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 - Figure out some way to specify a retention period for files
20 that no longer exist on the machine -- so that we maintain
21 say backups for 30 days, but if the file is deleted, we maintain
22 the last copy for 1 year.
23 - Have Bacula "poll" the tape to see if it is there.
24 - Add more of the config info to the tape label.
25 - Apparently cancel does not work for jobs waiting to be
27 - Implement bar code reader for autochangers
28 - Pull a canceled job from the Scheduling queue.
29 - Add a default File storage so that new users can do backup
30 and restores right away.
31 - Implement a Mount Command and an Unmount Command where
32 the users could specify a system command to be performed
33 to do the mount, after which Bacula could attempt to
34 read the device. This is for Removeable media such as a
36 - Most likely, this mount command would be invoked explicitly
37 by the user using the current Console "mount" and "unmount"
38 commands -- the Storage Daemon would do the right thing
39 depending on the exact nature of the device.
40 - As with tape drives, when Bacula wanted a new removable
41 disk mounted, it would unmount the old one, and send a message
42 to the user, who would then use "mount" as described above
43 once he had actually inserted the disk.
45 - Implement TCP/IP connection for MySQL
46 - Make some way so that if a machine is skipped because it is not up
47 that Bacula will continue retrying for a specified period of time --
49 - Figure out some way to estimate output size and to avoid splitting
50 a backup across two Volumes -- this could be useful for writing CDROMs
51 where you really prefer not to have it split -- not serious.
52 - Add chflags() code for FreeBSD file flags
53 - Add RunBeforeJob and RunAfterJob to the Client program.
54 - Have SD compute MD5 or SHA1 and compare to what FD computes.
55 - Make VolumeToCatalog calculate an MD5 or SHA1 from the
56 actual data on the Volume and compare it.
57 - Implement FileOptions (see end of this document)
58 - Implement Bacula plugins -- design API
59 - Make bcopy read through bad tape records.
60 - Fix read_record to handle multiple sessions.
61 - Program files (i.e. execute a program to read/write files).
62 Pass read date of last backup, size of file last time.
63 - Add Signature type to File DB record.
64 - Make Restore report an error if FD or SD term codes are not OK.
65 - CD into subdirectory when open()ing files for backup to
66 speed up things. Test with testfind().
67 - Add prefixlinks to where or not where absolute links to FD.
68 - Look at handling <> in smtp doesn't work with exim.
69 - Priority job to go to top of list.
70 - Implement Bar code handling
71 - Why is catreq.c:111 Find vol called twice for a job?
72 - Find out why Full saves run slower and slower (hashing?)
73 - Why are save/restore of device different sizes (sparse?) Yup! Fix it.
74 - Implement some way for the Console to dynamically create a job.
75 - Restore to a particular time -- e.g. before date, after date.
76 - Implement disk spooling
77 - Implement finer multiprocessing options.
78 - Solaris -I on tar for include list
79 - Enable avoid backing up archive device (findlib/find_one.c:128)
80 - Need a verbose mode in restore, perhaps to bsr.
81 - bscan without -v is too quiet -- perhaps show jobs.
82 - Add code to reject whole blocks if not wanted on restore.
83 - Start working on Base jobs.
84 - Make sure the MaxVolFiles is fully implemented in SD
85 - Check if both CatalogFiles and UseCatalog are set to SD.
86 - Check if we can increase Bacula FD priorty in Win2000
87 - Need return status on read_cb() from read_records(). Need multiple
88 records -- one per Job, maybe a JCR or some other structure with
90 - Figure out how to do a bare metal Windows restore
91 - Put system type returned by FD into catalog.
92 - Need to specify MaximumConcurrentJobs in the Job resource.
93 - Possibly add email to Watchdog if drive is unmounted too
94 long and a job is waiting on the drive.
95 - Strip trailing slashes from Include directory names in the FD.
96 - Use read_record.c in SD code.
97 - Why don't we get an error message from Win32 FD when bootstrap
98 file cannot be created for restore command?
99 - When Marking a file in Restore that is a hard link, also
100 mark the link so that the data will be reloaded.
101 - Restore program that errors in SD due to no tape reports
102 OK incorrectly in output.
103 - After unmount, if restore job started, ask to mount.
104 - Convert all %x substitution variables, which are hard to remember
105 and read to %(variable-name). Idea from TMDA.
