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.
19 - Make sure that all do_prompt() calls in Dir check for
20 -1 (error) and -2 (cancel) returns.
21 - Look at -D_FORTIFY_SOURCE=2
22 - Add Win32 FileSet definition somewhere
23 - Look at fixing restore status stats in SD.
24 - Make selection of Database used in restore correspond to
26 - Implement a mode that says when a hard read error is
27 encountered, read many times (as it currently does), and if the
28 block cannot be read, skip to the next block, and try again. If
29 that fails, skip to the next file and try again, ...
31 create table LevelType (LevelType binary(1), LevelTypeLong tinyblob);
32 insert into LevelType (LevelType,LevelTypeLong) values
36 - Add ACL to restore only to original location.
37 - Add a recursive mark command (rmark) to restore.
38 - "Minimum Job Interval = nnn" sets minimum interval between Jobs
39 of the same level and does not permit multiple simultaneous
40 running of that Job (i.e. lets any previous invocation finish
41 before doing Interval testing).
42 - Look at simplifying File exclusions.
43 - Fix store_yesno to be store_bitmask.
44 - New directive "Delete purged Volumes"
45 - new pool XXX with ScratchPoolId = MyScratchPool's PoolId and
46 let it fill itself, and RecyclePoolId = XXX's PoolId so I can
47 see if it become stable and I just have to supervise
49 - If I want to remove this pool, I set RecyclePoolId = MyScratchPool's
50 PoolId, and when it is empty remove it.
51 - Figure out how to recycle Scratch volumes back to the Scratch
54 - Allow Check Labels to be used with Bacula labels.
55 - "Resuming" a failed backup (lost line for example) by using the
56 failed backup as a sort of "base" job.
58 - Email to the user when the tape is about to need changing x
59 days before it needs changing.
60 - Command to show next tape that will be used for a job even
61 if the job is not scheduled.
62 --- create_file.c.orig Fri Jul 8 12:13:05 2005
63 +++ create_file.c Fri Jul 8 12:13:07 2005
65 attr->ofname, be.strerror());
68 + } else if(S_ISSOCK(attr->statp.st_mode)) {
69 + Dmsg1(200, "Skipping socket: %s\n", attr->ofname);
71 Dmsg1(200, "Restore node: %s\n", attr->ofname);
72 if (mknod(attr->ofname, attr->statp.st_mode, attr->statp.st_rdev) != 0 && errno != EEXIST) {
73 - From: Arunav Mandal <amandal@trolltech.com>
74 1. When jobs are running and bacula for some reason crashes or if I do a
75 restart it remembers and jobs it was running before it crashed or restarted
76 as of now I loose all jobs if I restart it.
78 2. When spooling and in the midway if client is disconnected for instance a
79 laptop bacula completely discard the spool. It will be nice if it can write
80 that spool to tape so there will be some backups for that client if not all.
82 3. We have around 150 clients machines it will be nice to have a option to
83 upgrade all the client machines bacula version automatically.
85 4. Atleast one connection should be reserved for the bconsole so at heavy load
86 I should connect to the director via bconsole which at sometimes I can't
88 5. Another most important feature that is missing, say at 10am I manually
89 started backup of client abc and it was a full backup since client abc has
90 no backup history and at 10.30am bacula again automatically started backup of
91 client abc as that was in the schedule. So now we have 2 multiple Full
92 backups of the same client and if we again try to start a full backup of
93 client backup abc bacula won't complain. That should be fixed.
95 - Fix bpipe.c so that it does not modify results pointer.
96 ***FIXME*** calling sequence should be changed.
98 #3 Migration (Move, Copy, Archive Jobs)
99 #7 Single Job Writing to Multiple Storage Devices
100 - Reserve blocks other restore jobs when first cannot connect
102 - Add true/false to conf same as yes/no
103 - For Windows disaster recovery see http://unattended.sf.net/
104 - regardless of the retention period, Bacula will not prune the
105 last Full, Diff, or Inc File data until a month after the
106 retention period for the last Full backup that was done.
107 - update volume=xxx --- add status=Full
108 - Remove old spool files on startup.
109 - Exclude SD spool/working directory.
110 - Refuse to prune last valid Full backup. Same goes for Catalog.
112 - Make a callback when Rerun failed levels is called.
113 - Give Python program access to Scheduled jobs.
114 - Add setting Volume State via Python.
115 - Python script to save with Python, not save, save with Bacula.
116 - Python script to do backup.
118 - Change the Priority, Client, Storage, JobStatus (error)
119 at the start of a job.
120 - Why is SpoolDirectory = /home/bacula/spool; not reported
121 as an error when writing a DVD?
122 - Make bootstrap file handle multiple MediaTypes (SD)
123 - Remove all old Device resource code in Dir and code to pass it
124 back in SD -- better, rework it to pass back device statistics.
125 - Check locking of resources -- be sure to lock devices where previously
126 resources were locked.
127 - The last part is left in the spool dir.
130 - In restore don't compare byte count on a raw device -- directory
131 entry does not contain bytes.
132 - To mark files as deleted, run essentially a Verify to disk, and
133 when a file is found missing (MarkId != JobId), then create
134 a new File record with FileIndex == -1. This could be done
135 by the FD at the same time as the backup.
139 MA = (last_MA * 3 + rate) / 4
140 rate = (bytes - last_bytes) / (runtime - last_runtime)
141 - Max Vols limit in Pool off by one?
142 - Implement Files/Bytes,... stats for restore job.
143 - Implement Total Bytes Written, ... for restore job.
144 - Despool attributes simultaneously with data in a separate
145 thread, rejoined at end of data spooling.
146 - 7. Implement new Console commands to allow offlining/reserving drives,
147 and possibly manipulating the autochanger (much asked for).
148 - Add start/end date editing in messages (%t %T, %e?) ...
149 - Add ClientDefs similar to JobDefs.
150 - Print more info when bextract -p accepts a bad block.
151 - Fix FD JobType to be set before RunBeforeJob in FD.
152 - Look at adding full Volume and Pool information to a Volume
153 label so that bscan can get *all* the info.
154 - If the user puts "Purge Oldest Volume = yes" or "Recycle Oldest Volume = yes"
155 and there is only one volume in the pool, refuse to do it -- otherwise
156 he fills the Volume, then immediately starts reusing it.
157 - Implement copies and stripes.
158 - Add history file to console.
