4 Documentation to do: (a little bit at a time)
5 - Document running a test version.
6 - Document query file format.
7 - Document static linking
8 - Document how to automatically backup all local partitions
9 - Document problems with Verify and pruning.
10 - Document how to use multiple databases.
13 Testing to do: (painful)
14 - that console command line options work
15 - blocksize recognition code.
16 - multiple simultaneous Volumes
19 - Implement a Mount Command and an Unmount Command where
20 the users could specify a system command to be performed
21 to do the mount, after which Bacula could attempt to
22 read the device. This is for Removeable media such as a
24 - Most likely, this mount command would be invoked explicitly
25 by the user using the current Console "mount" and "unmount"
26 commands -- the Storage Daemon would do the right thing
27 depending on the exact nature of the device.
28 - As with tape drives, when Bacula wanted a new removable
29 disk mounted, it would unmount the old one, and send a message
30 to the user, who would then use "mount" as described above
31 once he had actually inserted the disk.
33 - Implement TCP/IP connection for MySQL
34 - Make some way so that if a machine is skipped because it is not up
35 that Bacula will continue retrying for a specified period of time --
37 - Figure out some way to estimate output size and to avoid splitting
38 a backup across two Volumes -- this could be useful for writing CDROMs
39 where you really prefer not to have it split -- not serious.
40 - Add chflags() code for FreeBSD file flags
41 - Add RunBeforeJob and RunAfterJob to the Client program.
42 - Have SD compute MD5 or SHA1 and compare to what FD computes.
43 - Make VolumeToCatalog calculate an MD5 or SHA1 from the
44 actual data on the Volume and compare it.
45 - Implement FileOptions (see end of this document)
46 - Implement Bacula plugins -- design API
47 - Make bcopy read through bad tape records.
48 - Fix read_record to handle multiple sessions.
49 - Program files (i.e. execute a program to read/write files).
50 Pass read date of last backup, size of file last time.
51 - Add Signature type to File DB record.
52 - Make Restore report an error if FD or SD term codes are not OK.
53 - CD into subdirectory when open()ing files for backup to
54 speed up things. Test with testfind().
55 - Add prefixlinks to where or not where absolute links to FD.
56 - Look at handling <> in smtp doesn't work with exim.
57 - Priority job to go to top of list.
58 - Implement Bar code handling
59 - Why is catreq.c:111 Find vol called twice for a job?
60 - Find out why Full saves run slower and slower (hashing?)
61 - Why are save/restore of device different sizes (sparse?) Yup! Fix it.
62 - Implement some way for the Console to dynamically create a job.
63 - Restore to a particular time -- e.g. before date, after date.
64 - Implement disk spooling
65 - Implement finer multiprocessing options.
66 - Solaris -I on tar for include list
67 - Enable avoid backing up archive device (findlib/find_one.c:128)
68 - Need a verbose mode in restore, perhaps to bsr.
69 - bscan without -v is too quiet -- perhaps show jobs.
70 - Add code to reject whole blocks if not wanted on restore.
71 - Start working on Base jobs.
72 - Make sure the MaxVolFiles is fully implemented in SD
73 - Check if both CatalogFiles and UseCatalog are set to SD.
74 - Check if we can increase Bacula FD priorty in Win2000
75 - Need return status on read_cb() from read_records(). Need multiple
76 records -- one per Job, maybe a JCR or some other structure with
78 - Figure out how to do a bare metal Windows restore
79 - Put system type returned by FD into catalog.
80 - Need to specify MaximumConcurrentJobs in the Job resource.
81 - Possibly add email to Watchdog if drive is unmounted too
82 long and a job is waiting on the drive.
83 - Strip trailing slashes from Include directory names in the FD.
84 - Use read_record.c in SD code.
85 - Why don't we get an error message from Win32 FD when bootstrap
86 file cannot be created for restore command?
87 - When Marking a file in Restore that is a hard link, also
88 mark the link so that the data will be reloaded.
89 - Restore program that errors in SD due to no tape reports
90 OK incorrectly in output.
91 - After unmount, if restore job started, ask to mount.
92 - Convert all %x substitution variables, which are hard to remember
93 and read to %(variable-name). Idea from TMDA.
94 - Add JobLevel in FD status (but make sure it is defined).
95 - Make Pool resource handle Counter resources.
96 - Remove NextId for SQLite. Optimize.
97 - Strip trailing / from Include
98 - Move all SQL statements into a single location.
