7 Version 1.37 Kern (see below)
8 ========================================================
11 - Does ClientRunAfterJob fail the job on a bad return code?
12 - Document cleaning up the spool files:
13 db, pid, state, bsr, mail, conmsg, spool
14 - Document the multiple-drive-changer.txt script.
15 - Pruning with Admin job.
16 - Does WildFile match against full name? Doc.
17 - %d and %v only valid on Director, not for ClientRunBefore/After.
20 - Make hardlink code at line 240 of find_one.c use binary search.
21 - Queue warning/error messages during restore so that they
22 are reported at the end of the report rather than being
23 hidden in the file listing ...
24 - A Volume taken from Scratch should take on the retention period
26 - Correct doc for Maximum Changer Wait (and others) accepting only
28 - Fix Maximum Changer Wait (and others) to accept qualifiers.
29 - Look at -D_FORTIFY_SOURCE=2
30 - Add Win32 FileSet definition somewhere
31 - Look at fixing restore status stats in SD.
32 - Make selection of Database used in restore correspond to
34 - Implement a mode that says when a hard read error is
35 encountered, read many times (as it currently does), and if the
36 block cannot be read, skip to the next block, and try again. If
37 that fails, skip to the next file and try again, ...
39 create table LevelType (LevelType binary(1), LevelTypeLong tinyblob);
40 insert into LevelType (LevelType,LevelTypeLong) values
44 - Add ACL to restore only to original location.
45 - Add a recursive mark command (rmark) to restore.
46 - "Minimum Job Interval = nnn" sets minimum interval between Jobs
47 of the same level and does not permit multiple simultaneous
48 running of that Job (i.e. lets any previous invocation finish
49 before doing Interval testing).
50 - Look at simplifying File exclusions.
51 - Fix store_yesno to be store_bitmask.
52 - New directive "Delete purged Volumes"
53 - new pool XXX with ScratchPoolId = MyScratchPool's PoolId and
54 let it fill itself, and RecyclePoolId = XXX's PoolId so I can
55 see if it become stable and I just have to supervise
57 - If I want to remove this pool, I set RecyclePoolId = MyScratchPool's
58 PoolId, and when it is empty remove it.
59 - Figure out how to recycle Scratch volumes back to the Scratch
62 - Allow Check Labels to be used with Bacula labels.
63 - "Resuming" a failed backup (lost line for example) by using the
64 failed backup as a sort of "base" job.
66 - Email to the user when the tape is about to need changing x
67 days before it needs changing.
68 - Command to show next tape that will be used for a job even
69 if the job is not scheduled.
70 --- create_file.c.orig Fri Jul 8 12:13:05 2005
71 +++ create_file.c Fri Jul 8 12:13:07 2005
73 attr->ofname, be.strerror());
76 + } else if(S_ISSOCK(attr->statp.st_mode)) {
77 + Dmsg1(200, "Skipping socket: %s\n", attr->ofname);
79 Dmsg1(200, "Restore node: %s\n", attr->ofname);
80 if (mknod(attr->ofname, attr->statp.st_mode, attr->statp.st_rdev) != 0 && errno != EEXIST) {
81 - From: Arunav Mandal <amandal@trolltech.com>
82 1. When jobs are running and bacula for some reason crashes or if I do a
83 restart it remembers and jobs it was running before it crashed or restarted
84 as of now I loose all jobs if I restart it.
86 2. When spooling and in the midway if client is disconnected for instance a
87 laptop bacula completely discard the spool. It will be nice if it can write
88 that spool to tape so there will be some backups for that client if not all.
90 3. We have around 150 clients machines it will be nice to have a option to
91 upgrade all the client machines bacula version automatically.
93 4. Atleast one connection should be reserved for the bconsole so at heavy load
94 I should connect to the director via bconsole which at sometimes I can't
96 5. Another most important feature that is missing, say at 10am I manually
97 started backup of client abc and it was a full backup since client abc has
98 no backup history and at 10.30am bacula again automatically started backup of
99 client abc as that was in the schedule. So now we have 2 multiple Full
100 backups of the same client and if we again try to start a full backup of
101 client backup abc bacula won't complain. That should be fixed.
103 - Fix bpipe.c so that it does not modify results pointer.
104 ***FIXME*** calling sequence should be changed.
106 #3 Migration (Move, Copy, Archive Jobs)
107 #7 Single Job Writing to Multiple Storage Devices
108 - Reserve blocks other restore jobs when first cannot connect
110 - Add true/false to conf same as yes/no
111 - For Windows disaster recovery see http://unattended.sf.net/
112 - regardless of the retention period, Bacula will not prune the
113 last Full, Diff, or Inc File data until a month after the
114 retention period for the last Full backup that was done.
115 - update volume=xxx --- add status=Full
116 - Remove old spool files on startup.
117 - Exclude SD spool/working directory.
118 - Refuse to prune last valid Full backup. Same goes for Catalog.
120 - Make a callback when Rerun failed levels is called.
121 - Give Python program access to Scheduled jobs.
122 - Add setting Volume State via Python.
123 - Python script to save with Python, not save, save with Bacula.
124 - Python script to do backup.
126 - Change the Priority, Client, Storage, JobStatus (error)
127 at the start of a job.
128 - Why is SpoolDirectory = /home/bacula/spool; not reported
129 as an error when writing a DVD?
130 - Make bootstrap file handle multiple MediaTypes (SD)
131 - Remove all old Device resource code in Dir and code to pass it
132 back in SD -- better, rework it to pass back device statistics.
133 - Check locking of resources -- be sure to lock devices where previously
134 resources were locked.
135 - The last part is left in the spool dir.
138 - In restore don't compare byte count on a raw device -- directory
139 entry does not contain bytes.
140 - To mark files as deleted, run essentially a Verify to disk, and
141 when a file is found missing (MarkId != JobId), then create
142 a new File record with FileIndex == -1. This could be done
143 by the FD at the same time as the backup.
147 MA = (last_MA * 3 + rate) / 4
148 rate = (bytes - last_bytes) / (runtime - last_runtime)
149 - Max Vols limit in Pool off by one?
150 - Implement Files/Bytes,... stats for restore job.
151 - Implement Total Bytes Written, ... for restore job.
152 - Despool attributes simultaneously with data in a separate
153 thread, rejoined at end of data spooling.
154 - 7. Implement new Console commands to allow offlining/reserving drives,
155 and possibly manipulating the autochanger (much asked for).
156 - Add start/end date editing in messages (%t %T, %e?) ...