106 - Add JobLevel in FD status (but make sure it is defined).
107 - Make Pool resource handle Counter resources.
108 - Remove NextId for SQLite. Optimize.
109 - Strip trailing / from Include
110 - Move all SQL statements into a single location.
111 - Add UA rc and history files.
112 - put termcap (used by console) in ./configure and
113 allow -with-termcap-dir.
114 - Enhance time and size scanning routines.
115 - Fix Autoprune for Volumes to respect need for full save.
116 - Fix Win32 config file definition name on /install
117 - No READLINE_SRC if found in alternate directory.
118 - Add Client FS/OS id (Linux, Win95/98, ...).
119 - Test a second language e.g. french.
120 - Compare tape to Client files (attributes, or attributes and data)
121 - Make all database Ids 64 bit.
122 - Write an applet for Linux.
123 - Add estimate to Console commands
124 - Find solution to blank filename (i.e. path only) problem.
125 - Implement new daemon communications protocol.
126 - Remove PoolId from Job table, it exists in Media.
127 - Allow console commands to detach or run in background.
128 - Fix status delay on storage daemon during rewind.
129 - Add SD message variables to control operator wait time
130 - Maximum Operator Wait
131 - Minimum Message Interval
132 - Maximum Message Interval
133 - Send Operator message when cannot read tape label.
134 - Verify level=Volume (scan only), level=Data (compare of data to file).
135 Verify level=Catalog, level=InitCatalog
137 - Add keyword search to show command in Console.
138 - Fix Win2000 error with no messages during startup.
139 - Events : tape has more than xxx bytes.
140 - Restrict characters permitted in a Resource name.
141 - Complete code in Bacula Resources -- this will permit
142 reading a new config file at any time.
143 - Handle ctl-c in Console
144 - Implement LabelTemplate (at least first cut).
145 - Implement script driven addition of File daemon to config files.
146 - Think about how to make Bacula work better with File (non-tape) archives.
147 - Write Unix emulator for Windows.
149 - Implement new serialize subroutines
150 send(socket, "string", &Vol, "uint32", &i, NULL)
151 - Audit all UA commands to ensure that we always prompt where possible.
152 - If ./btape is called without /dev, assume argument is a Storage resource name.
153 - Put memory utilization in Status output of each daemon
154 if full status requested or if some level of debug on.
155 - Make database type selectable by .conf files i.e. at runtime
156 - gethostbyname failure in bnet_connect() continues
157 generating errors -- should stop.
158 - Set flag for uname -a. Add to Volume label.
159 - Implement throttled work queue.
160 - Check for EOT at ENOSPC or EIO or ENXIO (unix Pc)
161 - Allow multiple Storage specifications (or multiple names on
162 a single Storage specification) in the Job record. Thus a job
163 can be backed up to a number of storage devices.
164 - Implement dump label to UA
165 - Concept of VolumeSet during restore which is a list
166 of Volume names needed.
167 - Restore files modified after date
168 - Restore file modified before date
169 - Emergency restore info:
171 - Backup working directory
173 - Restore -- do nothing but show what would happen
174 - SET LD_RUN_PATH=$HOME/mysql/lib/mysql
175 - Implement Restore FileSet=
176 - Create a protocol.h and protocol.c where all protocol messages
178 - If SD cannot open a drive, make it periodically retry.
179 - Remove duplicate fields from jcr (e.g. jcr.level and jcr.jr.Level, ...).
180 - Timout a job or terminate if link goes down, or reopen link and query.
181 - Find general solution for sscanf size problems (as well
182 as sprintf. Do at run time?
183 - Concept of precious tapes (cannot be reused).
184 - Make bcopy copy with a single tape drive.
185 - Permit changing ownership during restore.
187 - Autolabel should be specified by DIR instead of SD.
188 - Find out how to get the system tape block limits, e.g.:
189 Apr 22 21:22:10 polymatou kernel: st1: Block limits 1 - 245760 bytes.
190 Apr 22 21:22:10 polymatou kernel: st0: Block limits 2 - 16777214 bytes.
193 - AutoScan (check checksum of tape)
194 - Format command = "format /dev/nst0"
198 - Seek resolution (usually corresponds to buffer size)
199 - EODErrorCode=ENOSPC or code
200 - Partial Read error code
201 - Partial write error code
202 - Nonformatted read error
203 - Nonformatted write error
204 - WriteProtected error
208 - IgnoreCloseErrors=yes
218 - FD sends unsaved file list to Director at end of job (see
220 - File daemon should build list of files skipped, and then
221 at end of save retry and report any errors.