159 - Each file on tape creates a JobMedia record. Peter has 4 million
160 files spread over 10000 tape files and four tapes. A restore takes
161 16 hours to build the restore list.
162 - Add and option to check if the file size changed during backup.
163 - Make sure SD deletes spool files on error exit.
164 - Delete old spool files when SD starts.
165 - When labeling tapes, if you enter 000026, Bacula uses
166 the tape index rather than the Volume name 000026.
167 - Add offline tape command to Bacula console.
169 Enter MediaId or Volume name: 32
170 Enter new Volume name: DLT-20Dec04
171 Automatically selected Pool: Default
172 Connecting to Storage daemon DLTDrive at 192.168.68.104:9103 ...
173 Sending relabel command from "DLT-28Jun03" to "DLT-20Dec04" ...
174 block.c:552 Write error at 0:0 on device /dev/nst0. ERR=Bad file descriptor.
175 Error writing final EOF to tape. This tape may not be readable.
176 dev.c:1207 ioctl MTWEOF error on /dev/nst0. ERR=Permission denied.
177 askdir.c:219 NULL Volume name. This shouldn't happen!!!
178 3912 Failed to label Volume: ERR=dev.c:1207 ioctl MTWEOF error on /dev/nst0. ERR=Permission denied.
179 Label command failed for Volume DLT-20Dec04.
180 Do not forget to mount the drive!!!
181 - Bug: if a job is manually scheduled to run later, it does not appear
182 in any status report and cannot be cancelled.
184 ==== Keeping track of deleted files ====
185 My "trick" for keeping track of deletions is the following.
186 Assuming the user turns on this option, after all the files
187 have been backed up, but before the job has terminated, the
188 FD will make a pass through all the files and send their
189 names to the DIR (*exactly* the same as what a Verify job
190 currently does). This will probably be done at the same
191 time the files are being sent to the SD avoiding a second
192 pass. The DIR will then compare that to what is stored in
193 the catalog. Any files in the catalog but not in what the
194 FD sent will receive a catalog File entry that indicates
195 that at that point in time the file was deleted.
197 During a restore, any file initially picked up by some
198 backup (Full, ...) then subsequently having a File entry
199 marked "delete" will be removed from the tree, so will not
200 be restored. If a file with the same name is later OK it
201 will be inserted in the tree -- this already happens. All
202 will be consistent except for possible changes during the
205 Since I'm on the subject, some of you may be wondering what
206 the utility of the in memory tree is if you are going to
207 restore everything (at least it comes up from time to time
208 on the list). Well, it is still *very* useful because it
209 allows only the last item found for a particular filename
210 (full path) to be entered into the tree, and thus if a file
211 is backed up 10 times, only the last copy will be restored.
212 I recently (last Friday) restored a complete directory, and
213 the Full and all the Differential and Incremental backups
214 spanned 3 Volumes. The first Volume was not even mounted
215 because all the files had been updated and hence backed up
216 since the Full backup was made. In this case, the tree
217 saved me a *lot* of time.
219 Make sure this information is stored on the tape too so
220 that it can be restored directly from the tape.
222 Comments from Martin Simmons (I think they are all covered):
223 Ok, that should cover the basics. There are few issues though:
225 - Restore will depend on the catalog. I think it is better to include the
226 extra data in the backup as well, so it can be seen by bscan and bextract.
228 - I'm not sure if it will preserve multiple hard links to the same inode. Or
229 maybe adding or removing links will cause the data to be dumped again?
231 - I'm not sure if it will handle renamed directories. Possibly it will work
232 by dumping the whole tree under a renamed directory?
234 - It remains to be seen how the backup performance of the DIR's will be
235 affected when comparing the catalog for a large filesystem.
240 - Add Pool/Storage override regression test.
241 - Add delete JobId to regression.
242 - Add a regression test for dbcheck.
243 - New test to add bscan to four-concurrent-jobs regression,
244 i.e. after the four-concurrent jobs zap the
245 database as is done in the bscan-test, then use bscan to
246 restore the database, do a restore and compare with the
248 - Add restore of specific JobId to regression (item 3
249 on the restore prompt)
250 - Add IPv6 to regression
251 - Add database test to regression. Test each function like delete,
254 - AntiVir can slow down backups on Win32 systems.
255 - Win32 systems with FAT32 can be much slower than NTFS for
256 more than 1000 files per directory.
260 - A HOLD command to stop all jobs from starting.
261 - A PAUSE command to pause all running jobs ==> release the
263 - Media Type = LTO,LTO-2,LTO-3
264 Media Type Read = LTO,LTO2,LTO3
265 Media Type Write = LTO2, LTO3
267 === From Carsten Menke <bootsy52@gmx.net>
269 Following is a list of what I think in the situations where I'm faced with,
270 could be a usefull enhancement to bacula, which I'm certain other users will
271 benefit from as well.
273 1. NextJob/NextJobs Directive within a Job Resource in the form of
274 NextJobs = job1,job2.
277 I currently solved the problem with running multiple jobs each after each
278 by setting the Max Wait Time for a job to 8 hours, and give
279 the jobs different Priorities. However, there scenarios where
280 1 Job is directly depending on another job, so if the former job fails,
281 the job after it needn't to be run
282 while maybe other jobs should run despite of that
285 A Backup Job and a Verify job, if the backup job fails there is no need to run
286 the verify job, as the backup job already failed. However, one may like
287 to backup the Catalog to disk despite of that the main backup job failed.
290 I see that this is related to the Event Handlers which are on the ToDo
291 list, also it is maybe a good idea to check for the return value and
292 execute different actions based on the return value
295 3. offline capability to bconsole
298 Currently I use a script which I execute within the last Job via the
299 RunAfterJob Directive, to release and eject the tape.
300 So I have to call bconsole "release=Storage-Name" and afterwards
301 mt -f /dev/nst0 eject to get the tape out.
303 If I have multiple Storage Devices, than these may not be /dev/nst0 and
304 I have to modify the script or call it with parameters etc.
305 This would actually not be needed, as everything is already defined
306 in bacula-sd.conf and if I can invoke bconsole with the
307 storage name via $1 in the script than I'm done and information is
310 4. %s for Storage Name added to the chars being substituted in "RunAfterJob"
314 For the reason mentioned in 3. to have the ability to call a
315 script with /scripts/foobar %s and in the script use $1
316 to pass the Storage Name to bconsole
318 5. Setting Volume State within a Job Resource
321 Instead of using "Maximum Volume Jobs" in the Pool Resource,
322 I would have the possibilty to define
323 in a Job Resource that after this certain job is run, the Volume State
324 should be set to "Volume State = Used", this give more flexibility (IMHO).