99 - Add UA rc and history files.
100 - put termcap (used by console) in ./configure and
101 allow -with-termcap-dir.
102 - Enhance time and size scanning routines.
103 - Fix Autoprune for Volumes to respect need for full save.
104 - Fix Win32 config file definition name on /install
105 - No READLINE_SRC if found in alternate directory.
106 - Add Client FS/OS id (Linux, Win95/98, ...).
107 - Test a second language e.g. french.
108 - Compare tape to Client files (attributes, or attributes and data)
109 - Make all database Ids 64 bit.
110 - Write an applet for Linux.
111 - Add estimate to Console commands
112 - Find solution to blank filename (i.e. path only) problem.
113 - Implement new daemon communications protocol.
114 - Remove PoolId from Job table, it exists in Media.
115 - Allow console commands to detach or run in background.
116 - Fix status delay on storage daemon during rewind.
117 - Add SD message variables to control operator wait time
118 - Maximum Operator Wait
119 - Minimum Message Interval
120 - Maximum Message Interval
121 - Send Operator message when cannot read tape label.
122 - Verify level=Volume (scan only), level=Data (compare of data to file).
123 Verify level=Catalog, level=InitCatalog
125 - Add keyword search to show command in Console.
126 - Fix Win2000 error with no messages during startup.
127 - Events : tape has more than xxx bytes.
128 - Restrict characters permitted in a Resource name.
129 - Complete code in Bacula Resources -- this will permit
130 reading a new config file at any time.
131 - Handle ctl-c in Console
132 - Implement LabelTemplate (at least first cut).
133 - Implement script driven addition of File daemon to config files.
134 - Think about how to make Bacula work better with File (non-tape) archives.
135 - Write Unix emulator for Windows.
137 - Implement new serialize subroutines
138 send(socket, "string", &Vol, "uint32", &i, NULL)
139 - Audit all UA commands to ensure that we always prompt where possible.
140 - If ./btape is called without /dev, assume argument is a Storage resource name.
141 - Put memory utilization in Status output of each daemon
142 if full status requested or if some level of debug on.
143 - Make database type selectable by .conf files i.e. at runtime
144 - gethostbyname failure in bnet_connect() continues
145 generating errors -- should stop.
146 - Set flag for uname -a. Add to Volume label.
147 - Implement throttled work queue.
148 - Check for EOT at ENOSPC or EIO or ENXIO (unix Pc)
149 - Allow multiple Storage specifications (or multiple names on
150 a single Storage specification) in the Job record. Thus a job
151 can be backed up to a number of storage devices.
152 - Implement dump label to UA
153 - Concept of VolumeSet during restore which is a list
154 of Volume names needed.
155 - Restore files modified after date
156 - Restore file modified before date
157 - Emergency restore info:
159 - Backup working directory
161 - Restore -- do nothing but show what would happen
162 - SET LD_RUN_PATH=$HOME/mysql/lib/mysql
163 - Implement Restore FileSet=
164 - Create a protocol.h and protocol.c where all protocol messages
166 - If SD cannot open a drive, make it periodically retry.
167 - Remove duplicate fields from jcr (e.g. jcr.level and jcr.jr.Level, ...).
168 - Timout a job or terminate if link goes down, or reopen link and query.
169 - Find general solution for sscanf size problems (as well
170 as sprintf. Do at run time?
171 - Concept of precious tapes (cannot be reused).
172 - Make bcopy copy with a single tape drive.
173 - Permit changing ownership during restore.
175 - Autolabel should be specified by DIR instead of SD.
176 - Find out how to get the system tape block limits, e.g.:
177 Apr 22 21:22:10 polymatou kernel: st1: Block limits 1 - 245760 bytes.
178 Apr 22 21:22:10 polymatou kernel: st0: Block limits 2 - 16777214 bytes.
181 - AutoScan (check checksum of tape)
182 - Format command = "format /dev/nst0"
186 - Seek resolution (usually corresponds to buffer size)
187 - EODErrorCode=ENOSPC or code
188 - Partial Read error code
189 - Partial write error code
190 - Nonformatted read error
191 - Nonformatted write error
192 - WriteProtected error
196 - IgnoreCloseErrors=yes
206 - FD sends unsaved file list to Director at end of job (see
208 - File daemon should build list of files skipped, and then
209 at end of save retry and report any errors.