157 - Add ClientDefs similar to JobDefs.
158 - Print more info when bextract -p accepts a bad block.
159 - Fix FD JobType to be set before RunBeforeJob in FD.
160 - Look at adding full Volume and Pool information to a Volume
161 label so that bscan can get *all* the info.
162 - If the user puts "Purge Oldest Volume = yes" or "Recycle Oldest Volume = yes"
163 and there is only one volume in the pool, refuse to do it -- otherwise
164 he fills the Volume, then immediately starts reusing it.
165 - Implement copies and stripes.
166 - Add history file to console.
167 - Each file on tape creates a JobMedia record. Peter has 4 million
168 files spread over 10000 tape files and four tapes. A restore takes
169 16 hours to build the restore list.
170 - Add and option to check if the file size changed during backup.
171 - Make sure SD deletes spool files on error exit.
172 - Delete old spool files when SD starts.
173 - When labeling tapes, if you enter 000026, Bacula uses
174 the tape index rather than the Volume name 000026.
175 - Add offline tape command to Bacula console.
177 Enter MediaId or Volume name: 32
178 Enter new Volume name: DLT-20Dec04
179 Automatically selected Pool: Default
180 Connecting to Storage daemon DLTDrive at 192.168.68.104:9103 ...
181 Sending relabel command from "DLT-28Jun03" to "DLT-20Dec04" ...
182 block.c:552 Write error at 0:0 on device /dev/nst0. ERR=Bad file descriptor.
183 Error writing final EOF to tape. This tape may not be readable.
184 dev.c:1207 ioctl MTWEOF error on /dev/nst0. ERR=Permission denied.
185 askdir.c:219 NULL Volume name. This shouldn't happen!!!
186 3912 Failed to label Volume: ERR=dev.c:1207 ioctl MTWEOF error on /dev/nst0. ERR=Permission denied.
187 Label command failed for Volume DLT-20Dec04.
188 Do not forget to mount the drive!!!
189 - Bug: if a job is manually scheduled to run later, it does not appear
190 in any status report and cannot be cancelled.
192 ==== Keeping track of deleted files ====
193 My "trick" for keeping track of deletions is the following.
194 Assuming the user turns on this option, after all the files
195 have been backed up, but before the job has terminated, the
196 FD will make a pass through all the files and send their
197 names to the DIR (*exactly* the same as what a Verify job
198 currently does). This will probably be done at the same
199 time the files are being sent to the SD avoiding a second
200 pass. The DIR will then compare that to what is stored in
201 the catalog. Any files in the catalog but not in what the
202 FD sent will receive a catalog File entry that indicates
203 that at that point in time the file was deleted.
205 During a restore, any file initially picked up by some
206 backup (Full, ...) then subsequently having a File entry
207 marked "delete" will be removed from the tree, so will not
208 be restored. If a file with the same name is later OK it
209 will be inserted in the tree -- this already happens. All
210 will be consistent except for possible changes during the
213 Since I'm on the subject, some of you may be wondering what
214 the utility of the in memory tree is if you are going to
215 restore everything (at least it comes up from time to time
216 on the list). Well, it is still *very* useful because it
217 allows only the last item found for a particular filename
218 (full path) to be entered into the tree, and thus if a file
219 is backed up 10 times, only the last copy will be restored.
220 I recently (last Friday) restored a complete directory, and
221 the Full and all the Differential and Incremental backups
222 spanned 3 Volumes. The first Volume was not even mounted
223 because all the files had been updated and hence backed up
224 since the Full backup was made. In this case, the tree
225 saved me a *lot* of time.
227 Make sure this information is stored on the tape too so
228 that it can be restored directly from the tape.
230 Comments from Martin Simmons (I think they are all covered):
231 Ok, that should cover the basics. There are few issues though:
233 - Restore will depend on the catalog. I think it is better to include the
234 extra data in the backup as well, so it can be seen by bscan and bextract.
236 - I'm not sure if it will preserve multiple hard links to the same inode. Or
237 maybe adding or removing links will cause the data to be dumped again?
239 - I'm not sure if it will handle renamed directories. Possibly it will work
240 by dumping the whole tree under a renamed directory?
242 - It remains to be seen how the backup performance of the DIR's will be
243 affected when comparing the catalog for a large filesystem.
248 - Add Pool/Storage override regression test.
249 - Add delete JobId to regression.
250 - Add a regression test for dbcheck.
251 - New test to add bscan to four-concurrent-jobs regression,
252 i.e. after the four-concurrent jobs zap the
253 database as is done in the bscan-test, then use bscan to
254 restore the database, do a restore and compare with the
256 - Add restore of specific JobId to regression (item 3
257 on the restore prompt)
258 - Add IPv6 to regression
259 - Add database test to regression. Test each function like delete,
262 - AntiVir can slow down backups on Win32 systems.
263 - Win32 systems with FAT32 can be much slower than NTFS for
264 more than 1000 files per directory.
268 - A HOLD command to stop all jobs from starting.
269 - A PAUSE command to pause all running jobs ==> release the
271 - Media Type = LTO,LTO-2,LTO-3
272 Media Type Read = LTO,LTO2,LTO3
273 Media Type Write = LTO2, LTO3
275 === From Carsten Menke <bootsy52@gmx.net>
277 Following is a list of what I think in the situations where I'm faced with,
278 could be a usefull enhancement to bacula, which I'm certain other users will
279 benefit from as well.
281 1. NextJob/NextJobs Directive within a Job Resource in the form of
282 NextJobs = job1,job2.
285 I currently solved the problem with running multiple jobs each after each
286 by setting the Max Wait Time for a job to 8 hours, and give
287 the jobs different Priorities. However, there scenarios where
288 1 Job is directly depending on another job, so if the former job fails,
289 the job after it needn't to be run
290 while maybe other jobs should run despite of that
293 A Backup Job and a Verify job, if the backup job fails there is no need to run
294 the verify job, as the backup job already failed. However, one may like
295 to backup the Catalog to disk despite of that the main backup job failed.
298 I see that this is related to the Event Handlers which are on the ToDo
299 list, also it is maybe a good idea to check for the return value and
300 execute different actions based on the return value
303 3. offline capability to bconsole
306 Currently I use a script which I execute within the last Job via the
307 RunAfterJob Directive, to release and eject the tape.
308 So I have to call bconsole "release=Storage-Name" and afterwards
309 mt -f /dev/nst0 eject to get the tape out.