326 6. Localization of Bacula Messages
329 Unfortunatley many,many people I work with don't speak english very well.
330 So if at least the Reporting messages would be localized then they
331 would understand that they have to change the tape,etc. etc.
333 I volunteer to do the german translations, and if I can convince my wife also
334 french and Morre (western african language).
336 7. OK, this is evil, probably bound to security risks and maybe not possible
337 due to the design of bacula.
339 Implementation of Backtics ( `command` ) for shell comand execution to
340 the "Label Format" Directive.
344 Currently I have defined BACULA_DAY_OF_WEEK="day1|day2..." resulting in
345 Label Format = "HolyBackup-${BACULA_DAY_OF_WEEK[${WeekDay}]}". If I could
346 use backticks than I could use "Label Format = HolyBackup-`date +%A` to have
347 the localized name for the day of the week appended to the
348 format string. Then I have the tape labeled automatically with weekday
349 name in the correct language.
351 - Yes, that is surely the case. I probably should turn those into Warning
352 errors. In addition, you just made me think that it might not be bad to
353 add an option to check the file size after backing up the file and
354 report if it changes. This would be done as an option because it would
357 Kern, good idea. If you do do that, mention in the output: file
358 shrunk, or file expanded, just to make it obvious to the user
359 (without having to the refer to file size), just how the file size
362 Would this option be for all file, or just one file? Or a fileset?
363 - Make output from status use html table tags for nicely
364 presenting in a browser.
365 - Can one write tapes faster with 8192 byte block sizes?
366 - Document security problems with the same password for everyone in
367 rpm and Win32 releases.
368 - Browse generations of files.
369 - I've seen an error when my catalog's File table fills up. I
370 then have to recreate the File table with a larger maximum row
371 size. Relevant information is at
372 http://dev.mysql.com/doc/mysql/en/Full_table.html ; I think the
373 "Installing and Configuring MySQL" chapter should talk a bit
374 about this potential problem, and recommend a solution.
375 - For Solaris must use POSIX awk.
376 - Want speed of writing to tape while despooling.
377 - Supported autochanger:
385 Wangtek 6525ES (SCSI-1 QIC drive, 525MB), under Linux 2.4.something,
386 bacula 1.36.0/1 works with blocksize 16k INSIDE bacula-sd.conf.
387 - Add regex from http://www.pcre.org to Bacula for Win32.
388 - Use only shell tools no make in CDROM package.
389 - Include within include does it work?
390 - Implement a Pool of type Cleaning?
391 - Implement VolReadTime and VolWriteTime in SD
392 - Modify Backing up Your Database to include a bootstrap file.
393 - Think about making certain database errors fatal.
394 - Look at correcting the time jump in the scheduler for daylight
395 savings time changes.
396 - Add a "real" timer to network connections.
397 - Promote to Full = Time period
398 - Check dates entered by user for correctness (month/day/... ranges)
399 - Compress restore Volume listing by date and first file.
400 - Look at patches/bacula_db.b2z postgresql that loops during restore.
402 - Perhaps add read/write programs and/or plugins to FileSets.
403 - How to handle backing up portables ...
404 - Add some sort of guaranteed Interval for upgrading jobs.
405 - Can we write the state file after every job terminates? On Win32
406 the system crashes and the state file is not updated.
409 Documentation to do: (any release a little bit at a time)
410 - Doc to do unmount before removing magazine.
411 - Alternative to static linking "ldd prog" save all binaries listed,
412 restore them and point LD_LIBRARY_PATH to them.
413 - Document add "</dev/null >/dev/null 2>&1" to the bacula-fd command line
414 - Document query file format.
415 - Add more documentation for bsr files.
416 - Document problems with Verify and pruning.
417 - Document how to use multiple databases.
418 - VXA drives have a "cleaning required"
419 indicator, but Exabyte recommends preventive cleaning after every 75
422 In this context, it should be noted that Exabyte has a command-line
423 vxatool utility available for free download. (The current version is
424 vxatool-3.72.) It can get diagnostic info, read, write and erase tapes,
425 test the drive, unload tapes, change drive settings, flash new firmware,
427 Of particular interest in this context is that vxatool <device> -i will
428 report, among other details, the time since last cleaning in tape motion
429 minutes. This information can be retrieved (and settings changed, for
430 that matter) through the generic-SCSI device even when Bacula has the
431 regular tape device locked. (Needless to say, I don't recommend
432 changing tape settings while a job is running.)
433 - Lookup HP cleaning recommendations.
434 - Lookup HP tape replacement recommendations (see trouble shooting autochanger)
435 - Document doing table repair
438 ===================================
439 - Add macro expansions in JobDefs.
440 Run Before Job = "SomeFile %{Level} %{Client}"
441 Write Bootstrap="/some/dir/%{JobName}_%{Client}.bsr"
442 - Use non-blocking network I/O but if no data is available, use
444 - Use gather write() for network I/O.
445 - Autorestart on crash.
446 - Add bandwidth limiting.
447 - Add acks every once and a while from the SD to keep
448 the line from timing out.
449 - When an error in input occurs and conio beeps, you can back
450 up through the prompt.
451 - Detect fixed tape block mode during positioning by looking at
452 block numbers in btape "test". Possibly adjust in Bacula.
453 - Fix list volumes to output volume retention in some other
454 units, perhaps via a directive.
455 - If opening a tape in read/write mode fails attempt to open
456 it in read-only mode, and mark the tape for read only.
457 - Allow Simultaneous Priorities = yes => run up to Max concurrent jobs even
458 with multiple priorities.
459 - If you use restore replace=never, the directory attributes for
460 non-existent directories will not be restored properly.
462 - see lzma401.zip in others directory for new compression
464 - Minimal autochanger handling in Bacula and in btape.
465 - Look into how tar does not save sockets and the possiblity of
466 not saving them in Bacula (Martin Simmons reported this).
467 - Fix restore jobs so that multiple jobs can run if they
468 are not using the same tape(s).
469 - Allow the user to select JobType for manual pruning/purging.
470 - bscan does not put first of two volumes back with all info in
472 - Implement the FreeBSD nodump flag in chflags.