222 - Write a Storage daemon that uses pipes and
223 standard Unix programs to write to the tape.
225 - Need something that monitors the JCR queue and
226 times out jobs by asking the deamons where they are.
227 - Enhance Jmsg code to permit buffering and saving to disk.
228 - device driver = "xxxx" for drives.
229 - restart: paranoid: read label fsf to
230 eom read append block, and go
231 super-paranoid: read label, read all files
232 in between, read append block, and go
233 verify: backspace, read append block, and go
234 permissive: same as above but frees drive
235 if tape is not valid.
237 - Ensure that /dev/null works
238 - Need report class for messages. Perhaps
239 report resource where report=group of messages
240 - enhance scan_attrib and rename scan_jobtype, and
241 fill in code for "since" option
242 - Director needs a time after which the report status is sent
243 anyway -- or better yet, a retry time for the job.
244 Don't reschedule a job if previous incarnation is still running.
245 - Some way to automatically backup everything is needed????
246 - Need a structure for pending actions:
248 - termination status (part of buffered msgs?)
249 - Concept of grouping Storage devices and job can use
250 any of a number of devices
252 Read, Write, Clean, Delete
253 - Login to Bacula; Bacula users with different permissions:
254 owner, group, user, quotas
255 - Store info on each file system type (probably in the job header on tape.
256 This could be the output of df; or perhaps some sort of /etc/mtab record.
259 - Design at hierarchial storage for Bacula.
260 - Implement FSM (File System Modules).
261 - Identify unchanged or "system" files and save them to a
262 special tape thus removing them from the standard
263 backup FileSet -- BASE backup.
264 - Turn virutally all sprintfs into snprintfs.
265 - Heartbeat between daemons.
266 - Audit M_ error codes to ensure they are correct and consistent.
267 - Add variable break characters to lex analyzer.
268 Either a bit mask or a string of chars so that
269 the caller can change the break characters.
270 - Make a single T_BREAK to replace T_COMMA, etc.
271 - Ensure that File daemon and Storage daemon can
272 continue a save if the Director goes down (this
273 is NOT currently the case). Must detect socket error,
274 buffer messages for later.
275 - Enhance time/duration input to allow multiple qualifiers e.g. 3d2h
276 - Add ability to backup to two Storage devices (two SD sessions) at
277 the same time -- e.g. onsite, offsite.
278 - Add the ability to consolidate old backup sets (basically do a restore
279 to tape and appropriately update the catalog). Compress Volume sets.
280 Might need to spool via file is only one drive is available.
281 - Compress or consolidate Volumes of old possibly deleted files. Perhaps
282 someway to do so with every volume that has less than x% valid
287 Bacula Projects Roadmap
289 last update 5 January 2003
291 Item 1: Multiple simultaneous Jobs. (done)
292 Done -- Restore part needs better implementation to work correctly
294 What: Permit multiple simultaneous jobs in Bacula.
296 Why: An enterprise level solution needs to go fast without the
297 need for the system administrator to carefully tweak
298 timing. Based on the benchmarks, during a full
299 backup, NetWorker typically hit 10 times the bandwidth to
300 the tape compared to Bacula--largely. This is probably due to
301 running parallel jobs and multi-threaded filling of buffers
302 and writing them to tape. This should also make things work
303 better when you have a mix of fast and slow machines backing
306 Notes: Bacula was designed to run multiple simultaneous jobs. Thus
307 implementing this is a matter of some small cleanups and
311 Item 2: Make the Storage daemon use intermediate file storage to buffer data.
312 Deferred -- not necessary yet.
314 What: If data is coming into the SD too fast, buffer it to
315 disk if the user has configured this option.
317 Why: This would be nice, especially if it more or less falls out
318 when implementing (1) above. If not, it probably should not
319 be given a high priority because fundamentally the backup time
320 is limited by the tape bandwidth. Even though you may finish a
321 client job quicker by spilling to disk, you still have to
322 eventually get it onto tape. If intermediate disk buffering
323 allows us to improve write bandwidth to tape, it may make
326 Notes: Whether or not this is implemented will depend upon performance
327 testing after item 1 is implemented.