311 If I have multiple Storage Devices, than these may not be /dev/nst0 and
312 I have to modify the script or call it with parameters etc.
313 This would actually not be needed, as everything is already defined
314 in bacula-sd.conf and if I can invoke bconsole with the
315 storage name via $1 in the script than I'm done and information is
318 4. %s for Storage Name added to the chars being substituted in "RunAfterJob"
322 For the reason mentioned in 3. to have the ability to call a
323 script with /scripts/foobar %s and in the script use $1
324 to pass the Storage Name to bconsole
326 5. Setting Volume State within a Job Resource
329 Instead of using "Maximum Volume Jobs" in the Pool Resource,
330 I would have the possibilty to define
331 in a Job Resource that after this certain job is run, the Volume State
332 should be set to "Volume State = Used", this give more flexibility (IMHO).
334 6. Localization of Bacula Messages
337 Unfortunatley many,many people I work with don't speak english very well.
338 So if at least the Reporting messages would be localized then they
339 would understand that they have to change the tape,etc. etc.
341 I volunteer to do the german translations, and if I can convince my wife also
342 french and Morre (western african language).
344 7. OK, this is evil, probably bound to security risks and maybe not possible
345 due to the design of bacula.
347 Implementation of Backtics ( `command` ) for shell comand execution to
348 the "Label Format" Directive.
352 Currently I have defined BACULA_DAY_OF_WEEK="day1|day2..." resulting in
353 Label Format = "HolyBackup-${BACULA_DAY_OF_WEEK[${WeekDay}]}". If I could
354 use backticks than I could use "Label Format = HolyBackup-`date +%A` to have
355 the localized name for the day of the week appended to the
356 format string. Then I have the tape labeled automatically with weekday
357 name in the correct language.
359 - Yes, that is surely the case. I probably should turn those into Warning
360 errors. In addition, you just made me think that it might not be bad to
361 add an option to check the file size after backing up the file and
362 report if it changes. This would be done as an option because it would
365 Kern, good idea. If you do do that, mention in the output: file
366 shrunk, or file expanded, just to make it obvious to the user
367 (without having to the refer to file size), just how the file size
370 Would this option be for all file, or just one file? Or a fileset?
371 - Make output from status use html table tags for nicely
372 presenting in a browser.
373 - Can one write tapes faster with 8192 byte block sizes?
374 - Document security problems with the same password for everyone in
375 rpm and Win32 releases.
376 - Browse generations of files.
377 - I've seen an error when my catalog's File table fills up. I
378 then have to recreate the File table with a larger maximum row
379 size. Relevant information is at
380 http://dev.mysql.com/doc/mysql/en/Full_table.html ; I think the
381 "Installing and Configuring MySQL" chapter should talk a bit
382 about this potential problem, and recommend a solution.
383 - For Solaris must use POSIX awk.
384 - Want speed of writing to tape while despooling.
385 - Supported autochanger:
393 Wangtek 6525ES (SCSI-1 QIC drive, 525MB), under Linux 2.4.something,
394 bacula 1.36.0/1 works with blocksize 16k INSIDE bacula-sd.conf.
395 - Add regex from http://www.pcre.org to Bacula for Win32.
396 - Use only shell tools no make in CDROM package.
397 - Include within include does it work?
398 - Implement a Pool of type Cleaning?
399 - Implement VolReadTime and VolWriteTime in SD
400 - Modify Backing up Your Database to include a bootstrap file.
401 - Think about making certain database errors fatal.
402 - Look at correcting the time jump in the scheduler for daylight
403 savings time changes.
404 - Add a "real" timer to network connections.
405 - Promote to Full = Time period
406 - Check dates entered by user for correctness (month/day/... ranges)
407 - Compress restore Volume listing by date and first file.
408 - Look at patches/bacula_db.b2z postgresql that loops during restore.
410 - Perhaps add read/write programs and/or plugins to FileSets.
411 - How to handle backing up portables ...
412 - Add some sort of guaranteed Interval for upgrading jobs.
413 - Can we write the state file after every job terminates? On Win32
414 the system crashes and the state file is not updated.
417 Documentation to do: (any release a little bit at a time)
418 - Doc to do unmount before removing magazine.
419 - Alternative to static linking "ldd prog" save all binaries listed,
420 restore them and point LD_LIBRARY_PATH to them.
421 - Document add "</dev/null >/dev/null 2>&1" to the bacula-fd command line
422 - Document query file format.
423 - Add more documentation for bsr files.
424 - Document problems with Verify and pruning.
425 - Document how to use multiple databases.
426 - VXA drives have a "cleaning required"
427 indicator, but Exabyte recommends preventive cleaning after every 75
430 In this context, it should be noted that Exabyte has a command-line
431 vxatool utility available for free download. (The current version is
432 vxatool-3.72.) It can get diagnostic info, read, write and erase tapes,
433 test the drive, unload tapes, change drive settings, flash new firmware,
435 Of particular interest in this context is that vxatool <device> -i will
436 report, among other details, the time since last cleaning in tape motion
437 minutes. This information can be retrieved (and settings changed, for
438 that matter) through the generic-SCSI device even when Bacula has the
439 regular tape device locked. (Needless to say, I don't recommend
440 changing tape settings while a job is running.)
441 - Lookup HP cleaning recommendations.
442 - Lookup HP tape replacement recommendations (see trouble shooting autochanger)
443 - Document doing table repair
446 ===================================
447 - Add macro expansions in JobDefs.
448 Run Before Job = "SomeFile %{Level} %{Client}"
449 Write Bootstrap="/some/dir/%{JobName}_%{Client}.bsr"
450 - Use non-blocking network I/O but if no data is available, use
452 - Use gather write() for network I/O.
453 - Autorestart on crash.
454 - Add bandwidth limiting.
455 - Add acks every once and a while from the SD to keep
456 the line from timing out.
457 - When an error in input occurs and conio beeps, you can back
458 up through the prompt.
459 - Detect fixed tape block mode during positioning by looking at
460 block numbers in btape "test". Possibly adjust in Bacula.
461 - Fix list volumes to output volume retention in some other
462 units, perhaps via a directive.
463 - If opening a tape in read/write mode fails attempt to open
464 it in read-only mode, and mark the tape for read only.
465 - Allow Simultaneous Priorities = yes => run up to Max concurrent jobs even
466 with multiple priorities.
467 - If you use restore replace=never, the directory attributes for
468 non-existent directories will not be restored properly.
470 - see lzma401.zip in others directory for new compression
472 - Minimal autochanger handling in Bacula and in btape.