473 - Figure out how to make named console messages go only to that
474 console and to the non-restricted console (new console class?).
475 - Make restricted console prompt for password if *ask* is set or
476 perhaps if password is undefined.
477 - Implement "from ISO-date/time every x hours/days/weeks/months" in
480 ==== from Marc Schoechlin
481 - the help-command should be more verbose
482 (it should explain the paramters of the different
484 -> its time-comsuming to consult the manual anytime
485 you need a special parameter
486 -> maybe its more easy to maintain this, if the
487 descriptions of that commands are outsourced to
489 - the cd-command should allow complete paths
490 i.e. cd /foo/bar/foo/bar
491 -> if a customer mails me the path to a certain file,
492 its faster to enter the specified directory
493 - if the password is not configured in bconsole.conf
494 you should be asked for it.
495 -> sometimes you like to do restore on a customer-machine
496 which shouldnt know the password for bacula.
497 -> adding the password to the file favours admins
498 to forget to remove the password after usage
500 the protection of that file is less important
501 - long-listed-output of commands should be scrollable
502 like the unix more/less-command does
503 -> if someone runs 200 and more machines, the lists could
504 be a little long and complex
505 - command-output should be shown column by column
506 to reduce scrolling and to increase clarity
508 - lsmark should list the selected files with full
510 - wildcards for selecting and file and directories would be nice
511 - any actions should be interuptable with STRG+C
512 - command-expansion would be pretty cool
514 - When the replace Never option is set, new directory permissions
515 are not restored. See bug 213. To fix this requires creating a
516 list of newly restored directories so that those directory
517 permissions *can* be restored.
518 - Compaction of Disk space by "migrating" Volumes that have pruned
519 Jobs (what criteria? size, #jobs, time).
520 - Add prune all command
521 - Document fact that purge can destroy a part of a restore by purging
522 one volume while others remain valid -- perhaps mark Jobs.
523 - Add multiple-media-types.txt
524 - look at mxt-changer.html
525 - Make ? do a help command (no return needed).
526 - Implement restore directory.
527 - Document streams and how to implement them.
528 - Try not to re-backup a file if a new hard link is added.
529 - Add feature to backup hard links only, but not the data.
530 - Fix stream handling to be simpler.
531 - Add Priority and Bootstrap to Run a Job.
532 - Eliminate Restore "Run Restore Job" prompt by allowing new "run command
534 - Remove View FileSet button from Run a Job dialog.
535 - Handle prompt for restore job at end of Restore command.
536 - Add display of total selected files to Restore window.
537 - Add tree pane to left of window.
538 - Add progress meter.
539 - Max wait time or max run time causes seg fault -- see runtime-bug.txt
540 - Document writing to a CD/DVD with Bacula.
541 - Add a "base" package to the window installer for pthreadsVCE.dll
542 which is needed by all packages.
543 - Add message to user to check for fixed block size when the forward
544 space test fails in btape.
545 - When unmarking a directory check if all files below are unmarked and
546 then remove the + flag -- in the restore tree.
547 - Possibly implement: Action = Unmount Device="TapeDrive1" in Admin jobs.
548 - Setup lrrd graphs: (http://www.linpro.no/projects/lrrd/) Mike Acar.
549 - Revisit the question of multiple Volumes (disk) on a single device.
550 - Add a block copy option to bcopy.
551 - Investigate adding Mac Resource Forks.
552 - Finish work on Gnome restore GUI.
553 - Fix "llist jobid=xx" where no fileset or client exists.
554 - For each job type (Admin, Restore, ...) require only the really necessary
555 fields.- Pass Director resource name as an option to the Console.
556 - Add a "batch" mode to the Console (no unsolicited queries, ...).
557 - Add a .list all files in the restore tree (probably also a list all files)
558 Do both a long and short form.
559 - Allow browsing the catalog to see all versions of a file (with
560 stat data on each file).
561 - Restore attributes of directory if replace=never set but directory
563 - Use SHA1 on authentication if possible.
564 - See comtest-xxx.zip for Windows code to talk to USB.
565 - Add John's appended files:
566 Appended = { /files/server/logs/http/*log }
567 and such files would be treated as follows.On a FULL backup, they would
568 be backed up like any other file.On an INCREMENTAL backup, where a
569 previous INCREMENTAL or FULL was already in thecatalogue and the length
570 of the file wasgreater than the length of the last backup, only thedata
571 added since the last backup will be dumped.On an INCREMENTAL backup, if
572 the length of the file is less than thelength of the file with the same
573 name last backed up, the completefile is dumped.On Windows systems, with
574 creation date of files, we can be evensmarter about this and not count
575 entirely upon the length.On a restore, the full and all incrementals
576 since it will beapplied in sequence to restore the file.
577 - Check new HAVE_WIN32 open bits.
578 - Check if the tape has moved before writing.
579 - Handling removable disks -- see below:
580 - Keep track of tape use time, and report when cleaning is necessary.
581 - Add FromClient and ToClient keywords on restore command (or
582 BackupClient RestoreClient).
583 - Implement a JobSet, which groups any number of jobs. If the
584 JobSet is started, all the jobs are started together.
585 Allow Pool, Level, and Schedule overrides.
586 - Enhance cancel to timeout BSOCK packets after a specific delay.
587 - Do scheduling by UTC using gmtime_r() in run_conf, scheduler, and
588 ua_status.!!! Thanks to Alan Brown for this tip.
589 - Look at updating Volume Jobs so that Max Volume Jobs = 1 will work
590 correctly for multiple simultaneous jobs.
591 - Correct code so that FileSet MD5 is calculated for < and | filename
593 - Implement the Media record flag that indicates that the Volume does disk
595 - Implement VolAddr, which is used when Volume is addressed like a disk,
596 and form it from VolFile and VolBlock.
597 - Make multiple restore jobs for multiple media types specifying
598 the proper storage type.
599 - Fix fast block rejection (stored/read_record.c:118). It passes a null
600 pointer (rec) to try_repositioning().
601 - Look at extracting Win data from BackupRead.
602 - Implement RestoreJobRetention? Maybe better "JobRetention" in a Job,
603 which would take precidence over the Catalog "JobRetention".
604 - Implement Label Format in Add and Label console commands.
605 - Possibly up network buffers to 65K. Put on variable.
606 - Put email tape request delays on one or more variables. User wants
607 to cancel the job after a certain time interval. Maximum Mount Wait?