330 Item 3: Write the bscan program -- also write a bcopy program.
333 What: Write a program that reads a Bacula tape and puts all the
334 appropriate data into the catalog. This allows recovery
335 from a tape that is no longer in the database, or it allows
336 re-creation of a database if lost.
338 Why: This is a fundamental robustness and disaster recovery tool
339 which will increase the comfort level of a sysadmin
340 considering adopting Bacula.
342 Notes: A skeleton of this program already exists, but much work
343 needs to be done. Implementing this will also make apparent
344 any deficiencies in the current Bacula tape format.
347 Item 4: Implement Base jobs.
349 What: A base job is sort of like a Full save except that you
350 will want the FileSet to contain only files that are unlikely
351 to change in the future (i.e. a snapshot of most of your
352 system after installing it). After the base job has been run,
353 when you are doing a Full save, you can specify to exclude
354 all files saved by the base job that have not been modified.
356 Why: This is something none of the competition does, as far as we know
357 (except BackupPC, which is a Perl program that saves to disk
358 only). It is big win for the user, it makes Bacula stand out
359 as offering a unique optimization that immediately saves time
362 Notes: Big savings in tape usage. Will require more resources because
363 the e. DIR must send FD a list of files/attribs, and the FD must
364 search the list and compare it for each file to be saved.
367 Item 5: Implement Label templates
369 What: This is a mechanism whereby Bacula can automatically create
370 a tape label for new tapes according to a detailed specification
371 provided by the user.
373 Why: It is a major convenience item for folks who use automated label
376 Notes: Bacula already has a working form of automatic tape label
377 creation, but it is very crude. The design for the complete
378 tape labeling project is already documented in the manual.
381 Item 6: Write a regression script.
384 What: This is an automatic script that runs and tests as many features
385 of Bacula as possible. The output is compared to previous
386 versions of Bacula and any differences are reported.
388 Why: This is an enormous help in preventing introduction of new
389 errors in parts of the program that already work correctly.
391 Notes: This probably should be ranked higher, it's something the typical
392 user doesn't see. Depending on how it's implemented, it may
393 make sense to defer it until the archival tape format and
394 user interface mature.
397 Item 7: GUI for interactive restore
398 Item 8: GUI for interactive backup
400 What: The current interactive restore is implemented with a tty
401 interface. It would be much nicer to be able to "see" the
402 list of files backed up in typical GUI tree format.
403 The same mechanism could also be used for creating
404 ad-hoc backup FileSets (item 8).
406 Why: Ease of use -- especially for the end user.
408 Notes: Rather than implementing in Gtk, we probably should go directly
409 for a Browser implementation, even if doing so meant the
410 capability wouldn't be available until much later. Not only
411 is there the question of Windows sites, most
412 Solaris/HP/IRIX, etc, shops can't currently run Gtk programs
413 without installing lots of stuff admins are very wary about.
414 Real sysadmins will always use the command line anyway, and
415 the user who's doing an interactive restore or backup of his
416 own files will in most cases be on a Windows machine running
420 Item 9: Add SSL to daemon communications.
422 What: This provides for secure communications between the daemons.
424 Why: This would allow doing backup across the Internet without
425 privacy concerns (or with much less concern).
427 Notes: The vast majority of near term potential users will be backing up
428 a single site over a LAN and, correctly or not, they probably
429 won't be concerned with security, at least not enough to go to
430 the trouble to set up keys, etc. to screw things down. We suspect
431 that many users genuinely interested in multi-site backup
432 already run some form of VPN software in their internetwork
433 connections, and are willing to delegate security to that layer.
436 Item 10: Define definitive tape format.
439 What: Define that definitive tape format that will not change
440 for the next millennium.
442 Why: Stability, security.
444 Notes: See notes for item 11 below.
447 Item 11: New daemon communication protocol.
449 What: The current daemon to daemon protocol is basically an ASCII
450 printf() and sending the buffer. On the receiving end, the
451 buffer is sscanf()ed to unpack it. The new scheme would
452 be a binary format that allows quick packing and unpacking
453 of any data type with named fields.