473 - Look into how tar does not save sockets and the possiblity of
474 not saving them in Bacula (Martin Simmons reported this).
475 - Fix restore jobs so that multiple jobs can run if they
476 are not using the same tape(s).
477 - Allow the user to select JobType for manual pruning/purging.
478 - bscan does not put first of two volumes back with all info in
480 - Implement the FreeBSD nodump flag in chflags.
481 - Figure out how to make named console messages go only to that
482 console and to the non-restricted console (new console class?).
483 - Make restricted console prompt for password if *ask* is set or
484 perhaps if password is undefined.
485 - Implement "from ISO-date/time every x hours/days/weeks/months" in
488 ==== from Marc Schoechlin
489 - the help-command should be more verbose
490 (it should explain the paramters of the different
492 -> its time-comsuming to consult the manual anytime
493 you need a special parameter
494 -> maybe its more easy to maintain this, if the
495 descriptions of that commands are outsourced to
497 - the cd-command should allow complete paths
498 i.e. cd /foo/bar/foo/bar
499 -> if a customer mails me the path to a certain file,
500 its faster to enter the specified directory
501 - if the password is not configured in bconsole.conf
502 you should be asked for it.
503 -> sometimes you like to do restore on a customer-machine
504 which shouldnt know the password for bacula.
505 -> adding the password to the file favours admins
506 to forget to remove the password after usage
508 the protection of that file is less important
509 - long-listed-output of commands should be scrollable
510 like the unix more/less-command does
511 -> if someone runs 200 and more machines, the lists could
512 be a little long and complex
513 - command-output should be shown column by column
514 to reduce scrolling and to increase clarity
516 - lsmark should list the selected files with full
518 - wildcards for selecting and file and directories would be nice
519 - any actions should be interuptable with STRG+C
520 - command-expansion would be pretty cool
522 - When the replace Never option is set, new directory permissions
523 are not restored. See bug 213. To fix this requires creating a
524 list of newly restored directories so that those directory
525 permissions *can* be restored.
526 - Compaction of Disk space by "migrating" Volumes that have pruned
527 Jobs (what criteria? size, #jobs, time).
528 - Add prune all command
529 - Document fact that purge can destroy a part of a restore by purging
530 one volume while others remain valid -- perhaps mark Jobs.
531 - Add multiple-media-types.txt
532 - look at mxt-changer.html
533 - Make ? do a help command (no return needed).
534 - Implement restore directory.
535 - Document streams and how to implement them.
536 - Try not to re-backup a file if a new hard link is added.
537 - Add feature to backup hard links only, but not the data.
538 - Fix stream handling to be simpler.
539 - Add Priority and Bootstrap to Run a Job.
540 - Eliminate Restore "Run Restore Job" prompt by allowing new "run command
542 - Remove View FileSet button from Run a Job dialog.
543 - Handle prompt for restore job at end of Restore command.
544 - Add display of total selected files to Restore window.
545 - Add tree pane to left of window.
546 - Add progress meter.
547 - Max wait time or max run time causes seg fault -- see runtime-bug.txt
548 - Document writing to a CD/DVD with Bacula.
549 - Add a "base" package to the window installer for pthreadsVCE.dll
550 which is needed by all packages.
551 - Add message to user to check for fixed block size when the forward
552 space test fails in btape.
553 - When unmarking a directory check if all files below are unmarked and
554 then remove the + flag -- in the restore tree.
555 - Possibly implement: Action = Unmount Device="TapeDrive1" in Admin jobs.
556 - Setup lrrd graphs: (http://www.linpro.no/projects/lrrd/) Mike Acar.
557 - Revisit the question of multiple Volumes (disk) on a single device.
558 - Add a block copy option to bcopy.
559 - Investigate adding Mac Resource Forks.
560 - Finish work on Gnome restore GUI.
561 - Fix "llist jobid=xx" where no fileset or client exists.
562 - For each job type (Admin, Restore, ...) require only the really necessary
563 fields.- Pass Director resource name as an option to the Console.
564 - Add a "batch" mode to the Console (no unsolicited queries, ...).
565 - Add a .list all files in the restore tree (probably also a list all files)
566 Do both a long and short form.
567 - Allow browsing the catalog to see all versions of a file (with
568 stat data on each file).
569 - Restore attributes of directory if replace=never set but directory
571 - Use SHA1 on authentication if possible.
572 - See comtest-xxx.zip for Windows code to talk to USB.
573 - Add John's appended files:
574 Appended = { /files/server/logs/http/*log }
575 and such files would be treated as follows.On a FULL backup, they would
576 be backed up like any other file.On an INCREMENTAL backup, where a
577 previous INCREMENTAL or FULL was already in thecatalogue and the length
578 of the file wasgreater than the length of the last backup, only thedata
579 added since the last backup will be dumped.On an INCREMENTAL backup, if
580 the length of the file is less than thelength of the file with the same
581 name last backed up, the completefile is dumped.On Windows systems, with
582 creation date of files, we can be evensmarter about this and not count
583 entirely upon the length.On a restore, the full and all incrementals
584 since it will beapplied in sequence to restore the file.
585 - Check new HAVE_WIN32 open bits.
586 - Check if the tape has moved before writing.
587 - Handling removable disks -- see below:
588 - Keep track of tape use time, and report when cleaning is necessary.
589 - Add FromClient and ToClient keywords on restore command (or
590 BackupClient RestoreClient).
591 - Implement a JobSet, which groups any number of jobs. If the
592 JobSet is started, all the jobs are started together.
593 Allow Pool, Level, and Schedule overrides.
594 - Enhance cancel to timeout BSOCK packets after a specific delay.
595 - Do scheduling by UTC using gmtime_r() in run_conf, scheduler, and
596 ua_status.!!! Thanks to Alan Brown for this tip.
597 - Look at updating Volume Jobs so that Max Volume Jobs = 1 will work
598 correctly for multiple simultaneous jobs.
599 - Correct code so that FileSet MD5 is calculated for < and | filename
601 - Implement the Media record flag that indicates that the Volume does disk
603 - Implement VolAddr, which is used when Volume is addressed like a disk,
604 and form it from VolFile and VolBlock.
605 - Make multiple restore jobs for multiple media types specifying
606 the proper storage type.
607 - Fix fast block rejection (stored/read_record.c:118). It passes a null
608 pointer (rec) to try_repositioning().
609 - Look at extracting Win data from BackupRead.
610 - Implement RestoreJobRetention? Maybe better "JobRetention" in a Job,
611 which would take precidence over the Catalog "JobRetention".