608 - Job, Client, Device, Pool, or Volume?
609 Is it possible to make this a directive which is *optional* in multiple
610 resources, like Level? If so, I think I'd make it an optional directive
611 in Job, Client, and Pool, with precedence such that Job overrides Client
612 which in turn overrides Pool.
614 - New Storage specifications:
615 - Want to write to multiple storage devices simultaneously
616 - Want to write to multiple storage devices sequentially (in one job)
617 - Want to read/write simultaneously
618 - Key is MediaType -- it must match
620 Passed to SD as a sort of BSR record called Storage Specification
624 MediaType -> Next MediaType
626 Device -> Next Device
628 Allow multiple Storage specifications
636 Allow Multiple Pool specifications (note, Pool currently
638 Allow Multiple MediaType specifications in Dir conf
639 Allow Multiple Device specifications in Dir conf
640 Perhaps keep this in a single SSR
641 Tie a Volume to a specific device by using a MediaType that
642 is contained in only one device.
643 In SD allow Device to have Multiple MediaTypes
645 - Ideas from Jerry Scharf:
646 First let's point out some big pluses that bacula has for this
648 more importantly it's active. Thank you so much for that
649 even more important, it's not flaky
650 it has an open access catalog, opening many possibilities
651 it's pushing toward heterogeneous systems capability
653 Macintosh file client
654 macs are an interesting niche, but I fear a server is a rathole
655 working bare iron recovery for windows
656 the option for inc/diff backups not reset on fileset revision
657 a) use both change and inode update time against base time
658 b) do the full catalog check (expensive but accurate)
659 sizing guide (how much system is needed to back up N systems/files)
660 consultants on using bacula in building a disaster recovery system
662 or how to get at fancy things that one could do with bacula
663 logwatch code for bacula logs (or similar)
664 linux distro inclusion of bacula (brings good and bad, but necessary)
665 win2k/XP server capability (icky but you asked)
666 support for Oracle database ??
668 - Look at adding SQL server and Exchange support for Windows.
669 - Make dev->file and dev->block_num signed integers so that -1 can
670 be an invalid value which happens with BSR.
671 - Create VolAddr for disk files in place of VolFile and VolBlock. This
672 is needed to properly specify ranges.
673 - Add progress of files/bytes to SD and FD.
674 - Print warning message if FileId > 4 billion
675 - do a "messages" before the first prompt in Console
676 - Client does not show busy during Estimate command.
677 - Implement Console mtx commands.
678 - Implement a Mount Command and an Unmount Command where
679 the users could specify a system command to be performed
680 to do the mount, after which Bacula could attempt to
681 read the device. This is for Removeable media such as a CDROM.
682 - Most likely, this mount command would be invoked explicitly
683 by the user using the current Console "mount" and "unmount"
684 commands -- the Storage Daemon would do the right thing
685 depending on the exact nature of the device.
686 - As with tape drives, when Bacula wanted a new removable
687 disk mounted, it would unmount the old one, and send a message
688 to the user, who would then use "mount" as described above
689 once he had actually inserted the disk.
690 - Implement dump/print label to UA
691 - Spool to disk only when the tape is full, then when a tape is hung move
693 - bextract is sending everything to the log file ****FIXME****
694 - Allow multiple Storage specifications (or multiple names on
695 a single Storage specification) in the Job record. Thus a job
696 can be backed up to a number of storage devices.
697 - Implement some way for the File daemon to contact the Director
698 to start a job or pass its DHCP obtained IP number.
699 - Implement a query tape prompt/replace feature for a console
700 - Copy console @ code to gnome2-console
701 - Make AES the only encryption algorithm see
702 http://csrc.nist.gov/CryptoToolkit/aes/). It's
703 an officially adopted standard, has survived peer
704 review, and provides keys up to 256 bits.
705 - Take a careful look at SetACL http://setacl.sourceforge.net
706 - Make tree walk routines like cd, ls, ... more user friendly
707 by handling spaces better.
708 - Make sure that Bacula rechecks the tape after the 20 min wait.
709 - Set IO_NOWAIT on Bacula TCP/IP packets.
710 - Try doing a raw partition backup and restore by mounting a
713 Yes, it would allow to highly automatic the request for new tapes. If a
714 tape is empty, bacula reads the barcodes (native or simulated), and if
715 an unused tape is found, it runs the label command with all the
716 necessary parameters.
718 By the way can bacula automatically "move" an empty/purged volume say
719 in the "short" pool to the "long" pool if this pool runs out of volume
721 - What to do about "list files job=xxx".
722 - Get and test MySQL 4.0
723 - Look at how fuser works and /proc/PID/fd that is how Nic found the
724 file descriptor leak in Bacula.
725 - Implement WrapCounters in Counters.
726 - Add heartbeat from FD to SD if hb interval expires.
727 - Can we dynamically change FileSets?
728 - If pool specified to label command and Label Format is specified,
729 automatically generate the Volume name.
730 - Why can't SQL do the filename sort for restore?
731 - Add ExhautiveRestoreSearch
732 - Look at the possibility of loading only the necessary
733 data into the restore tree (i.e. do it one directory at a
734 time as the user walks through the tree).
735 - Possibly use the hash code if the user selects all for a restore command.
736 - Fix "restore all" to bypass building the tree.
737 - Prohibit backing up archive device (findlib/find_one.c:128)
738 - Implement Release Device in the Job resource to unmount a drive.
739 - Implement Acquire Device in the Job resource to mount a drive,
740 be sure this works with admin jobs so that the user can get
741 prompted to insert the correct tape. Possibly some way to say to
742 run the job but don't save the files.
743 - Make things like list where a file is saved case independent for
746 - Use autochanger to handle multiple devices.
747 - On Windows with very long path names, it may be impossible to create
748 a file (and thus restore it) because the total length is too long.
749 We must cd into the directory then create the file without the
751 - Implement a Recycle command
752 - Test a second language e.g. french.
753 - Start working on Base jobs.
754 - Implement UnsavedFiles DB record.
755 - From Phil Stracchino:
756 It would probably be a per-client option, and would be called
757 something like, say, "Automatically purge obsoleted jobs". What it
758 would do is, when you successfully complete a Differential backup of a
759 client, it would automatically purge all Incremental backups for that
760 client that are rendered redundant by that Differential. Likewise,
761 when a Full backup on a client completed, it would automatically purge
762 all Differential and Incremental jobs obsoleted by that Full backup.