210 - Write a Storage daemon that uses pipes and
211 standard Unix programs to write to the tape.
213 - Need something that monitors the JCR queue and
214 times out jobs by asking the deamons where they are.
215 - Enhance Jmsg code to permit buffering and saving to disk.
216 - device driver = "xxxx" for drives.
217 - restart: paranoid: read label fsf to
218 eom read append block, and go
219 super-paranoid: read label, read all files
220 in between, read append block, and go
221 verify: backspace, read append block, and go
222 permissive: same as above but frees drive
223 if tape is not valid.
225 - Ensure that /dev/null works
226 - Need report class for messages. Perhaps
227 report resource where report=group of messages
228 - enhance scan_attrib and rename scan_jobtype, and
229 fill in code for "since" option
230 - Director needs a time after which the report status is sent
231 anyway -- or better yet, a retry time for the job.
232 Don't reschedule a job if previous incarnation is still running.
233 - Some way to automatically backup everything is needed????
234 - Need a structure for pending actions:
236 - termination status (part of buffered msgs?)
237 - Concept of grouping Storage devices and job can use
238 any of a number of devices
240 Read, Write, Clean, Delete
241 - Login to Bacula; Bacula users with different permissions:
242 owner, group, user, quotas
243 - Store info on each file system type (probably in the job header on tape.
244 This could be the output of df; or perhaps some sort of /etc/mtab record.
247 - Design at hierarchial storage for Bacula.
248 - Implement FSM (File System Modules).
249 - Identify unchanged or "system" files and save them to a
250 special tape thus removing them from the standard
251 backup FileSet -- BASE backup.
252 - Turn virutally all sprintfs into snprintfs.
253 - Heartbeat between daemons.
254 - Audit M_ error codes to ensure they are correct and consistent.
255 - Add variable break characters to lex analyzer.
256 Either a bit mask or a string of chars so that
257 the caller can change the break characters.
258 - Make a single T_BREAK to replace T_COMMA, etc.
259 - Ensure that File daemon and Storage daemon can
260 continue a save if the Director goes down (this
261 is NOT currently the case). Must detect socket error,
262 buffer messages for later.
263 - Enhance time/duration input to allow multiple qualifiers e.g. 3d2h
264 - Add ability to backup to two Storage devices (two SD sessions) at
265 the same time -- e.g. onsite, offsite.
266 - Add the ability to consolidate old backup sets (basically do a restore
267 to tape and appropriately update the catalog). Compress Volume sets.
268 Might need to spool via file is only one drive is available.
269 - Compress or consolidate Volumes of old possibly deleted files. Perhaps
270 someway to do so with every volume that has less than x% valid
275 Bacula Projects Roadmap
277 last update 5 January 2003
279 Item 1: Multiple simultaneous Jobs. (done)
280 Done -- Restore part needs better implementation to work correctly
282 What: Permit multiple simultaneous jobs in Bacula.
284 Why: An enterprise level solution needs to go fast without the
285 need for the system administrator to carefully tweak
286 timing. Based on the benchmarks, during a full
287 backup, NetWorker typically hit 10 times the bandwidth to
288 the tape compared to Bacula--largely. This is probably due to
289 running parallel jobs and multi-threaded filling of buffers
290 and writing them to tape. This should also make things work
291 better when you have a mix of fast and slow machines backing
294 Notes: Bacula was designed to run multiple simultaneous jobs. Thus
295 implementing this is a matter of some small cleanups and
299 Item 2: Make the Storage daemon use intermediate file storage to buffer data.
300 Deferred -- not necessary yet.
302 What: If data is coming into the SD too fast, buffer it to
303 disk if the user has configured this option.
305 Why: This would be nice, especially if it more or less falls out
306 when implementing (1) above. If not, it probably should not
307 be given a high priority because fundamentally the backup time
308 is limited by the tape bandwidth. Even though you may finish a
309 client job quicker by spilling to disk, you still have to
310 eventually get it onto tape. If intermediate disk buffering
311 allows us to improve write bandwidth to tape, it may make
314 Notes: Whether or not this is implemented will depend upon performance
315 testing after item 1 is implemented.
318 Item 3: Write the bscan program -- also write a bcopy program.
321 What: Write a program that reads a Bacula tape and puts all the
322 appropriate data into the catalog. This allows recovery
323 from a tape that is no longer in the database, or it allows
324 re-creation of a database if lost.
