
328 > - the ability to spool parts of a backup (not the whole client)
330 This may change in the future, but for the moment, it will spool
331 either to a job high water mark, or until the directory is full
332 (reaches max spool size or I/O error). It will then write to tape,
333 truncate the spool file, and begin spooling again.
335 > - spooling while writing to tape
337 Not within a job, but yes, if you run concurrent jobs -- each is a
338 different thread. Within a job could be a feature, but *much* later.
340 > - parallel spooling (like parallel jobs/ concurrent jobs) of clients
342 Yes, this is one of my main motivations for doing it (aside from
343 eliminating tape "shoe shine" during incremental backups.
345 > - flushing a backup that only went to disk (like amflush in amanda)
347 This will be a future feature, since spooling is different from backing
348 up to disk. The future feature will be "migration" which will move a job
349 from one backup Volume to another.
351 - New Storage specifications:
352 Passed to SD as a sort of BSR record called Storage Specification
356 MediaType -> Next MediaType
358 Device -> Next Device
359 Write Copy Resource that makes a copy of a resource.
361 Allow multiple Storage specifications
369 Allow Multiple Pool specifications (note, Pool currently
371 Allow Multiple MediaType specifications
372 Allow Multiple Device specifications
373 Perhaps keep this in a single SSR
374 Tie a Volume to a specific device by using a MediaType that
375 is contained in only one device.
376 In SD allow Device to have Multiple MediaTypes
379 - Look at www.nu2.nu/pebuilder as a helper for full windows
381 Ideas from Jerry Scharf:
382 First let's point out some big pluses that bacula has for this
384 more importantly it's active. Thank you so much for that
385 even more important, it's not flaky
386 it has an open access catalog, opening many possibilities
387 it's pushing toward heterogeneous systems capability
389 I don't remember an include file directive for config files
390 (not filesets, actual config directives)
391 can you check the configs without starting the daemon?
392 some warnings about possible common mistakes
394 doing the testing and blessing of concurrent backup writes
395 this is absolutely necessary in the enterprise
396 easy user recovery GUI with full access checking
397 Macintosh file client
398 macs are an interesting niche, but I fear a server is a rathole
399 working bare iron recovery for windows
400 much better handling on running config changes
401 thinking through the logic of what happens to jobs in progress
402 the option for inc/diff backups not reset on fileset revision
403 a) use both change and inode update time against base time
404 b) do the full catalog check (expensive but accurate)
405 sizing guide (how much system is needed to back up N systems/files)
406 consultants on using bacula in building a disaster recovery system
408 or how to get at fancy things that one could do with bacula
409 logwatch code for bacula logs (or similar)
410 linux distro inclusion of bacula (brings good and bad, but necessary)
411 win2k/XP server capability (icky but you asked)
412 support for Oracle database ??
414 - Look at adding SQL server and Exchange support for Windows.
415 - Restore: Enter Filename: 'C:/Documents and Settings/Comercial/My
416 Documents/MOP/formulário de registro BELAS ARTES.doc' causes Bacula to
418 - Each DVD-RAM disk would be a volume, just like each tape is
419 a volume. It's a 4.7GB media with random access, but there's nothing about
420 it that I can see that makes it so different than a tape from bacula's
421 perspective. Why couldn't I back up to a bare floppy as a volume (ignoring
423 - Make dev->file and dev->block_num signed integers so that -1 can
424 be an invalid value which happens with BSR.
425 - Create VolAddr for disk files in place of VolFile and VolBlock. This
426 is needed to properly specify ranges.
427 - Print bsmtp output to job report so that problems will be seen.
428 - Pass the number of files to be restored to the FD for reporting
429 - Add progress of files/bytes to SD and FD.
430 - Don't continue Restore if no files selected.
431 - Print warning message if FileId > 4 billion
432 - do a "messages" before the first prompt in Console
433 - Add a date and time stamp at the beginning of every line in the
434 Job report (Volker Sauer).
435 - Client does not show busy during Estimate command.
436 - Implement Console mtx commands.
437 - Add a default DB password to MySQL.
438 GRANT all privileges ON bacula.* TO bacula@localhost IDENTIFIED BY
441 - Implement a Mount Command and an Unmount Command where
442 the users could specify a system command to be performed
443 to do the mount, after which Bacula could attempt to
444 read the device. This is for Removeable media such as a CDROM.
445 - Most likely, this mount command would be invoked explicitly
446 by the user using the current Console "mount" and "unmount"
447 commands -- the Storage Daemon would do the right thing
448 depending on the exact nature of the device.