612 - Implement Label Format in Add and Label console commands.
613 - Possibly up network buffers to 65K. Put on variable.
614 - Put email tape request delays on one or more variables. User wants
615 to cancel the job after a certain time interval. Maximum Mount Wait?
616 - Job, Client, Device, Pool, or Volume?
617 Is it possible to make this a directive which is *optional* in multiple
618 resources, like Level? If so, I think I'd make it an optional directive
619 in Job, Client, and Pool, with precedence such that Job overrides Client
620 which in turn overrides Pool.
622 - New Storage specifications:
623 - Want to write to multiple storage devices simultaneously
624 - Want to write to multiple storage devices sequentially (in one job)
625 - Want to read/write simultaneously
626 - Key is MediaType -- it must match
628 Passed to SD as a sort of BSR record called Storage Specification
632 MediaType -> Next MediaType
634 Device -> Next Device
636 Allow multiple Storage specifications
644 Allow Multiple Pool specifications (note, Pool currently
646 Allow Multiple MediaType specifications in Dir conf
647 Allow Multiple Device specifications in Dir conf
648 Perhaps keep this in a single SSR
649 Tie a Volume to a specific device by using a MediaType that
650 is contained in only one device.
651 In SD allow Device to have Multiple MediaTypes
653 - Ideas from Jerry Scharf:
654 First let's point out some big pluses that bacula has for this
656 more importantly it's active. Thank you so much for that
657 even more important, it's not flaky
658 it has an open access catalog, opening many possibilities
659 it's pushing toward heterogeneous systems capability
661 Macintosh file client
662 macs are an interesting niche, but I fear a server is a rathole
663 working bare iron recovery for windows
664 the option for inc/diff backups not reset on fileset revision
665 a) use both change and inode update time against base time
666 b) do the full catalog check (expensive but accurate)
667 sizing guide (how much system is needed to back up N systems/files)
668 consultants on using bacula in building a disaster recovery system
670 or how to get at fancy things that one could do with bacula
671 logwatch code for bacula logs (or similar)
672 linux distro inclusion of bacula (brings good and bad, but necessary)
673 win2k/XP server capability (icky but you asked)
674 support for Oracle database ??
676 - Look at adding SQL server and Exchange support for Windows.
677 - Make dev->file and dev->block_num signed integers so that -1 can
678 be an invalid value which happens with BSR.
679 - Create VolAddr for disk files in place of VolFile and VolBlock. This
680 is needed to properly specify ranges.
681 - Add progress of files/bytes to SD and FD.
682 - Print warning message if FileId > 4 billion
683 - do a "messages" before the first prompt in Console
684 - Client does not show busy during Estimate command.
685 - Implement Console mtx commands.
686 - Implement a Mount Command and an Unmount Command where
687 the users could specify a system command to be performed
688 to do the mount, after which Bacula could attempt to
689 read the device. This is for Removeable media such as a CDROM.
690 - Most likely, this mount command would be invoked explicitly
691 by the user using the current Console "mount" and "unmount"
692 commands -- the Storage Daemon would do the right thing
693 depending on the exact nature of the device.
694 - As with tape drives, when Bacula wanted a new removable
695 disk mounted, it would unmount the old one, and send a message
696 to the user, who would then use "mount" as described above
697 once he had actually inserted the disk.
698 - Implement dump/print label to UA
699 - Spool to disk only when the tape is full, then when a tape is hung move
701 - bextract is sending everything to the log file ****FIXME****
702 - Allow multiple Storage specifications (or multiple names on
703 a single Storage specification) in the Job record. Thus a job
704 can be backed up to a number of storage devices.
705 - Implement some way for the File daemon to contact the Director
706 to start a job or pass its DHCP obtained IP number.
707 - Implement a query tape prompt/replace feature for a console
708 - Copy console @ code to gnome2-console
709 - Make AES the only encryption algorithm see
710 http://csrc.nist.gov/CryptoToolkit/aes/). It's
711 an officially adopted standard, has survived peer
712 review, and provides keys up to 256 bits.
713 - Take a careful look at SetACL http://setacl.sourceforge.net
714 - Make tree walk routines like cd, ls, ... more user friendly
715 by handling spaces better.
716 - Make sure that Bacula rechecks the tape after the 20 min wait.
717 - Set IO_NOWAIT on Bacula TCP/IP packets.
718 - Try doing a raw partition backup and restore by mounting a
721 Yes, it would allow to highly automatic the request for new tapes. If a
722 tape is empty, bacula reads the barcodes (native or simulated), and if
723 an unused tape is found, it runs the label command with all the
724 necessary parameters.
726 By the way can bacula automatically "move" an empty/purged volume say
727 in the "short" pool to the "long" pool if this pool runs out of volume
729 - What to do about "list files job=xxx".
730 - Get and test MySQL 4.0
731 - Look at how fuser works and /proc/PID/fd that is how Nic found the
732 file descriptor leak in Bacula.
733 - Implement WrapCounters in Counters.
734 - Add heartbeat from FD to SD if hb interval expires.
735 - Can we dynamically change FileSets?
736 - If pool specified to label command and Label Format is specified,
737 automatically generate the Volume name.
738 - Why can't SQL do the filename sort for restore?
739 - Add ExhautiveRestoreSearch
740 - Look at the possibility of loading only the necessary
741 data into the restore tree (i.e. do it one directory at a
742 time as the user walks through the tree).
743 - Possibly use the hash code if the user selects all for a restore command.
744 - Fix "restore all" to bypass building the tree.
745 - Prohibit backing up archive device (findlib/find_one.c:128)
746 - Implement Release Device in the Job resource to unmount a drive.
747 - Implement Acquire Device in the Job resource to mount a drive,
748 be sure this works with admin jobs so that the user can get
749 prompted to insert the correct tape. Possibly some way to say to
750 run the job but don't save the files.
751 - Make things like list where a file is saved case independent for
754 - Use autochanger to handle multiple devices.
755 - On Windows with very long path names, it may be impossible to create
756 a file (and thus restore it) because the total length is too long.
757 We must cd into the directory then create the file without the
759 - Implement a Recycle command
760 - Test a second language e.g. french.
761 - Start working on Base jobs.
762 - Implement UnsavedFiles DB record.
763 - From Phil Stracchino:
764 It would probably be a per-client option, and would be called
765 something like, say, "Automatically purge obsoleted jobs". What it
766 would do is, when you successfully complete a Differential backup of a
767 client, it would automatically purge all Incremental backups for that
768 client that are rendered redundant by that Differential. Likewise,
769 when a Full backup on a client completed, it would automatically purge
770 all Differential and Incremental jobs obsoleted by that Full backup.