763 This would let people minimize the number of tapes they're keeping on
764 hand without having to master the art of retention times.
765 - When doing a Backup send all attributes back to the Director, who
766 would then figure out what files have been deleted.
767 - Currently in mount.c:236 the SD simply creates a Volume. It should have
768 explicit permission to do so. It should also mark the tape in error
769 if there is an error.
770 - Cancel waiting for Client connect in SD if FD goes away.
772 - Implement timeout in response() when it should come quickly.
773 - Implement a Slot priority (loaded/not loaded).
774 - Implement "vacation" Incremental only saves.
775 - Implement create "FileSet"?
776 - Add prefixlinks to where or not where absolute links to FD.
777 - Issue message to mount a new tape before the rewind.
778 - Simplified client job initiation for portables.
779 - If SD cannot open a drive, make it periodically retry.
780 - Add more of the config info to the tape label.
782 - If tape is marked read-only, then try opening it read-only rather than
783 failing, and remember that it cannot be written.
784 - Refine SD waiting output:
785 Device is being positioned
786 > Device is being positioned for append
787 > Device is being positioned to file x
789 - Figure out some way to estimate output size and to avoid splitting
790 a backup across two Volumes -- this could be useful for writing CDROMs
791 where you really prefer not to have it split -- not serious.
792 - Have SD compute MD5 or SHA1 and compare to what FD computes.
793 - Make VolumeToCatalog calculate an MD5 or SHA1 from the
794 actual data on the Volume and compare it.
795 - Implement Bacula plugins -- design API
796 - Make bcopy read through bad tape records.
797 - Program files (i.e. execute a program to read/write files).
798 Pass read date of last backup, size of file last time.
799 - Add Signature type to File DB record.
800 - CD into subdirectory when open()ing files for backup to
801 speed up things. Test with testfind().
802 - Priority job to go to top of list.
803 - Why are save/restore of device different sizes (sparse?) Yup! Fix it.
804 - Implement some way for the Console to dynamically create a job.
805 - Solaris -I on tar for include list
806 - Need a verbose mode in restore, perhaps to bsr.
807 - bscan without -v is too quiet -- perhaps show jobs.
808 - Add code to reject whole blocks if not wanted on restore.
809 - Check if we can increase Bacula FD priorty in Win2000
810 - Make sure the MaxVolFiles is fully implemented in SD
811 - Check if both CatalogFiles and UseCatalog are set to SD.
812 - Possibly add email to Watchdog if drive is unmounted too
813 long and a job is waiting on the drive.
814 - After unmount, if restore job started, ask to mount.
815 - Add UA rc and history files.
816 - put termcap (used by console) in ./configure and
817 allow -with-termcap-dir.
818 - Fix Autoprune for Volumes to respect need for full save.
819 - Compare tape to Client files (attributes, or attributes and data)
820 - Make all database Ids 64 bit.
821 - Allow console commands to detach or run in background.
822 - Fix status delay on storage daemon during rewind.
823 - Add SD message variables to control operator wait time
824 - Maximum Operator Wait
825 - Minimum Message Interval
826 - Maximum Message Interval
827 - Send Operator message when cannot read tape label.
828 - Verify level=Volume (scan only), level=Data (compare of data to file).
829 Verify level=Catalog, level=InitCatalog
831 - Add keyword search to show command in Console.
832 - Events : tape has more than xxx bytes.
833 - Complete code in Bacula Resources -- this will permit
834 reading a new config file at any time.
835 - Handle ctl-c in Console
836 - Implement script driven addition of File daemon to config files.
837 - Think about how to make Bacula work better with File (non-tape) archives.
838 - Write Unix emulator for Windows.
839 - Put memory utilization in Status output of each daemon
840 if full status requested or if some level of debug on.
841 - Make database type selectable by .conf files i.e. at runtime
842 - Set flag for uname -a. Add to Volume label.
843 - Restore files modified after date
844 - SET LD_RUN_PATH=$HOME/mysql/lib/mysql
845 - Remove duplicate fields from jcr (e.g. jcr.level and jcr.jr.Level, ...).
846 - Timout a job or terminate if link goes down, or reopen link and query.
847 - Concept of precious tapes (cannot be reused).
848 - Make bcopy copy with a single tape drive.
849 - Permit changing ownership during restore.
852 > My suggestion: Add a feature on the systray menu-icon menu to request
853 > an immediate backup now. This would be useful for laptop users who may
854 > not be on the network when the regular scheduled backup is run.
856 > My wife's suggestion: Add a setting to the win32 client to allow it to
857 > shut down the machine after backup is complete (after, of course,
858 > displaying a "System will shut down in one minute, click here to cancel"
859 > warning dialog). This would be useful for sites that want user
860 > woorkstations to be shut down overnight to save power.
863 - Autolabel should be specified by DIR instead of SD.
866 - AutoScan (check checksum of tape)
867 - Format command = "format /dev/nst0"
871 - Seek resolution (usually corresponds to buffer size)
872 - EODErrorCode=ENOSPC or code
873 - Partial Read error code
874 - Partial write error code
875 - Nonformatted read error
876 - Nonformatted write error
877 - WriteProtected error
881 - IgnoreCloseErrors=yes
891 - FD sends unsaved file list to Director at end of job (see
893 - File daemon should build list of files skipped, and then
894 at end of save retry and report any errors.
895 - Write a Storage daemon that uses pipes and
896 standard Unix programs to write to the tape.
898 - Need something that monitors the JCR queue and
899 times out jobs by asking the deamons where they are.
900 - Enhance Jmsg code to permit buffering and saving to disk.
901 - device driver = "xxxx" for drives.
903 - Ensure that /dev/null works
904 - Need report class for messages. Perhaps
905 report resource where report=group of messages
906 - enhance scan_attrib and rename scan_jobtype, and
907 fill in code for "since" option
908 - Director needs a time after which the report status is sent
909 anyway -- or better yet, a retry time for the job.
910 - Don't reschedule a job if previous incarnation is still running.
911 - Some way to automatically backup everything is needed????
912 - Need a structure for pending actions:
914 - termination status (part of buffered msgs?)
916 Read, Write, Clean, Delete
917 - Login to Bacula; Bacula users with different permissions:
918 owner, group, user, quotas
919 - Store info on each file system type (probably in the job header on tape.