326 Why: This is a fundamental robustness and disaster recovery tool
327 which will increase the comfort level of a sysadmin
328 considering adopting Bacula.
330 Notes: A skeleton of this program already exists, but much work
331 needs to be done. Implementing this will also make apparent
332 any deficiencies in the current Bacula tape format.
335 Item 4: Implement Base jobs.
337 What: A base job is sort of like a Full save except that you
338 will want the FileSet to contain only files that are unlikely
339 to change in the future (i.e. a snapshot of most of your
340 system after installing it). After the base job has been run,
341 when you are doing a Full save, you can specify to exclude
342 all files saved by the base job that have not been modified.
344 Why: This is something none of the competition does, as far as we know
345 (except BackupPC, which is a Perl program that saves to disk
346 only). It is big win for the user, it makes Bacula stand out
347 as offering a unique optimization that immediately saves time
350 Notes: Big savings in tape usage. Will require more resources because
351 the e. DIR must send FD a list of files/attribs, and the FD must
352 search the list and compare it for each file to be saved.
355 Item 5: Implement Label templates
357 What: This is a mechanism whereby Bacula can automatically create
358 a tape label for new tapes according to a detailed specification
359 provided by the user.
361 Why: It is a major convenience item for folks who use automated label
364 Notes: Bacula already has a working form of automatic tape label
365 creation, but it is very crude. The design for the complete
366 tape labeling project is already documented in the manual.
369 Item 6: Write a regression script.
372 What: This is an automatic script that runs and tests as many features
373 of Bacula as possible. The output is compared to previous
374 versions of Bacula and any differences are reported.
376 Why: This is an enormous help in preventing introduction of new
377 errors in parts of the program that already work correctly.
379 Notes: This probably should be ranked higher, it's something the typical
380 user doesn't see. Depending on how it's implemented, it may
381 make sense to defer it until the archival tape format and
382 user interface mature.
385 Item 7: GUI for interactive restore
386 Item 8: GUI for interactive backup
388 What: The current interactive restore is implemented with a tty
389 interface. It would be much nicer to be able to "see" the
390 list of files backed up in typical GUI tree format.
391 The same mechanism could also be used for creating
392 ad-hoc backup FileSets (item 8).
394 Why: Ease of use -- especially for the end user.
396 Notes: Rather than implementing in Gtk, we probably should go directly
397 for a Browser implementation, even if doing so meant the
398 capability wouldn't be available until much later. Not only
399 is there the question of Windows sites, most
400 Solaris/HP/IRIX, etc, shops can't currently run Gtk programs
401 without installing lots of stuff admins are very wary about.
402 Real sysadmins will always use the command line anyway, and
403 the user who's doing an interactive restore or backup of his
404 own files will in most cases be on a Windows machine running
408 Item 9: Add SSL to daemon communications.
410 What: This provides for secure communications between the daemons.
412 Why: This would allow doing backup across the Internet without
413 privacy concerns (or with much less concern).
415 Notes: The vast majority of near term potential users will be backing up
416 a single site over a LAN and, correctly or not, they probably
417 won't be concerned with security, at least not enough to go to
418 the trouble to set up keys, etc. to screw things down. We suspect
419 that many users genuinely interested in multi-site backup
420 already run some form of VPN software in their internetwork
421 connections, and are willing to delegate security to that layer.
424 Item 10: Define definitive tape format.
427 What: Define that definitive tape format that will not change
428 for the next millennium.
430 Why: Stability, security.
432 Notes: See notes for item 11 below.
435 Item 11: New daemon communication protocol.
437 What: The current daemon to daemon protocol is basically an ASCII
438 printf() and sending the buffer. On the receiving end, the
439 buffer is sscanf()ed to unpack it. The new scheme would
440 be a binary format that allows quick packing and unpacking
441 of any data type with named fields.
443 Why: Using binary packing would be faster. Named fields will permit
444 error checking to ensure that what is sent is what the
445 receiver really wants.
447 Notes: These are internal improvements in the interest of the
448 long-term stability and evolution of the program. On the one
449 hand, the sooner they're done, the less code we have to rip
450 up when the time comes to install them. On the other hand, they
451 don't bring an immediately perceptible benefit to potential
452 users. Item 10 and possibly item 11 should be deferred until Bacula
453 is well established with a growing user community more or
454 less happy with the feature set. At that time, it will make a
455 good "next generation" upgrade in the interest of data
461 ======================================================
463 It is somewhat like a Full save becomes an incremental since
464 the Base job (or jobs) plus other non-base files.