455 Why: Using binary packing would be faster. Named fields will permit
456 error checking to ensure that what is sent is what the
457 receiver really wants.
459 Notes: These are internal improvements in the interest of the
460 long-term stability and evolution of the program. On the one
461 hand, the sooner they're done, the less code we have to rip
462 up when the time comes to install them. On the other hand, they
463 don't bring an immediately perceptible benefit to potential
464 users. Item 10 and possibly item 11 should be deferred until Bacula
465 is well established with a growing user community more or
466 less happy with the feature set. At that time, it will make a
467 good "next generation" upgrade in the interest of data
473 ======================================================
475 It is somewhat like a Full save becomes an incremental since
476 the Base job (or jobs) plus other non-base files.
478 - New BaseFile table that contains:
479 JobId, BaseJobId, FileId (from Base).
480 i.e. for each base file that exists but is not saved because
481 it has not changed, the File daemon sends the JobId, BaseId,
482 and FileId back to the Director who creates the DB entry.
483 - To initiate a Base save, the Director sends the FD
484 the FileId, and full filename for each file in the Base.
485 - When the FD finds a Base file, he requests the Director to
486 send him the full File entry (stat packet plus MD5), or
487 conversely, the FD sends it to the Director and the Director
488 says yes or no. This can be quite rapid if the FileId is kept
489 by the FD for each Base Filename.
490 - It is probably better to have the comparison done by the FD
491 despite the fact that the File entry must be sent across the
493 - An alternative would be to send the FD the whole File entry
494 from the start. The disadvantage is that it requires a lot of
495 space. The advantage is that it requires less communications
497 - The Job record must be updated to indicate that one or more
499 - At end of Job, FD returns:
500 1. Count of base files/bytes not written to tape (i.e. matches)
501 2. Count of base file that were saved i.e. had changed.
502 - No tape record would be written for a Base file that matches, in the
503 same way that no tape record is written for Incremental jobs where
504 the file is not saved because it is unchanged.
505 - On a restore, all the Base file records must explicitly be
506 found from the BaseFile tape. I.e. for each Full save that is marked
507 to have one or more Base Jobs, search the BaseFile for all occurrences
509 - An optimization might be to make the BaseFile have:
515 This would avoid the need to explicitly fetch each File record for
516 the Base job. The Base Job record will be fetched to get the
517 VolSessionId and VolSessionTime.
518 =========================================================
520 ==========================================================
522 For each Incremental job that is run, there may be files that
523 were found but not saved because they were locked (this applies
524 only to Windows). Such a system could send back to the Director
525 a list of Unsaved files.
527 - New UnSavedFiles table that contains:
531 - Then in the next Incremental job, the list of Unsaved Files will be
532 feed to the FD, who will ensure that they are explicitly chosen even
533 if standard date/time check would not have selected them.
534 =============================================================
538 =============================================================
540 Request For Comments For File Backup Options
543 Subject: File Backup Options
546 A few days ago, a Bacula user who is backing up to file volumes and
547 using compression asked if it was possible to suppress compressing
548 all .gz files since it was a waste of CPU time. Although Bacula
549 currently permits using different options (compression, ...) on
550 a directory by directory basis, it cannot do it on a file by
551 file basis, which is clearly what was desired.
553 Proposed Implementation:
554 To solve this problem, I propose the following:
556 - Add a new Director resource type called FileOptions.
558 - The FileOptions resource will have records for all
559 options that can currently be specified on the Include record
560 (in a FileSet). Examples below.
562 - The FileOptions resource will permit an exclude option as well
563 as a number of additional options.
565 - The heart of the FileOptions resource is the ability to
566 supply any number of ApplyTo records which specify POSIX
567 regular expressions. These ApplyTo regular expressions are
568 applied to the fully qualified filename (path and all). If
569 one matches, then the FileOptions will be used.
571 - When an ApplyTo specification matches an included file, the
572 options specified in the FileOptions resource will override
573 the default options specified on the Include record.
575 - Include records will be modified to permit referencing one or
576 more FileOptions resources. The FileOptions will be used
577 in the order listed on the Include record and the first
578 one that matches will be applied.
580 - Options (or specifications) currently supplied on the Include
581 record will be deprecated (i.e. removed in a later version a
582 year or so from now).
584 - The Exclude record will be deprecated as the same functionality
585 can be obtained by using an Exclude = yes in the FileOptions.