920 This could be the output of df; or perhaps some sort of /etc/mtab record.
922 ========= ideas ===============
923 From: "Jerry K. Schieffer" <jerry@skylinetechnology.com>
924 To: <kern@sibbald.com>
925 Subject: RE: [Bacula-users] future large programming jobs
926 Date: Thu, 26 Feb 2004 11:34:54 -0600
928 I noticed the subject thread and thought I would offer the following
929 merely as sources of ideas, i.e. something to think about, not even as
930 strong as a request. In my former life (before retiring) I often
931 dealt with backups and storage management issues/products as a
932 developer and as a consultant. I am currently migrating my personal
933 network from amanda to bacula specifically because of the ability to
934 cross media boundaries during storing backups.
935 Are you familiar with the commercial product called ADSM (I think IBM
936 now sells it under the Tivoli label)? It has a couple of interesting
937 ideas that may apply to the following topics.
939 1. Migration: Consider that when you need to restore a system, there
940 may be pressure to hurry. If all the information for a single client
941 can eventually end up on the same media (and in chronological order),
942 the restore is facillitated by not having to search past information
943 from other clients. ADSM has the concept of "client affinity" that
944 may be associated with it's storage pools. It seems to me that this
945 concept (as an optional feature) might fit in your architecture for
948 ADSM also has the concept of defining one or more storage pools as
949 "copy pools" (almost mirrors, but only in the sense of contents).
950 These pools provide the ability to have duplicte data stored both
951 onsite and offsite. The copy process can be scheduled to be handled
952 by their storage manager during periods when there is no backup
953 activity. Again, the migration process might be a place to consider
954 implementing something like this.
957 > It strikes me that it would be very nice to be able to do things
959 > have the Job(s) backing up the machines run, and once they have all
960 > completed, start a migration job to copy the data from disks Volumes
962 > a tape library and then to offsite storage. Maybe this can already
964 > done with some careful scheduling and Job prioritzation; the events
965 > mechanism described below would probably make it very easy.
967 This is the goal. In the first step (before events), you simply
969 the Migration to tape later.
971 2. Base jobs: In ADSM, each copy of each stored file is tracked in
972 the database. Once a file (unique by path and metadata such as dates,
973 size, ownership, etc.) is in a copy pool, no more copies are made. In
974 other words, when you start ADSM, it begins like your concept of a
975 base job. After that it is in the "incremental" mode. You can
976 configure the number of "generations" of files to be retained, plus a
977 retention date after which even old generations are purged. The
978 database tracks the contents of media and projects the percentage of
979 each volume that is valid. When the valid content of a volume drops
980 below a configured percentage, the valid data are migrated to another
981 volume and the old volume is marked as empty. Note, this requires
982 ADSM to have an idea of the contents of a client, i.e. marking the
983 database when an existing file was deleted, but this would solve your
984 issue of restoring a client without restoring deleted files.
986 This is pretty far from what bacula now does, but if you are going to
987 rip things up for Base jobs,.....
988 Also, the benefits of this are huge for very large shops, especially
989 with media robots, but are a pain for shops with manual media
993 > Base jobs sound pretty useful, but I'm not dying for them.
995 Nobody is dying for them, but when you see what it does, you will die
998 3. Restoring deleted files: Since I think my comments in (2) above
999 have low probability of implementation, I'll also suggest that you
1000 could approach the issue of deleted files by a mechanism of having the
1001 fd report to the dir, a list of all files on the client for every
1002 backup job. The dir could note in the database entry for each file
1003 the date that the file was seen. Then if a restore as of date X takes
1004 place, only files that exist from before X until after X would be
1005 restored. Probably the major cost here is the extra date container in
1006 each row of the files table.
1008 Thanks for "listening". I hope some of this helps. If you want to
1009 contact me, please send me an email - I read some but not all of the
1010 mailing list traffic and might miss a reply there.
1012 Please accept my compliments for bacula. It is doing a great job for
1013 me!! I sympathize with you in the need to wrestle with excelence in
1014 execution vs. excelence in feature inclusion.
1019 ==============================
1022 - Design at hierarchial storage for Bacula. Migration and Clone.
1023 - Implement FSM (File System Modules).
1024 - Audit M_ error codes to ensure they are correct and consistent.
1025 - Add variable break characters to lex analyzer.
1026 Either a bit mask or a string of chars so that
1027 the caller can change the break characters.
1028 - Make a single T_BREAK to replace T_COMMA, etc.
1029 - Ensure that File daemon and Storage daemon can
1030 continue a save if the Director goes down (this
1031 is NOT currently the case). Must detect socket error,
1032 buffer messages for later.
1033 - Enhance time/duration input to allow multiple qualifiers e.g. 3d2h
1034 - Add ability to backup to two Storage devices (two SD sessions) at
1035 the same time -- e.g. onsite, offsite.
1036 - Add the ability to consolidate old backup sets (basically do a restore
1037 to tape and appropriately update the catalog). Compress Volume sets.
1038 Might need to spool via file is only one drive is available.
1039 - Compress or consolidate Volumes of old possibly deleted files. Perhaps
1040 someway to do so with every volume that has less than x% valid
1044 Migration: Move a backup from one Volume to another
1045 Clone: Copy a backup -- two Volumes
1047 Bacula Migration is based on Jobs (apparently Networker is file by file).
1049 Migration triggered by:
1053 Highwater mark (keep total size)
1058 ======================================================
1060 It is somewhat like a Full save becomes an incremental since
1061 the Base job (or jobs) plus other non-base files.
1063 - A Base backup is same as Full backup, just different type.
1064 - New BaseFiles table that contains:
1066 BaseJobId - Base JobId referenced for this FileId (needed ???)
1067 JobId - JobId currently running
1068 FileId - File not backed up, exists in Base Job
1069 FileIndex - FileIndex from Base Job.
1070 i.e. for each base file that exists but is not saved because
1071 it has not changed, the File daemon sends the JobId, BaseId,
1072 FileId, FileIndex back to the Director who creates the DB entry.
1073 - To initiate a Base save, the Director sends the FD
1074 the FileId, and full filename for each file in the Base.
1075 - When the FD finds a Base file, he requests the Director to
1076 send him the full File entry (stat packet plus MD5), or
1077 conversely, the FD sends it to the Director and the Director
1078 says yes or no. This can be quite rapid if the FileId is kept
1079 by the FD for each Base Filename.