449 - As with tape drives, when Bacula wanted a new removable
450 disk mounted, it would unmount the old one, and send a message
451 to the user, who would then use "mount" as described above
452 once he had actually inserted the disk.
453 - Implement dump/print label to UA
454 - Implement disk spooling. Two parts: 1. Spool to disk then
455 immediately to tape to speed up tape operations. 2. Spool to
456 disk only when the tape is full, then when a tape is hung move
458 - Scratch Pool where the volumes can be re-assigned to any Pool.
459 - bextract is sending everything to the log file ****FIXME****
460 - Add Progress command that periodically reports the progress of
462 - Restrict characters permitted in a Resource name, and don't permit
464 - Allow multiple Storage specifications (or multiple names on
465 a single Storage specification) in the Job record. Thus a job
466 can be backed up to a number of storage devices.
467 - Implement some way for the File daemon to contact the Director
468 to start a job or pass its DHCP obtained IP number.
469 - Implement multiple Consoles.
470 - Implement a query tape prompt/replace feature for a console
472 Two jobs ready to go, first one blocked waiting for media
473 Cancel 2nd job ("waiting execution" one)
476 - Copy console @ code to gnome2-console
477 - Make AES the only encryption algorithm see
478 http://csrc.nist.gov/CryptoToolkit/aes/). It's
479 an officially adopted standard, has survived peer
480 review, and provides keys up to 256 bits.
481 - Think about how space could be freed up on a tape -- perhaps this
482 is a Merge or Compact feature that is needed.
483 - Modify FileSet, did not upgrade the current Increment job, but
484 waited for the next job to be upgraded.
485 - Take a careful look at SetACL http://setacl.sourceforge.net
486 - Implement a where command for the tree telling where a file
488 - Take a careful look at Level for the estimate command, maybe make
489 it a command line option.
490 - Add Volume name to "I cannot write on this volume because"
491 - Make tree walk routines like cd, ls, ... more user friendly
492 by handling spaces better.
493 - Write your PID file and chown root:wheel before drop.
494 - Make sure there is no symlink in a file before creating a
496 - Look at mktemp or mkstemp(3).
497 mktemp and mkstemp create files with predictable names too. That's
498 not the vulnerability. The vulnerability is in creating files without
499 using the O_EXCL flag, which means "only create this file if it doesn't
500 exist, including if the file is a dangling symlink."
502 It is *NOT* enough to do the equivalent of
504 if doesn't exist $filename
505 then create $filename
507 because between the test and the create another process could have
508 gotten the CPU and created the file. You must use atomic functions
509 (those that don't get interrupted by other processes) and O_EXCL is
510 the only way for this particular example.
511 - Automatically create pools, but instead of looking for what
512 in in Job records, walk through the pool resources.
513 - Check and double check tree code, why does it take so long?
514 - Add device name to "Current Volume not acceptable because ..."
515 - Make sure that Bacula rechecks the tape after the 20 min wait.
516 - Set IO_NOWAIT on Bacula TCP/IP packets.
517 - Try doing a raw partition backup and restore by mounting a
520 Yes, it would allow to highly automatic the request for new tapes. If a
521 tape is empty, bacula reads the barcodes (native or simulated), and if
522 an unused tape is found, it runs the label command with all the
523 necessary parameters.
525 By the way can bacula automatically "move" an empty/purged volume say
526 in the "short" pool to the "long" pool if this pool runs out of volume
528 - Eliminate orphaned jobs: dbcheck, normal pruning, delete job command.
529 Hm. Well, there are the remaining orphaned job records:
531 | 105 | Llioness Save | 0000-00-00 00:00:00 | B | D | 0 | 0 | f |
532 | 110 | Llioness Save | 0000-00-00 00:00:00 | B | I | 0 | 0 | f |
533 | 115 | Llioness Save | 2003-09-10 02:22:03 | B | I | 0 | 0 | A |
534 | 128 | Catalog Save | 2003-09-11 03:53:32 | B | I | 0 | 0 | C |
535 | 131 | Catalog Save | 0000-00-00 00:00:00 | B | I | 0 | 0 | f |
537 As you can see, three of the five are failures. I already deleted the
538 one restore and one other failure using the by-client option. Deciding
539 what is an orphaned job is a tricky problem though, I agree. All these
540 records have or had 0 files/ 0 bytes, except for the restore. With no
541 files, of course, I don't know of the job ever actually becomes
542 associated with a Volume.