771 This would let people minimize the number of tapes they're keeping on
772 hand without having to master the art of retention times.
773 - When doing a Backup send all attributes back to the Director, who
774 would then figure out what files have been deleted.
775 - Currently in mount.c:236 the SD simply creates a Volume. It should have
776 explicit permission to do so. It should also mark the tape in error
777 if there is an error.
778 - Cancel waiting for Client connect in SD if FD goes away.
780 - Implement timeout in response() when it should come quickly.
781 - Implement a Slot priority (loaded/not loaded).
782 - Implement "vacation" Incremental only saves.
783 - Implement create "FileSet"?
784 - Add prefixlinks to where or not where absolute links to FD.
785 - Issue message to mount a new tape before the rewind.
786 - Simplified client job initiation for portables.
787 - If SD cannot open a drive, make it periodically retry.
788 - Add more of the config info to the tape label.
790 - If tape is marked read-only, then try opening it read-only rather than
791 failing, and remember that it cannot be written.
792 - Refine SD waiting output:
793 Device is being positioned
794 > Device is being positioned for append
795 > Device is being positioned to file x
797 - Figure out some way to estimate output size and to avoid splitting
798 a backup across two Volumes -- this could be useful for writing CDROMs
799 where you really prefer not to have it split -- not serious.
800 - Have SD compute MD5 or SHA1 and compare to what FD computes.
801 - Make VolumeToCatalog calculate an MD5 or SHA1 from the
802 actual data on the Volume and compare it.
803 - Implement Bacula plugins -- design API
804 - Make bcopy read through bad tape records.
805 - Program files (i.e. execute a program to read/write files).
806 Pass read date of last backup, size of file last time.
807 - Add Signature type to File DB record.
808 - CD into subdirectory when open()ing files for backup to
809 speed up things. Test with testfind().
810 - Priority job to go to top of list.
811 - Why are save/restore of device different sizes (sparse?) Yup! Fix it.
812 - Implement some way for the Console to dynamically create a job.
813 - Solaris -I on tar for include list
814 - Need a verbose mode in restore, perhaps to bsr.
815 - bscan without -v is too quiet -- perhaps show jobs.
816 - Add code to reject whole blocks if not wanted on restore.
817 - Check if we can increase Bacula FD priorty in Win2000
818 - Make sure the MaxVolFiles is fully implemented in SD
819 - Check if both CatalogFiles and UseCatalog are set to SD.
820 - Possibly add email to Watchdog if drive is unmounted too
821 long and a job is waiting on the drive.
822 - After unmount, if restore job started, ask to mount.
823 - Add UA rc and history files.
824 - put termcap (used by console) in ./configure and
825 allow -with-termcap-dir.
826 - Fix Autoprune for Volumes to respect need for full save.
827 - Compare tape to Client files (attributes, or attributes and data)
828 - Make all database Ids 64 bit.
829 - Allow console commands to detach or run in background.
830 - Fix status delay on storage daemon during rewind.
831 - Add SD message variables to control operator wait time
832 - Maximum Operator Wait
833 - Minimum Message Interval
834 - Maximum Message Interval
835 - Send Operator message when cannot read tape label.
836 - Verify level=Volume (scan only), level=Data (compare of data to file).
837 Verify level=Catalog, level=InitCatalog
839 - Add keyword search to show command in Console.
840 - Events : tape has more than xxx bytes.
841 - Complete code in Bacula Resources -- this will permit
842 reading a new config file at any time.
843 - Handle ctl-c in Console
844 - Implement script driven addition of File daemon to config files.
845 - Think about how to make Bacula work better with File (non-tape) archives.
846 - Write Unix emulator for Windows.
847 - Put memory utilization in Status output of each daemon
848 if full status requested or if some level of debug on.
849 - Make database type selectable by .conf files i.e. at runtime
850 - Set flag for uname -a. Add to Volume label.
851 - Restore files modified after date
852 - SET LD_RUN_PATH=$HOME/mysql/lib/mysql
853 - Remove duplicate fields from jcr (e.g. jcr.level and jcr.jr.Level, ...).
854 - Timout a job or terminate if link goes down, or reopen link and query.
855 - Concept of precious tapes (cannot be reused).
856 - Make bcopy copy with a single tape drive.
857 - Permit changing ownership during restore.
860 > My suggestion: Add a feature on the systray menu-icon menu to request
861 > an immediate backup now. This would be useful for laptop users who may
862 > not be on the network when the regular scheduled backup is run.
864 > My wife's suggestion: Add a setting to the win32 client to allow it to
865 > shut down the machine after backup is complete (after, of course,
866 > displaying a "System will shut down in one minute, click here to cancel"
867 > warning dialog). This would be useful for sites that want user
868 > woorkstations to be shut down overnight to save power.
871 - Autolabel should be specified by DIR instead of SD.
874 - AutoScan (check checksum of tape)
875 - Format command = "format /dev/nst0"
879 - Seek resolution (usually corresponds to buffer size)
880 - EODErrorCode=ENOSPC or code
881 - Partial Read error code
882 - Partial write error code
883 - Nonformatted read error
884 - Nonformatted write error
885 - WriteProtected error
889 - IgnoreCloseErrors=yes
899 - FD sends unsaved file list to Director at end of job (see
901 - File daemon should build list of files skipped, and then
902 at end of save retry and report any errors.
903 - Write a Storage daemon that uses pipes and
904 standard Unix programs to write to the tape.
906 - Need something that monitors the JCR queue and
907 times out jobs by asking the deamons where they are.
908 - Enhance Jmsg code to permit buffering and saving to disk.
909 - device driver = "xxxx" for drives.
911 - Ensure that /dev/null works
912 - Need report class for messages. Perhaps
913 report resource where report=group of messages
914 - enhance scan_attrib and rename scan_jobtype, and
915 fill in code for "since" option
916 - Director needs a time after which the report status is sent
917 anyway -- or better yet, a retry time for the job.
918 - Don't reschedule a job if previous incarnation is still running.
919 - Some way to automatically backup everything is needed????
920 - Need a structure for pending actions:
922 - termination status (part of buffered msgs?)
924 Read, Write, Clean, Delete
925 - Login to Bacula; Bacula users with different permissions:
926 owner, group, user, quotas
927 - Store info on each file system type (probably in the job header on tape.