466 - New BaseFile table that contains:
467 JobId, BaseJobId, FileId (from Base).
468 i.e. for each base file that exists but is not saved because
469 it has not changed, the File daemon sends the JobId, BaseId,
470 and FileId back to the Director who creates the DB entry.
471 - To initiate a Base save, the Director sends the FD
472 the FileId, and full filename for each file in the Base.
473 - When the FD finds a Base file, he requests the Director to
474 send him the full File entry (stat packet plus MD5), or
475 conversely, the FD sends it to the Director and the Director
476 says yes or no. This can be quite rapid if the FileId is kept
477 by the FD for each Base Filename.
478 - It is probably better to have the comparison done by the FD
479 despite the fact that the File entry must be sent across the
481 - An alternative would be to send the FD the whole File entry
482 from the start. The disadvantage is that it requires a lot of
483 space. The advantage is that it requires less communications
485 - The Job record must be updated to indicate that one or more
487 - At end of Job, FD returns:
488 1. Count of base files/bytes not written to tape (i.e. matches)
489 2. Count of base file that were saved i.e. had changed.
490 - No tape record would be written for a Base file that matches, in the
491 same way that no tape record is written for Incremental jobs where
492 the file is not saved because it is unchanged.
493 - On a restore, all the Base file records must explicitly be
494 found from the BaseFile tape. I.e. for each Full save that is marked
495 to have one or more Base Jobs, search the BaseFile for all occurrences
497 - An optimization might be to make the BaseFile have:
503 This would avoid the need to explicitly fetch each File record for
504 the Base job. The Base Job record will be fetched to get the
505 VolSessionId and VolSessionTime.
506 =========================================================
508 ==========================================================
510 For each Incremental job that is run, there may be files that
511 were found but not saved because they were locked (this applies
512 only to Windows). Such a system could send back to the Director
513 a list of Unsaved files.
515 - New UnSavedFiles table that contains:
519 - Then in the next Incremental job, the list of Unsaved Files will be
520 feed to the FD, who will ensure that they are explicitly chosen even
521 if standard date/time check would not have selected them.
522 =============================================================
526 =============================================================
528 Request For Comments For File Backup Options
531 Subject: File Backup Options
534 A few days ago, a Bacula user who is backing up to file volumes and
535 using compression asked if it was possible to suppress compressing
536 all .gz files since it was a waste of CPU time. Although Bacula
537 currently permits using different options (compression, ...) on
538 a directory by directory basis, it cannot do it on a file by
539 file basis, which is clearly what was desired.
541 Proposed Implementation:
542 To solve this problem, I propose the following:
544 - Add a new Director resource type called FileOptions.
546 - The FileOptions resource will have records for all
547 options that can currently be specified on the Include record
548 (in a FileSet). Examples below.
550 - The FileOptions resource will permit an exclude option as well
551 as a number of additional options.
553 - The heart of the FileOptions resource is the ability to
554 supply any number of ApplyTo records which specify POSIX
555 regular expressions. These ApplyTo regular expressions are
556 applied to the fully qualified filename (path and all). If
557 one matches, then the FileOptions will be used.
559 - When an ApplyTo specification matches an included file, the
560 options specified in the FileOptions resource will override
561 the default options specified on the Include record.
563 - Include records will be modified to permit referencing one or
564 more FileOptions resources. The FileOptions will be used
565 in the order listed on the Include record and the first
566 one that matches will be applied.
568 - Options (or specifications) currently supplied on the Include
569 record will be deprecated (i.e. removed in a later version a
570 year or so from now).
572 - The Exclude record will be deprecated as the same functionality
573 can be obtained by using an Exclude = yes in the FileOptions.
576 The following records can appear in the FileOptions resource. An
577 asterisk preceding the name indicates a feature not currently
581 - Compression= (GZIP, ...)
582 - Signature= (MD5, SHA1, ...)