544 (I'm not sure if this is documented anywhere -- what are the meanings of
545 all the possible JobStatus codes?)
547 Looking at my database, it appears to me as though all the "orphaned"
548 jobs fit into one of two categories:
550 1) The Job record has a StartTime but no EndTime, and the job is not
553 2) The Job record has an EndTime, indicating that it completed, but
554 it has no associated JobMedia record.
557 This does suggest an approach. If failed jobs (or jobs that, for some
558 other reason, write no files) are associated with a volume via a
559 JobMedia record, then they should be purged when the associated volume
560 is purged. I see two ways to handle jobs that are NOT associated with a
563 1) purge them automatically whenever any volume is manually purged;
565 2) add an option to the purge command to manually purge all jobs with
566 no associated volume.
568 I think Restore jobs also fall into category 2 above .... so one might
569 want to make that "The Job record has an EndTime,, but no associated
570 JobMedia record, and is not a Restore job."
571 - make "btape /tmp" work.
572 - Make sure a rescheduled job is properly reported by status.
573 - Walk through the Pool records rather than the Job records
574 in dird.c to create/update pools.
575 - What to do about "list files job=xxx".
576 - When job rescheduled, status gives is waiting for Client Rufus
577 to connect to Storage File. Dir needs to inform SD that job
579 - Make Dmsg look at global before calling subroutine.
580 - Enable trace output at runtime for Win32
581 - Available volumes for autochangers (see patrick@baanboard.com 3 Sep 03
582 and 4 Sep) scan slots.
583 - Get and test MySQL 4.0
584 - Do a complete audit of all pthreads_mutex, cond, ... to ensure that
585 any that are dynamically initialized are destroyed when no longer used.
586 - Look at how fuser works and /proc/PID/fd that is how Nic found the
587 file descriptor leak in Bacula.
588 - Implement WrapCounters in Counters.
589 - Turn on SIGHUP in dird.c and test.
590 - Use system dependent calls to get more precise info on tape errors.
591 - Add heartbeat from FD to SD if hb interval expires.
592 - Suppress read error on blank tape when doing a label.
593 - Can we dynamically change FileSets?
594 - If pool specified to label command and Label Format is specified,
595 automatically generate the Volume name.
596 - Why can't SQL do the filename sort for restore?
597 - Look at libkse (man kse) for FreeBSD threading.
598 - Look into Microsoft Volume Shadowcopy Service VSS for backing
599 up system state components (Active Directory, System Volume, ...)
600 - Add ExhautiveRestoreSearch
601 - Look at the possibility of loading only the necessary
602 data into the restore tree (i.e. do it one directory at a
603 time as the user walks through the tree).
604 - Possibly use the hash code if the user selects all for a restore command.
605 - Orphaned Dir buffer at parse_conf.c:373 => store_dir
606 - Fix "restore all" to bypass building the tree.
607 - Prohibit backing up archive device (findlib/find_one.c:128)
608 - Implement Release Device in the Job resource to unmount a drive.
609 - Implement Acquire Device in the Job resource to mount a drive,
610 be sure this works with admin jobs so that the user can get
611 prompted to insert the correct tape. Possibly some way to say to
612 run the job but don't save the files.
613 - Implement FileOptions (see end of this document)
614 - Make things like list where a file is saved case independent for
617 - Bacula needs to propagate SD errors.
618 > > cluster-dir: Start Backup JobId 252, Job=REUTERS.2003-08-11_15.04.12
619 > > prod4-sd: REUTERS.2003-08-11_15.04.12 Error: Write error on device
620 /dev/nst0. ERR=Input/output error.
621 > > prod4-sd: REUTERS.2003-08-11_15.04.12 Error: Re-read of last block failed.
622 Last block=5162 Current block=5164.
623 > > prod4-sd: End of medium on Volume "REU007" Bytes=16,303,521,933
625 - Use autochanger to handle multiple devices.
626 - Add SuSE install doc to list.
627 - Check and rechedk "Invalid block number"
628 - Make bextract release the drive properly between tapes
629 so that an autochanger can be made to work.
630 - User wants to NOT backup up certain big files (email files).
631 - Maybe remove multiple simultaneous devices code in SD.
632 - On Windows with very long path names, it may be impossible to create
633 a file (and thus restore it) because the total length is too long.
634 We must cd into the directory then create the file without the
636 - lstat() is not going to work on Win32 for testing date.