928 This could be the output of df; or perhaps some sort of /etc/mtab record.
930 ========= ideas ===============
931 From: "Jerry K. Schieffer" <jerry@skylinetechnology.com>
932 To: <kern@sibbald.com>
933 Subject: RE: [Bacula-users] future large programming jobs
934 Date: Thu, 26 Feb 2004 11:34:54 -0600
936 I noticed the subject thread and thought I would offer the following
937 merely as sources of ideas, i.e. something to think about, not even as
938 strong as a request. In my former life (before retiring) I often
939 dealt with backups and storage management issues/products as a
940 developer and as a consultant. I am currently migrating my personal
941 network from amanda to bacula specifically because of the ability to
942 cross media boundaries during storing backups.
943 Are you familiar with the commercial product called ADSM (I think IBM
944 now sells it under the Tivoli label)? It has a couple of interesting
945 ideas that may apply to the following topics.
947 1. Migration: Consider that when you need to restore a system, there
948 may be pressure to hurry. If all the information for a single client
949 can eventually end up on the same media (and in chronological order),
950 the restore is facillitated by not having to search past information
951 from other clients. ADSM has the concept of "client affinity" that
952 may be associated with it's storage pools. It seems to me that this
953 concept (as an optional feature) might fit in your architecture for
956 ADSM also has the concept of defining one or more storage pools as
957 "copy pools" (almost mirrors, but only in the sense of contents).
958 These pools provide the ability to have duplicte data stored both
959 onsite and offsite. The copy process can be scheduled to be handled
960 by their storage manager during periods when there is no backup
961 activity. Again, the migration process might be a place to consider
962 implementing something like this.
965 > It strikes me that it would be very nice to be able to do things
967 > have the Job(s) backing up the machines run, and once they have all
968 > completed, start a migration job to copy the data from disks Volumes
970 > a tape library and then to offsite storage. Maybe this can already
972 > done with some careful scheduling and Job prioritzation; the events
973 > mechanism described below would probably make it very easy.
975 This is the goal. In the first step (before events), you simply
977 the Migration to tape later.
979 2. Base jobs: In ADSM, each copy of each stored file is tracked in
980 the database. Once a file (unique by path and metadata such as dates,
981 size, ownership, etc.) is in a copy pool, no more copies are made. In
982 other words, when you start ADSM, it begins like your concept of a
983 base job. After that it is in the "incremental" mode. You can
984 configure the number of "generations" of files to be retained, plus a
985 retention date after which even old generations are purged. The
986 database tracks the contents of media and projects the percentage of
987 each volume that is valid. When the valid content of a volume drops
988 below a configured percentage, the valid data are migrated to another
989 volume and the old volume is marked as empty. Note, this requires
990 ADSM to have an idea of the contents of a client, i.e. marking the
991 database when an existing file was deleted, but this would solve your
992 issue of restoring a client without restoring deleted files.
994 This is pretty far from what bacula now does, but if you are going to
995 rip things up for Base jobs,.....
996 Also, the benefits of this are huge for very large shops, especially
997 with media robots, but are a pain for shops with manual media
1001 > Base jobs sound pretty useful, but I'm not dying for them.
1003 Nobody is dying for them, but when you see what it does, you will die
1006 3. Restoring deleted files: Since I think my comments in (2) above
1007 have low probability of implementation, I'll also suggest that you
1008 could approach the issue of deleted files by a mechanism of having the
1009 fd report to the dir, a list of all files on the client for every
1010 backup job. The dir could note in the database entry for each file
1011 the date that the file was seen. Then if a restore as of date X takes
1012 place, only files that exist from before X until after X would be
1013 restored. Probably the major cost here is the extra date container in
1014 each row of the files table.
1016 Thanks for "listening". I hope some of this helps. If you want to
1017 contact me, please send me an email - I read some but not all of the
1018 mailing list traffic and might miss a reply there.
1020 Please accept my compliments for bacula. It is doing a great job for
1021 me!! I sympathize with you in the need to wrestle with excelence in
1022 execution vs. excelence in feature inclusion.
1027 ==============================
1030 - Design at hierarchial storage for Bacula. Migration and Clone.
1031 - Implement FSM (File System Modules).
1032 - Audit M_ error codes to ensure they are correct and consistent.
1033 - Add variable break characters to lex analyzer.
1034 Either a bit mask or a string of chars so that
1035 the caller can change the break characters.
1036 - Make a single T_BREAK to replace T_COMMA, etc.
1037 - Ensure that File daemon and Storage daemon can
1038 continue a save if the Director goes down (this
1039 is NOT currently the case). Must detect socket error,
1040 buffer messages for later.
1041 - Enhance time/duration input to allow multiple qualifiers e.g. 3d2h
1042 - Add ability to backup to two Storage devices (two SD sessions) at
1043 the same time -- e.g. onsite, offsite.
1044 - Add the ability to consolidate old backup sets (basically do a restore
1045 to tape and appropriately update the catalog). Compress Volume sets.
1046 Might need to spool via file is only one drive is available.
1047 - Compress or consolidate Volumes of old possibly deleted files. Perhaps
1048 someway to do so with every volume that has less than x% valid
1052 Migration: Move a backup from one Volume to another
1053 Clone: Copy a backup -- two Volumes
1055 Bacula Migration is based on Jobs (apparently Networker is file by file).
1057 Migration triggered by:
1061 Highwater mark (keep total size)
1066 ======================================================
1068 It is somewhat like a Full save becomes an incremental since
1069 the Base job (or jobs) plus other non-base files.
1071 - A Base backup is same as Full backup, just different type.
1072 - New BaseFiles table that contains:
1074 BaseJobId - Base JobId referenced for this FileId (needed ???)
1075 JobId - JobId currently running
1076 FileId - File not backed up, exists in Base Job
1077 FileIndex - FileIndex from Base Job.
1078 i.e. for each base file that exists but is not saved because
1079 it has not changed, the File daemon sends the JobId, BaseId,
1080 FileId, FileIndex back to the Director who creates the DB entry.
1081 - To initiate a Base save, the Director sends the FD
1082 the FileId, and full filename for each file in the Base.
1083 - When the FD finds a Base file, he requests the Director to
1084 send him the full File entry (stat packet plus MD5), or
1085 conversely, the FD sends it to the Director and the Director
1086 says yes or no. This can be quite rapid if the FileId is kept
1087 by the FD for each Base Filename.