588 The following records can appear in the FileOptions resource. An
589 asterisk preceding the name indicates a feature not currently
593 - Compression= (GZIP, ...)
594 - Signature= (MD5, SHA1, ...)
596 - OneFs= (yes/no) - remain on one filesystem
597 - Recurse= (yes/no) - recurse into subdirectories
598 - Sparse= (yes/no) - do sparse file backup
599 - *Exclude= (yes/no) - exclude file from being saved
600 - *Reader= (filename) - external read (backup) program
601 - *Plugin= (filename) - read/write plugin module
604 - verify= (ipnougsamc5) - verify options
607 - replace= (always/ifnewer/ifolder/never) - replace options currently
609 - *Writer= (filename) - external write (restore) program
613 Currently options specifying compression, MD5 signatures, recursion,
614 ... of a FileSet are supplied on the Include record. These will now
615 all be collected into a FileOptions resource, which will be
616 specified on the Include in place of the options. Multiple FileOptions
617 may be specified. Since the FileOptions contain regular expressions
618 that are applied to the full filename, this will give the ability
619 to specify backup options on a file by file basis to whatever level
628 Include = compression=GZIP signature=MD5 {
637 Include = FileOptions=Opts {
648 That's a lot more to do the same thing, but it gives the ability to
649 apply options on a file by file basis. For example, suppose you
650 want to compress all files but not any file with extensions .gz or .Z.
651 You could do so as follows:
655 Include = FileOptions=NoCompress FileOptions=Opts {
663 ApplyTo = /*.?*/ # matches all files
668 # Note multiple ApplyTos are ORed
669 ApplyTo = /*.gz/ # matches .gz files */
670 ApplyTo = /*.Z/ # matches .Z files */
673 Now, since the NoCompress FileOptions is specified first on the
674 Include line, any *.gz or *.Z file will have an MD5 signature computed,
675 but will not be compressed. For all other files, the NoCompress will not
676 match, so the Opts options will be used which will include GZIP
680 - Is it necessary to provide some means of ANDing regular expressions
681 and negation? (not currently planned)
683 e.g. ApplyTo = /*.gz/ && !/big.gz/
685 - I see that Networker has a "null" module which, if specified, does not
686 backup the file, but does make an record of the file in the catalog
687 so that the catalog will reflect an exact picture of the filesystem.
688 The result is that the file can be "seen" when "browsing" the save
689 sets, but it cannot be restored.
691 Is this really useful? Should it be implemented in Bacula?
694 After implementing the above, the user will be able to specify
695 on a file by file basis (using regular expressions) what options are
696 applied for the backup.
697 ====================================
699 =========================================
700 Proposal by Bill Sellers
702 Return-Path: <w.a.sellers@larc.nasa.gov>
703 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
704 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)
705 Message-Id: <5.1.0.14.2.20030114153452.028dbae8@pop.larc.nasa.gov>
706 X-Sender: w.a.sellers@pop.larc.nasa.gov
707 X-Mailer: QUALCOMM Windows Eudora Version 5.1
708 Date: Tue, 14 Jan 2003 16:30:18 -0500
709 To: Kern Sibbald <kern@sibbald.com>
710 From: Bill Sellers <w.a.sellers@larc.nasa.gov>
711 Subject: Re: [Bacula-users] Bacula remote storage?
712 In-Reply-To: <1042565382.1845.177.camel@rufus>
713 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>
715 Content-Type: text/plain; charset="us-ascii"; format=flowed
716 X-Annoyance-Filter-Junk-Probability: 0
717 X-Annoyance-Filter-Classification: Mail
718 At 06:29 PM 1/14/2003 +0100, you wrote:
721 >Well, if you cannot put a Bacula client on the machine,
722 >then it is a big problem. If you know of some software
723 >that can do what you want, let me know, because I
724 >really just don't know how to do it -- at least not
730 We have been able to get Amanda to use the HSM as a storage
731 device. Someone here wrote a driver for Amanda. BUT, Amanda doesn't
732 handle Windows systems very well (or at all without Samba). So I am
733 looking for a backup system that has a Windows client. I really like the
734 Windows integration of Bacula.
736 From the command line, its rather trivial to move the data around. We use
739 tar cf - ./files | gzip -c | rsh hsm dd of=path/file.tgz
741 or if you use GNU tar:
743 tar czf hsm:path/file.tgz ./files
745 One idea for you to consider; Sendmail offers pipes in the aliases file;
746 (mailpipe: "|/usr/bin/vacation root") and Perl supports pipes in the
747 "open" statement (open FILE, "|/bin/nroff -man";) Could you could make a
748 pipe available, as a storage device? Then we could use any command that
749 handles stdin as a storage destination.