637 - Implement a Recycle command
638 - Add client name to cram-md5 challenge so Director can immediately
639 verify if it is the correct client.
640 - Add JobLevel in FD status (but make sure it is defined).
641 - Audit all UA commands to ensure that we always prompt where possible.
642 - Check Jmsg in bnet, may not work, must dup bsock.
643 - Suppress Job Name in Jmsg for console
644 - Create Pools that are referenced in a Run statement at startup if possible.
645 - Use runbeforejob to unload, then reload a volume previously used,
646 then the next job run gets an error reading the drive.
647 - Make bootstrap filename unique.
648 - Test a second language e.g. french.
649 - Start working on Base jobs.
650 - Make "make binary-release" work from any directory.
651 - Implement UnsavedFiles DB record.
652 - Implement argc/argv for daemon command line scanning using table driven
654 - Implement table driven single argc/argv scanner to pickup all arguments.
655 Much like xxx_conf.c scan table.
656 keyword, handler(store_routine), store_address, code, flags, default.
657 - From Phil Stracchino:
658 It would probably be a per-client option, and would be called
659 something like, say, "Automatically purge obsoleted jobs". What it
660 would do is, when you successfully complete a Differential backup of a
661 client, it would automatically purge all Incremental backups for that
662 client that are rendered redundant by that Differential. Likewise,
663 when a Full backup on a client completed, it would automatically purge
664 all Differential and Incremental jobs obsoleted by that Full backup.
665 This would let people minimize the number of tapes they're keeping on
666 hand without having to master the art of retention times.
667 - Implement a M_SECURITY message class.
668 - When doing a Backup send all attributes back to the Director, who
669 would then figure out what files have been deleted.
670 - Currently in mount.c:236 the SD simply creates a Volume. It should have
671 explicit permission to do so. It should also mark the tape in error
672 if there is an error.
673 - Make sure all restore counters are working correctly in the FD.
674 - SD Bytes Read is wrong.
675 - Look at ALL higher level routines that call block.c to be sure
676 they don't expect something in errmsg.
677 - Investigate doing RAW backup of Win32 partition.
678 - Add thread specific data to hold the jcr -- send error messages from
679 low level routines by accessing it and using Jmsg().
680 - Cancel waiting for Client connect in SD if FD goes away.
682 - Examine Bare Metal restore problem (a FD crash exists somewhere ...).
683 - Implement timeout in response() when it should come quickly.
684 - Implement console @echo command.
685 - Implement a Slot priority (loaded/not loaded).
686 - Implement "vacation" Incremental only saves.
687 - Implement single pane restore (much like the Gftp panes).
688 - Implement Automatic Mount even in operator wait.
689 - Implement create "FileSet"?
690 - Fix watchdog pthread crash on Win32 (this is pthread_kill() Cygwin bug)
691 - Implement "scratch pool" where tapes are defined and can be
692 taken by any pool that needs them.
693 - Implement restore "current system", but take all files without
694 doing selection tree -- so that jobs without File records can
696 - Add prefixlinks to where or not where absolute links to FD.
697 - Issue message to mount a new tape before the rewind.
698 - Simplified client job initiation for portables.
699 - If SD cannot open a drive, make it periodically retry.
700 - Add more of the config info to the tape label.
702 - If tape is marked read-only, then try opening it read-only rather than
703 failing, and remember that it cannot be written.
704 - Refine SD waiting output:
705 Device is being positioned
706 > Device is being positioned for append
707 > Device is being positioned to file x
709 - Figure out some way to estimate output size and to avoid splitting
710 a backup across two Volumes -- this could be useful for writing CDROMs
711 where you really prefer not to have it split -- not serious.
712 - Have SD compute MD5 or SHA1 and compare to what FD computes.
713 - Make VolumeToCatalog calculate an MD5 or SHA1 from the
714 actual data on the Volume and compare it.
715 - Implement Bacula plugins -- design API
716 - Make bcopy read through bad tape records.
717 - Program files (i.e. execute a program to read/write files).
718 Pass read date of last backup, size of file last time.
719 - Add Signature type to File DB record.
720 - CD into subdirectory when open()ing files for backup to
721 speed up things. Test with testfind().
722 - Priority job to go to top of list.
723 - Why are save/restore of device different sizes (sparse?) Yup! Fix it.
724 - Implement some way for the Console to dynamically create a job.
725 - Restore to a particular time -- e.g. before date, after date.
726 - Solaris -I on tar for include list
727 - Need a verbose mode in restore, perhaps to bsr.