1088 - It is probably better to have the comparison done by the FD
1089 despite the fact that the File entry must be sent across the
1091 - An alternative would be to send the FD the whole File entry
1092 from the start. The disadvantage is that it requires a lot of
1093 space. The advantage is that it requires less communications
1095 - The Job record must be updated to indicate that one or more
1097 - At end of Job, FD returns:
1098 1. Count of base files/bytes not written to tape (i.e. matches)
1099 2. Count of base file that were saved i.e. had changed.
1100 - No tape record would be written for a Base file that matches, in the
1101 same way that no tape record is written for Incremental jobs where
1102 the file is not saved because it is unchanged.
1103 - On a restore, all the Base file records must explicitly be
1104 found from the BaseFile tape. I.e. for each Full save that is marked
1105 to have one or more Base Jobs, search the BaseFile for all occurrences
1107 - An optimization might be to make the BaseFile have:
1113 This would avoid the need to explicitly fetch each File record for
1114 the Base job. The Base Job record will be fetched to get the
1115 VolSessionId and VolSessionTime.
1116 =========================================================
1119 ==========================================================
1121 For each Incremental job that is run, there may be files that
1122 were found but not saved because they were locked (this applies
1123 only to Windows). Such a system could send back to the Director
1124 a list of Unsaved files.
1126 - New UnSavedFiles table that contains:
1130 - Then in the next Incremental job, the list of Unsaved Files will be
1131 feed to the FD, who will ensure that they are explicitly chosen even
1132 if standard date/time check would not have selected them.
1133 =============================================================
1137 Multiple drive autochanger data: see Alan Brown
1138 mtx -f xxx unloadStorage Element 1 is Already Full(drive 0 was empty)
1139 Unloading Data Transfer Element into Storage Element 1...source Element
1140 Address 480 is Empty
1142 (drive 0 was empty and so was slot 1)
1143 > mtx -f xxx load 15 0
1144 no response, just returns to the command prompt when complete.
1145 > mtx -f xxx status Storage Changer /dev/changer:2 Drives, 60 Slots ( 2 Import/Export )
1146 Data Transfer Element 0:Full (Storage Element 15 Loaded):VolumeTag = HX001
1147 Data Transfer Element 1:Empty
1148 Storage Element 1:Empty
1149 Storage Element 2:Full :VolumeTag=HX002
1150 Storage Element 3:Full :VolumeTag=HX003
1151 Storage Element 4:Full :VolumeTag=HX004
1152 Storage Element 5:Full :VolumeTag=HX005
1153 Storage Element 6:Full :VolumeTag=HX006
1154 Storage Element 7:Full :VolumeTag=HX007
1155 Storage Element 8:Full :VolumeTag=HX008
1156 Storage Element 9:Full :VolumeTag=HX009
1157 Storage Element 10:Full :VolumeTag=HX010
1158 Storage Element 11:Empty
1159 Storage Element 12:Empty
1160 Storage Element 13:Empty
1161 Storage Element 14:Empty
1162 Storage Element 15:Empty
1163 Storage Element 16:Empty....
1164 Storage Element 28:Empty
1165 Storage Element 29:Full :VolumeTag=CLNU01L1
1166 Storage Element 30:Empty....
1167 Storage Element 57:Empty
1168 Storage Element 58:Full :VolumeTag=NEX261L2
1169 Storage Element 59 IMPORT/EXPORT:Empty
1170 Storage Element 60 IMPORT/EXPORT:Empty
1172 Unloading Data Transfer Element into Storage Element 15...done
1174 (just to verify it remembers where it came from, however it can be
1175 overrriden with mtx unload {slotnumber} to go to any storage slot.)
1177 There needs to be a table of drive # to devices somewhere - If there are
1178 multiple changers or drives there may not be a 1:1 correspondance between
1179 changer drive number and system device name - and depending on the way the
1180 drives are hooked up to scsi busses, they may not be linearly numbered
1181 from an offset point either.something like
1183 Autochanger drives = 2
1184 Autochanger drive 0 = /dev/nst1
1185 Autochanger drive 1 = /dev/nst2
1186 IMHO, it would be _safest_ to use explicit mtx unload commands at all
1187 times, not just for multidrive changers. For a 1 drive changer, that's
1193 MTX's manpage (1.2.15):
1194 unload [<slotnum>] [ <drivenum> ]
1195 Unloads media from drive <drivenum> into slot
1196 <slotnum>. If <drivenum> is omitted, defaults to
1197 drive 0 (as do all commands). If <slotnum> is
1198 omitted, defaults to the slot that the drive was
1199 loaded from. Note that there's currently no way
1200 to say 'unload drive 1's media to the slot it
1201 came from', other than to explicitly use that
1202 slot number as the destination.AB
1208 undef# camcontrol devlist
1209 <WANGTEK 51000 SCSI M74H 12B3> at scbus0 target 2 lun 0 (pass0,sa0)
1210 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 0 (pass1,sa1)
1211 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 1 (pass2)
1213 tapeinfo -f /dev/sg0 with a bad tape in drive 1:
1214 [kern@rufus mtx-1.2.17kes]$ ./tapeinfo -f /dev/sg0
1215 Product Type: Tape Drive
1217 Product ID: 'C5713A '
1219 Attached Changer: No
1220 TapeAlert[3]: Hard Error: Uncorrectable read/write error.
1221 TapeAlert[20]: Clean Now: The tape drive neads cleaning NOW.
1228 Medium Type: Not Loaded
1231 DataCompEnabled: yes
1232 DataCompCapable: yes
1233 DataDeCompEnabled: yes
1240 Handling removable disks
1242 From: Karl Cunningham <karlc@keckec.com>
1244 My backups are only to hard disk these days, in removable bays. This is my
1245 idea of how a backup to hard disk would work more smoothly. Some of these
1246 things Bacula does already, but I mention them for completeness. If others
1247 have better ways to do this, I'd like to hear about it.
1249 1. Accommodate several disks, rotated similar to how tapes are. Identified
1250 by partition volume ID or perhaps by the name of a subdirectory.
1251 2. Abort & notify the admin if the wrong disk is in the bay.
1252 3. Write backups to different subdirectories for each machine to be backed
1254 4. Volumes (files) get created as needed in the proper subdirectory, one
1256 5. When a disk is recycled, remove or zero all old backup files. This is
1257 important as the disk being recycled may be close to full. This may be
1258 better done manually since the backup files for many machines may be
1259 scattered in many subdirectories.
1264 - Make sure that all do_prompt() calls in Dir check for
1265 -1 (error) and -2 (cancel) returns.