584 - OneFs= (yes/no) - remain on one filesystem
585 - Recurse= (yes/no) - recurse into subdirectories
586 - Sparse= (yes/no) - do sparse file backup
587 - *Exclude= (yes/no) - exclude file from being saved
588 - *Reader= (filename) - external read (backup) program
589 - *Plugin= (filename) - read/write plugin module
592 - verify= (ipnougsamc5) - verify options
595 - replace= (always/ifnewer/ifolder/never) - replace options currently
597 - *Writer= (filename) - external write (restore) program
601 Currently options specifying compression, MD5 signatures, recursion,
602 ... of a FileSet are supplied on the Include record. These will now
603 all be collected into a FileOptions resource, which will be
604 specified on the Include in place of the options. Multiple FileOptions
605 may be specified. Since the FileOptions contain regular expressions
606 that are applied to the full filename, this will give the ability
607 to specify backup options on a file by file basis to whatever level
616 Include = compression=GZIP signature=MD5 {
625 Include = FileOptions=Opts {
636 That's a lot more to do the same thing, but it gives the ability to
637 apply options on a file by file basis. For example, suppose you
638 want to compress all files but not any file with extensions .gz or .Z.
639 You could do so as follows:
643 Include = FileOptions=NoCompress FileOptions=Opts {
651 ApplyTo = /*.?*/ # matches all files
656 # Note multiple ApplyTos are ORed
657 ApplyTo = /*.gz/ # matches .gz files */
658 ApplyTo = /*.Z/ # matches .Z files */
661 Now, since the NoCompress FileOptions is specified first on the
662 Include line, any *.gz or *.Z file will have an MD5 signature computed,
663 but will not be compressed. For all other files, the NoCompress will not
664 match, so the Opts options will be used which will include GZIP
668 - Is it necessary to provide some means of ANDing regular expressions
669 and negation? (not currently planned)
671 e.g. ApplyTo = /*.gz/ && !/big.gz/
673 - I see that Networker has a "null" module which, if specified, does not
674 backup the file, but does make an record of the file in the catalog
675 so that the catalog will reflect an exact picture of the filesystem.
676 The result is that the file can be "seen" when "browsing" the save
677 sets, but it cannot be restored.
679 Is this really useful? Should it be implemented in Bacula?
682 After implementing the above, the user will be able to specify
683 on a file by file basis (using regular expressions) what options are
684 applied for the backup.
685 ====================================
687 =========================================
688 Proposal by Bill Sellers
690 Return-Path: <w.a.sellers@larc.nasa.gov>
691 Received: from post.larc.nasa.gov (post.larc.nasa.gov [128.155.4.45]) by matou.sibbald.com (8.11.6/8.11.6) with ESMTP id h0ELUIm07622 for <kern@sibbald.com>; Tue, 14 Jan 2003 22:30:18 +0100
692 Received: from Baron.larc.nasa.gov (baron.larc.nasa.gov [128.155.40.132]) by post.larc.nasa.gov (pohub4.6) with ESMTP id QAA09768 for <kern@sibbald.com>; Tue, 14 Jan 2003 16:30:14 -0500 (EST)
693 Message-Id: <5.1.0.14.2.20030114153452.028dbae8@pop.larc.nasa.gov>
694 X-Sender: w.a.sellers@pop.larc.nasa.gov
695 X-Mailer: QUALCOMM Windows Eudora Version 5.1
696 Date: Tue, 14 Jan 2003 16:30:18 -0500
697 To: Kern Sibbald <kern@sibbald.com>
698 From: Bill Sellers <w.a.sellers@larc.nasa.gov>
699 Subject: Re: [Bacula-users] Bacula remote storage?
700 In-Reply-To: <1042565382.1845.177.camel@rufus>
701 References: <5.1.0.14.2.20030114113004.0293a210@pop.larc.nasa.gov> <5.1.0.14.2.20030113170650.028dad88@pop.larc.nasa.gov> <5.1.0.14.2.20030113170650.028dad88@pop.larc.nasa.gov> <5.1.0.14.2.20030114113004.0293a210@pop.larc.nasa.gov>
703 Content-Type: text/plain; charset="us-ascii"; format=flowed
704 X-Annoyance-Filter-Junk-Probability: 0
705 X-Annoyance-Filter-Classification: Mail
706 At 06:29 PM 1/14/2003 +0100, you wrote:
709 >Well, if you cannot put a Bacula client on the machine,
710 >then it is a big problem. If you know of some software
711 >that can do what you want, let me know, because I
712 >really just don't know how to do it -- at least not
718 We have been able to get Amanda to use the HSM as a storage
719 device. Someone here wrote a driver for Amanda. BUT, Amanda doesn't
720 handle Windows systems very well (or at all without Samba). So I am
721 looking for a backup system that has a Windows client. I really like the
722 Windows integration of Bacula.
724 From the command line, its rather trivial to move the data around. We use
727 tar cf - ./files | gzip -c | rsh hsm dd of=path/file.tgz
729 or if you use GNU tar:
731 tar czf hsm:path/file.tgz ./files
733 One idea for you to consider; Sendmail offers pipes in the aliases file;
734 (mailpipe: "|/usr/bin/vacation root") and Perl supports pipes in the
735 "open" statement (open FILE, "|/bin/nroff -man";) Could you could make a
736 pipe available, as a storage device? Then we could use any command that
737 handles stdin as a storage destination.