728 - bscan without -v is too quiet -- perhaps show jobs.
729 - Add code to reject whole blocks if not wanted on restore.
730 - Check if we can increase Bacula FD priorty in Win2000
731 - Make sure the MaxVolFiles is fully implemented in SD
732 - Check if both CatalogFiles and UseCatalog are set to SD.
733 - Figure out how to do a bare metal Windows restore
734 - Possibly add email to Watchdog if drive is unmounted too
735 long and a job is waiting on the drive.
736 - Restore program that errs in SD due to no tape, reports
737 OK incorrectly in output.
738 - After unmount, if restore job started, ask to mount.
739 - Convert all %x substitution variables, which are hard to remember
740 and read to %(variable-name). Idea from TMDA.
741 - Remove NextId for SQLite. Optimize.
742 - Move all SQL statements into a single location.
743 - Add UA rc and history files.
744 - put termcap (used by console) in ./configure and
745 allow -with-termcap-dir.
746 - Fix Autoprune for Volumes to respect need for full save.
747 - Fix Win32 config file definition name on /install
748 - Compare tape to Client files (attributes, or attributes and data)
749 - Make all database Ids 64 bit.
750 - Write an applet for Linux.
751 - Allow console commands to detach or run in background.
752 - Fix status delay on storage daemon during rewind.
753 - Add SD message variables to control operator wait time
754 - Maximum Operator Wait
755 - Minimum Message Interval
756 - Maximum Message Interval
757 - Send Operator message when cannot read tape label.
758 - Verify level=Volume (scan only), level=Data (compare of data to file).
759 Verify level=Catalog, level=InitCatalog
761 - Add keyword search to show command in Console.
762 - Events : tape has more than xxx bytes.
763 - Complete code in Bacula Resources -- this will permit
764 reading a new config file at any time.
765 - Handle ctl-c in Console
766 - Implement script driven addition of File daemon to config files.
767 - Think about how to make Bacula work better with File (non-tape) archives.
768 - Write Unix emulator for Windows.
769 - Put memory utilization in Status output of each daemon
770 if full status requested or if some level of debug on.
771 - Make database type selectable by .conf files i.e. at runtime
772 - Set flag for uname -a. Add to Volume label.
773 - Implement throttled work queue.
774 - Restore files modified after date
775 - SET LD_RUN_PATH=$HOME/mysql/lib/mysql
776 - Implement Restore FileSet=
777 - Create a protocol.h and protocol.c where all protocol messages
779 - Remove duplicate fields from jcr (e.g. jcr.level and jcr.jr.Level, ...).
780 - Timout a job or terminate if link goes down, or reopen link and query.
781 - Concept of precious tapes (cannot be reused).
782 - Make bcopy copy with a single tape drive.
783 - Permit changing ownership during restore.
786 > My suggestion: Add a feature on the systray menu-icon menu to request
787 > an immediate backup now. This would be useful for laptop users who may
788 > not be on the network when the regular scheduled backup is run.
790 > My wife's suggestion: Add a setting to the win32 client to allow it to
791 > shut down the machine after backup is complete (after, of course,
792 > displaying a "System will shut down in one minute, click here to cancel"
793 > warning dialog). This would be useful for sites that want user
794 > woorkstations to be shut down overnight to save power.
797 - Autolabel should be specified by DIR instead of SD.
800 - AutoScan (check checksum of tape)
801 - Format command = "format /dev/nst0"
805 - Seek resolution (usually corresponds to buffer size)
806 - EODErrorCode=ENOSPC or code
807 - Partial Read error code
808 - Partial write error code
809 - Nonformatted read error
810 - Nonformatted write error
811 - WriteProtected error
815 - IgnoreCloseErrors=yes
825 - FD sends unsaved file list to Director at end of job (see
827 - File daemon should build list of files skipped, and then
828 at end of save retry and report any errors.
829 - Write a Storage daemon that uses pipes and
830 standard Unix programs to write to the tape.
832 - Need something that monitors the JCR queue and
833 times out jobs by asking the deamons where they are.
834 - Enhance Jmsg code to permit buffering and saving to disk.
835 - device driver = "xxxx" for drives.
837 - Ensure that /dev/null works
838 - Need report class for messages. Perhaps
839 report resource where report=group of messages
840 - enhance scan_attrib and rename scan_jobtype, and
841 fill in code for "since" option
842 - Director needs a time after which the report status is sent
843 anyway -- or better yet, a retry time for the job.