1080 - It is probably better to have the comparison done by the FD
1081 despite the fact that the File entry must be sent across the
1083 - An alternative would be to send the FD the whole File entry
1084 from the start. The disadvantage is that it requires a lot of
1085 space. The advantage is that it requires less communications
1087 - The Job record must be updated to indicate that one or more
1089 - At end of Job, FD returns:
1090 1. Count of base files/bytes not written to tape (i.e. matches)
1091 2. Count of base file that were saved i.e. had changed.
1092 - No tape record would be written for a Base file that matches, in the
1093 same way that no tape record is written for Incremental jobs where
1094 the file is not saved because it is unchanged.
1095 - On a restore, all the Base file records must explicitly be
1096 found from the BaseFile tape. I.e. for each Full save that is marked
1097 to have one or more Base Jobs, search the BaseFile for all occurrences
1099 - An optimization might be to make the BaseFile have:
1105 This would avoid the need to explicitly fetch each File record for
1106 the Base job. The Base Job record will be fetched to get the
1107 VolSessionId and VolSessionTime.
1108 =========================================================
1111 ==========================================================
1113 For each Incremental job that is run, there may be files that
1114 were found but not saved because they were locked (this applies
1115 only to Windows). Such a system could send back to the Director
1116 a list of Unsaved files.
1118 - New UnSavedFiles table that contains:
1122 - Then in the next Incremental job, the list of Unsaved Files will be
1123 feed to the FD, who will ensure that they are explicitly chosen even
1124 if standard date/time check would not have selected them.
1125 =============================================================
1129 Multiple drive autochanger data: see Alan Brown
1130 mtx -f xxx unloadStorage Element 1 is Already Full(drive 0 was empty)
1131 Unloading Data Transfer Element into Storage Element 1...source Element
1132 Address 480 is Empty
1134 (drive 0 was empty and so was slot 1)
1135 > mtx -f xxx load 15 0
1136 no response, just returns to the command prompt when complete.
1137 > mtx -f xxx status Storage Changer /dev/changer:2 Drives, 60 Slots ( 2 Import/Export )
1138 Data Transfer Element 0:Full (Storage Element 15 Loaded):VolumeTag = HX001
1139 Data Transfer Element 1:Empty
1140 Storage Element 1:Empty
1141 Storage Element 2:Full :VolumeTag=HX002
1142 Storage Element 3:Full :VolumeTag=HX003
1143 Storage Element 4:Full :VolumeTag=HX004
1144 Storage Element 5:Full :VolumeTag=HX005
1145 Storage Element 6:Full :VolumeTag=HX006
1146 Storage Element 7:Full :VolumeTag=HX007
1147 Storage Element 8:Full :VolumeTag=HX008
1148 Storage Element 9:Full :VolumeTag=HX009
1149 Storage Element 10:Full :VolumeTag=HX010
1150 Storage Element 11:Empty
1151 Storage Element 12:Empty
1152 Storage Element 13:Empty
1153 Storage Element 14:Empty
1154 Storage Element 15:Empty
1155 Storage Element 16:Empty....
1156 Storage Element 28:Empty
1157 Storage Element 29:Full :VolumeTag=CLNU01L1
1158 Storage Element 30:Empty....
1159 Storage Element 57:Empty
1160 Storage Element 58:Full :VolumeTag=NEX261L2
1161 Storage Element 59 IMPORT/EXPORT:Empty
1162 Storage Element 60 IMPORT/EXPORT:Empty
1164 Unloading Data Transfer Element into Storage Element 15...done
1166 (just to verify it remembers where it came from, however it can be
1167 overrriden with mtx unload {slotnumber} to go to any storage slot.)
1169 There needs to be a table of drive # to devices somewhere - If there are
1170 multiple changers or drives there may not be a 1:1 correspondance between
1171 changer drive number and system device name - and depending on the way the
1172 drives are hooked up to scsi busses, they may not be linearly numbered
1173 from an offset point either.something like
1175 Autochanger drives = 2
1176 Autochanger drive 0 = /dev/nst1
1177 Autochanger drive 1 = /dev/nst2
1178 IMHO, it would be _safest_ to use explicit mtx unload commands at all
1179 times, not just for multidrive changers. For a 1 drive changer, that's
1185 MTX's manpage (1.2.15):
1186 unload [<slotnum>] [ <drivenum> ]
1187 Unloads media from drive <drivenum> into slot
1188 <slotnum>. If <drivenum> is omitted, defaults to
1189 drive 0 (as do all commands). If <slotnum> is
1190 omitted, defaults to the slot that the drive was
1191 loaded from. Note that there's currently no way
1192 to say 'unload drive 1's media to the slot it
1193 came from', other than to explicitly use that
1194 slot number as the destination.AB
1200 undef# camcontrol devlist
1201 <WANGTEK 51000 SCSI M74H 12B3> at scbus0 target 2 lun 0 (pass0,sa0)
1202 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 0 (pass1,sa1)
1203 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 1 (pass2)
1205 tapeinfo -f /dev/sg0 with a bad tape in drive 1:
1206 [kern@rufus mtx-1.2.17kes]$ ./tapeinfo -f /dev/sg0
1207 Product Type: Tape Drive
1209 Product ID: 'C5713A '
1211 Attached Changer: No
1212 TapeAlert[3]: Hard Error: Uncorrectable read/write error.
1213 TapeAlert[20]: Clean Now: The tape drive neads cleaning NOW.
1220 Medium Type: Not Loaded
1223 DataCompEnabled: yes
1224 DataCompCapable: yes
1225 DataDeCompEnabled: yes
1232 Handling removable disks
1234 From: Karl Cunningham <karlc@keckec.com>
1236 My backups are only to hard disk these days, in removable bays. This is my
1237 idea of how a backup to hard disk would work more smoothly. Some of these
1238 things Bacula does already, but I mention them for completeness. If others
1239 have better ways to do this, I'd like to hear about it.
1241 1. Accommodate several disks, rotated similar to how tapes are. Identified
1242 by partition volume ID or perhaps by the name of a subdirectory.
1243 2. Abort & notify the admin if the wrong disk is in the bay.
1244 3. Write backups to different subdirectories for each machine to be backed
1246 4. Volumes (files) get created as needed in the proper subdirectory, one
1248 5. When a disk is recycled, remove or zero all old backup files. This is
1249 important as the disk being recycled may be close to full. This may be
1250 better done manually since the backup files for many machines may be
1251 scattered in many subdirectories.
1255 === Done -- see kernsdone