744 #Password is not used in rsh, but might be used in ftp.
745 Device = "| gzip -c | rsh hsm dd of=path/file.tgz"
752 Password = "foobar&-"
753 Device = "| ncftpput -c hsm /path/file.bacula"
757 >If you have some local storage available, you could
758 >use Bacula to backup to disk volumes, then use some
759 >other software (ftp, scp) to move them to the HSM
760 >machine. However, this is a bit kludgy.
763 It is, but maybe worth a try. Is there some function in Bacula to put
764 variables in filenames? i.e. backup.2003-01-15.root
771 w.a.sellers@larc.nasa.gov
773 ==============================================
774 The Project for the above
776 I finally realized that this is not at all
777 the same as reader/writer programs or plugins,
778 which are alternate ways of accessing the
779 files to be backed up. Rather, it is an alternate
780 form of storage device, and I have always planned
781 that Bacula should be able to handle all sorts
784 So, I propose the following phases:
786 1. OK from you to invest some time in testing
787 this as I implement it (requires that you
788 know how to download from the SourceForge
789 cvs -- which I imagine is a piece of cake
792 2. Dumb implementation by allowing a device to
793 be a fifo for write only.
794 Reason: easy to implement, proof of concept.
796 3. Try reading from fifo but with fixed block
798 Reason: proof of concept, easy to implement.
800 4. Extend reading from fifo (restores) to handle
802 Reason: requires some delicate low level coding
803 which could destabilize all of Bacula.
805 5. Implementation of above but to a program. E.g.
806 Device = "|program" (not full pipeline).
807 Reason: routines already exist, and program can
808 be a shell script which contains anything.
810 6. Add full pipeline as a possibility. E.g.
811 Device = "| gzip -c | rsh hsm dd of=path/file.tgz"
812 Reason: needs additional coding to implement full
813 pipeline (must fire off either a shell or all
814 programs and connect their pipes).
816 There are a good number of details in each step
817 that I have left out, but I will specify them at
818 every stage, and there may be a few changes as things
819 evolve. I expect that to get to stage 5 will take a
820 few weeks, and at that point, you will have
821 everything you need (just inside a script).
822 Stage 6 will probably take longer, but if this
823 project pleases you, what we do for 5 should
824 be adequate for some time.
828 =============================================
830 Done: (see kernsdone for more)
831 - Look into Pruning/purging problems or why there seem to
832 be so many files listed each night.
833 - Fix cancel in find_one -- need jcr.
834 - Cancel does not work for restore in FD.
835 - Write SetJobStatus() function so cancel status not lost.
836 - Add include list to end of chain in findlib
837 - Zap sd_auth_key after use
838 - Add Bar code reading capabilities (new mtx-changer)
839 - Figure out some way to automatically backup all local partitions
840 - Make hash table for linked files in findlib/find_one.c:161
842 - Rewrite find_one.c to use only pool_memory instead of
843 alloca and malloc (probably not necessary).
844 - Make sure btraceback goes into /sbin not sysconf directory.
845 - InitVerify is getting pruned and it shouldn't (document it)
846 - Make 1.28c release ??? NO do 1.29 directly
847 - Set timeout on opening fifo for save or restore (findlib)
848 - Document FIFO storage device.
849 - Document fifo and | and <
852 - Get correct error status from run_program or open_bpipe().
853 - Restrict permissions on File Volumes (now 0640).
854 - Umasked 022 daemons
855 - Fix restore of hard linked file.
856 - Figure out how to allow multiple simultaneous file Volumes on a single device.
857 - Implement multiple simultaneous file Volumes on a single device.
858 - Cleanup db_update_media and db_update_pool
859 - Flush all the daemon messages at the end of every job.
860 - Change stat1= fgets()!=NULL to stat1=fgest()==NULL; in
861 run_program -- bpipe.c