844 - Don't reschedule a job if previous incarnation is still running.
845 - Some way to automatically backup everything is needed????
846 - Need a structure for pending actions:
848 - termination status (part of buffered msgs?)
850 Read, Write, Clean, Delete
851 - Login to Bacula; Bacula users with different permissions:
852 owner, group, user, quotas
853 - Store info on each file system type (probably in the job header on tape.
854 This could be the output of df; or perhaps some sort of /etc/mtab record.
857 - Design at hierarchial storage for Bacula. Migration and Clone.
858 - Implement FSM (File System Modules).
859 - Audit M_ error codes to ensure they are correct and consistent.
860 - Add variable break characters to lex analyzer.
861 Either a bit mask or a string of chars so that
862 the caller can change the break characters.
863 - Make a single T_BREAK to replace T_COMMA, etc.
864 - Ensure that File daemon and Storage daemon can
865 continue a save if the Director goes down (this
866 is NOT currently the case). Must detect socket error,
867 buffer messages for later.
868 - Enhance time/duration input to allow multiple qualifiers e.g. 3d2h
869 - Add ability to backup to two Storage devices (two SD sessions) at
870 the same time -- e.g. onsite, offsite.
871 - Add the ability to consolidate old backup sets (basically do a restore
872 to tape and appropriately update the catalog). Compress Volume sets.
873 Might need to spool via file is only one drive is available.
874 - Compress or consolidate Volumes of old possibly deleted files. Perhaps
875 someway to do so with every volume that has less than x% valid
879 Migration: Move a backup from one Volume to another
880 Clone: Copy a backup -- two Volumes
882 Bacula Migration is based on Jobs (apparently Networker is file by file).
884 Migration triggered by:
888 Highwater mark (keep total size)
893 ======================================================
895 It is somewhat like a Full save becomes an incremental since
896 the Base job (or jobs) plus other non-base files.
898 - A Base backup is same as Full backup, just different type.
899 - New BaseFiles table that contains:
901 BaseJobId - Base JobId referenced for this FileId (needed ???)
902 JobId - JobId currently running
903 FileId - File not backed up, exists in Base Job
904 FileIndex - FileIndex from Base Job.
905 i.e. for each base file that exists but is not saved because
906 it has not changed, the File daemon sends the JobId, BaseId,
907 FileId, FileIndex back to the Director who creates the DB entry.
908 - To initiate a Base save, the Director sends the FD
909 the FileId, and full filename for each file in the Base.
910 - When the FD finds a Base file, he requests the Director to
911 send him the full File entry (stat packet plus MD5), or
912 conversely, the FD sends it to the Director and the Director
913 says yes or no. This can be quite rapid if the FileId is kept
914 by the FD for each Base Filename.
915 - It is probably better to have the comparison done by the FD
916 despite the fact that the File entry must be sent across the
918 - An alternative would be to send the FD the whole File entry
919 from the start. The disadvantage is that it requires a lot of
920 space. The advantage is that it requires less communications
922 - The Job record must be updated to indicate that one or more
924 - At end of Job, FD returns:
925 1. Count of base files/bytes not written to tape (i.e. matches)
926 2. Count of base file that were saved i.e. had changed.
927 - No tape record would be written for a Base file that matches, in the
928 same way that no tape record is written for Incremental jobs where
929 the file is not saved because it is unchanged.
930 - On a restore, all the Base file records must explicitly be
931 found from the BaseFile tape. I.e. for each Full save that is marked
932 to have one or more Base Jobs, search the BaseFile for all occurrences
934 - An optimization might be to make the BaseFile have:
940 This would avoid the need to explicitly fetch each File record for
941 the Base job. The Base Job record will be fetched to get the
942 VolSessionId and VolSessionTime.
943 =========================================================
946 ==========================================================
948 For each Incremental job that is run, there may be files that
949 were found but not saved because they were locked (this applies
950 only to Windows). Such a system could send back to the Director
951 a list of Unsaved files.
953 - New UnSavedFiles table that contains:
957 - Then in the next Incremental job, the list of Unsaved Files will be
958 feed to the FD, who will ensure that they are explicitly chosen even
959 if standard date/time check would not have selected them.
960 =============================================================
963 Done: (see kernsdone for more)
966 - John's Full save failed with 1.32c FD and 1.31 Dir no FD status,
967 and no error message.
968 - Add fd and st as Console keywords.