756 #Password is not used in rsh, but might be used in ftp.
757 Device = "| gzip -c | rsh hsm dd of=path/file.tgz"
764 Password = "foobar&-"
765 Device = "| ncftpput -c hsm /path/file.bacula"
769 >If you have some local storage available, you could
770 >use Bacula to backup to disk volumes, then use some
771 >other software (ftp, scp) to move them to the HSM
772 >machine. However, this is a bit kludgy.
775 It is, but maybe worth a try. Is there some function in Bacula to put
776 variables in filenames? i.e. backup.2003-01-15.root
783 w.a.sellers@larc.nasa.gov
785 ==============================================
786 The Project for the above
788 I finally realized that this is not at all
789 the same as reader/writer programs or plugins,
790 which are alternate ways of accessing the
791 files to be backed up. Rather, it is an alternate
792 form of storage device, and I have always planned
793 that Bacula should be able to handle all sorts
796 So, I propose the following phases:
798 1. OK from you to invest some time in testing
799 this as I implement it (requires that you
800 know how to download from the SourceForge
801 cvs -- which I imagine is a piece of cake
804 2. Dumb implementation by allowing a device to
805 be a fifo for write only.
806 Reason: easy to implement, proof of concept.
808 3. Try reading from fifo but with fixed block
810 Reason: proof of concept, easy to implement.
812 4. Extend reading from fifo (restores) to handle
814 Reason: requires some delicate low level coding
815 which could destabilize all of Bacula.
817 5. Implementation of above but to a program. E.g.
818 Device = "|program" (not full pipeline).
819 Reason: routines already exist, and program can
820 be a shell script which contains anything.
822 6. Add full pipeline as a possibility. E.g.
823 Device = "| gzip -c | rsh hsm dd of=path/file.tgz"
824 Reason: needs additional coding to implement full
825 pipeline (must fire off either a shell or all
826 programs and connect their pipes).
828 There are a good number of details in each step
829 that I have left out, but I will specify them at
830 every stage, and there may be a few changes as things
831 evolve. I expect that to get to stage 5 will take a
832 few weeks, and at that point, you will have
833 everything you need (just inside a script).
834 Stage 6 will probably take longer, but if this
835 project pleases you, what we do for 5 should
836 be adequate for some time.
840 =============================================
842 Done: (see kernsdone for more)
843 - Look into Pruning/purging problems or why there seem to
844 be so many files listed each night.
845 - Fix cancel in find_one -- need jcr.
846 - Cancel does not work for restore in FD.
847 - Write SetJobStatus() function so cancel status not lost.
848 - Add include list to end of chain in findlib
849 - Zap sd_auth_key after use
850 - Add Bar code reading capabilities (new mtx-changer)
851 - Figure out some way to automatically backup all local partitions
852 - Make hash table for linked files in findlib/find_one.c:161
854 - Rewrite find_one.c to use only pool_memory instead of
855 alloca and malloc (probably not necessary).
856 - Make sure btraceback goes into /sbin not sysconf directory.
857 - InitVerify is getting pruned and it shouldn't (document it)
858 - Make 1.28c release ??? NO do 1.29 directly
859 - Set timeout on opening fifo for save or restore (findlib)
860 - Document FIFO storage device.
861 - Document fifo and | and <
864 - Get correct error status from run_program or open_bpipe().
865 - Restrict permissions on File Volumes (now 0640).
866 - Umasked 022 daemons
867 - Fix restore of hard linked file.
868 - Figure out how to allow multiple simultaneous file Volumes on a single device.
869 - Implement multiple simultaneous file Volumes on a single device.
870 - Cleanup db_update_media and db_update_pool
871 - Flush all the daemon messages at the end of every job.
872 - Change stat1= fgets()!=NULL to stat1=fgest()==NULL; in
873 run_program -- bpipe.c