969 - Recycling volume with a Slot requires an operator intervention:
970 rufus-dir: Start Backup JobId 18, Job=kernsave.2003-11-01_21.23.52
971 rufus-dir: Pruned 1 Job on Volume Vol01 from catalog.
972 rufus-dir: There are no Jobs associated with Volume Vol01. Marking it purged.
973 rufus-dir: Recycled volume "Vol01"
974 rufus-sd: Please mount Volume "Vol01" on Storage Device "DDS-4" for Job kernsave.2003-11-01_21.23.52
975 Use "mount" command to release Job.
976 - Implement Dan's bacula script (email of 26 Oct).
977 - Add JobName= to VerifyToCatalog so that all verifies can be done at the end.
978 - Edit the Client/Storage name into authentication failure messages.
979 - Fix packet too big problem. This is most likely a Windows TCP stack
981 - Implement ClientRunBeforeJob and ClientRunAfterJob.
982 - Implement forward spacing block/file: position_device(bsr) --
983 just before read_block_from_device();
986 Multiple drive autochanger data: see Alan Brown
987 mtx -f xxx unloadStorage Element 1 is Already Full(drive 0 was empty)
988 Unloading Data Transfer Element into Storage Element 1...source Element
991 (drive 0 was empty and so was slot 1)
992 > mtx -f xxx load 15 0
993 no response, just returns to the command prompt when complete.
994 > mtx -f xxx status Storage Changer /dev/changer:2 Drives, 60 Slots ( 2 Import/Export )
995 Data Transfer Element 0:Full (Storage Element 15 Loaded):VolumeTag = HX001
996 Data Transfer Element 1:Empty
997 Storage Element 1:Empty
998 Storage Element 2:Full :VolumeTag=HX002
999 Storage Element 3:Full :VolumeTag=HX003
1000 Storage Element 4:Full :VolumeTag=HX004
1001 Storage Element 5:Full :VolumeTag=HX005
1002 Storage Element 6:Full :VolumeTag=HX006
1003 Storage Element 7:Full :VolumeTag=HX007
1004 Storage Element 8:Full :VolumeTag=HX008
1005 Storage Element 9:Full :VolumeTag=HX009
1006 Storage Element 10:Full :VolumeTag=HX010
1007 Storage Element 11:Empty
1008 Storage Element 12:Empty
1009 Storage Element 13:Empty
1010 Storage Element 14:Empty
1011 Storage Element 15:Empty
1012 Storage Element 16:Empty....
1013 Storage Element 28:Empty
1014 Storage Element 29:Full :VolumeTag=CLNU01L1
1015 Storage Element 30:Empty....
1016 Storage Element 57:Empty
1017 Storage Element 58:Full :VolumeTag=NEX261L2
1018 Storage Element 59 IMPORT/EXPORT:Empty
1019 Storage Element 60 IMPORT/EXPORT:Empty
1021 Unloading Data Transfer Element into Storage Element 15...done
1023 (just to verify it remembers where it came from, however it can be
1024 overrriden with mtx unload {slotnumber} to go to any storage slot.)
1026 There needs to be a table of drive # to devices somewhere - If there are
1027 multiple changers or drives there may not be a 1:1 correspondance between
1028 changer drive number and system device name - and depending on the way the
1029 drives are hooked up to scsi busses, they may not be linearly numbered
1030 from an offset point either.something like
1032 Autochanger drives = 2
1033 Autochanger drive 0 = /dev/nst1
1034 Autochanger drive 1 = /dev/nst2
1035 IMHO, it would be _safest_ to use explicit mtx unload commands at all
1036 times, not just for multidrive changers. For a 1 drive changer, that's
1042 MTX's manpage (1.2.15):
1043 unload [<slotnum>] [ <drivenum> ]
1044 Unloads media from drive <drivenum> into slot
1045 <slotnum>. If <drivenum> is omitted, defaults to
1046 drive 0 (as do all commands). If <slotnum> is
1047 omitted, defaults to the slot that the drive was
1048 loaded from. Note that there's currently no way
1049 to say 'unload drive 1's media to the slot it
1050 came from', other than to explicitly use that
1051 slot number as the destination.AB
1057 undef# camcontrol devlist
1058 <WANGTEK 51000 SCSI M74H 12B3> at scbus0 target 2 lun 0 (pass0,sa0)
1059 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 0 (pass1,sa1)
1060 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 1 (pass2)
1062 tapeinfo -f /dev/sg0 with a bad tape in drive 1:
1063 [kern@rufus mtx-1.2.17kes]$ ./tapeinfo -f /dev/sg0
1064 Product Type: Tape Drive
1066 Product ID: 'C5713A '
1068 Attached Changer: No
1069 TapeAlert[3]: Hard Error: Uncorrectable read/write error.
1070 TapeAlert[20]: Clean Now: The tape drive neads cleaning NOW.
1077 Medium Type: Not Loaded
1080 DataCompEnabled: yes
1081 DataCompCapable: yes
1082 DataDeCompEnabled: yes
1089 Handling removable disks
1091 From: Karl Cunningham <karlc@keckec.com>
1093 My backups are only to hard disk these days, in removable bays. This is my
1094 idea of how a backup to hard disk would work more smoothly. Some of these
1095 things Bacula does already, but I mention them for completeness. If others
1096 have better ways to do this, I'd like to hear about it.
1098 1. Accommodate several disks, rotated similar to how tapes are. Identified
1099 by partition volume ID or perhaps by the name of a subdirectory.
1100 2. Abort & notify the admin if the wrong disk is in the bay.
1101 3. Write backups to different subdirectories for each machine to be backed
1103 4. Volumes (files) get created as needed in the proper subdirectory, one
1105 5. When a disk is recycled, remove or zero all old backup files. This is
1106 important as the disk being recycled may be close to full. This may be
1107 better done manually since the backup files for many machines may be
1108 scattered in many subdirectories.
1113 - To pass Include 1 or two letter commands
1114 I Name Include name - first record
1115 B Name Base name - repeat
1118 E zzz Exclude expression (wild card)
1120 D "reader" Reader program
1121 T "writer" Writer program
1122 O Options In current commpressed format
1123 (compression, signature, onefs, recurse, sparse,
1124 replace, verify options, ...)
1126 B BaseName Start second option set any letter
1129 F Number Number of filenames to follow
1134 F Number Number of filenames to follow
1137 - Add regression of btape "fill"
1139 - Implement "label slots=1,2... pool=yyy barcodes"
1140 - Mark a non-removable disk Volume in error rather than asking sys op to
1142 - Implement Fixed storage LabelFormat test.
1143 - Add reporting in attr despooling.
1144 - Figure out how to use ssh or stunnel to protect Bacula communications.
1145 Add Dan's work to manual See ssl.txt
1148 - Implement fast tree insert (doubly linked list)
1149 - Disallow using Internal database
1150 - Feedback while the restore tree is being built.
1151 - Add detection of Whitebox release in ./configure
1152 - Add InChanger to "list media"
1153 - Implement second cut of SIGHUP in Dir
1154 (Implement resources on a single pointer)
1155 - Add regular expressions to FileSets (needs testing).
1156 - Test new despooling error recovery code when partition fills.
1157 - Sort Scheduled jobs status listing by start time.
1158 - Add priority to Scheduled jobs status listing.
1159 - Look at Win32 inc problem when new directory added files not saved???
1160 - Require resource names to be unique.
1161 - Make sure that all errors from libacl are printed.
1162 - Make FD run on Win95 if possible (not really possible).
1163 - Fix GROUP BYs in all SQL (done by Dan -- thanks)
1164 - Make entering multiple Storage names in Dir illegal.
1165 - Make btape release the drive during the "test" append.
1167 You have selected the following JobIds: 739,856,797
1168 Building directory tree for JobId 739 ...
1169 Building directory tree for JobId 856 ...
1170 Building directory tree for JobId 797 ...3
1171 Jobs inserted into the tree.
1172 - Fix errno handling in win32 compat routines.
1173 - Add better error codes to run_program (10000+)
1174 - Revisit and revise Disaster Recovery (fix SCSI and RAID
1176 - When passwords do not match, print message that points the
1178 - Do tape alerts -- see tapealert.txt
1179 - Find memory leaks in address_conf.c
1180 - Network order of port and ip addresses.
1181 - SetIP in dird/fd_cmds.c is broken.
1182 - don't use inet_addr() it is deprecated.
1183 - Fix code in filed.c for init_bsock();
1184 - Fix code in console.c for init_bsock();
1185 - Solve the termcap.h problem on Solaris configure.
1186 - Make Verify jobs require exclusive use of Volume as Restore
1188 - Fix restore ++++ that get intermingled with "Building directory tree"
1189 - Fix doc in dirdconf.html for @, where the file must
1190 have not a list of files, but
1193 - Test Win32 errno handling.
1194 - port 1.34.6-block.patch to 1.35