9 - Document running multiple Jobs
10 - Document that two Verifys at same time on same client do not work.
11 - Document how to recycle a tape in 7 days even if the backup takes a long time.
12 - Document default config file locations.
13 - Document better includes (does it cross file systems ?).
14 - Document specifically how to add new File daemon to config files.
15 - Document forcing a new tape to be used.
16 - Document "Error in message.c:500 Mail program terminated in error.
19 --bindir is wrong and does not reflect prefix= in the *_sqlite_* scripts
21 --top level configure options are not passed to the depkgs, particularly
23 --Also, it might be better to split the depkgs location from the --with-sqlite
25 --should be able to specify e.g. --with-sqlite=/opt/local and have it find
26 lib, bin, sbin for itself
27 I tried this and it didn't find sqlite.h
28 --sd.conf password does not match dir.conf storage password
31 - Fix catalog filename truncation in sql_get and elsewhere. Use
32 only a single filename split routine.
33 - Add command to reset VolFiles to a larger value (don't allow
34 a smaller number or print big warning).
35 - Make SD disallow writing on Volume with fewer files than in
37 - Make Restore report an error if FD or SD term codes are not OK.
38 - Convert all %x substitution variables, which are hard to remember
39 and read to %(variable-name)s. Idea from TMDA.
40 - Report volume write rate.
41 - Fix db_get_job_volume_names() to return array of strings.
42 - Report compression % and other compression statistics if turned on.
43 - Add JobLevel in FD status (but make sure it is defined).
44 - Eliminate MySQL shared libraries from smtp and daemons not
46 - Pass "Catalog Files = no" to storage daemon to eliminate
48 - Make Pool resource handle Counter resources.
49 - Remove NextId for SQLite. Optimize.
50 - Fix gethostbyname() to use gethostbyname_r()
51 - Cleanup path/filename separation in sql_get.c and sql_create.c
52 - Implement ./configure --with-client-only
53 - Strip trailing / from Include
54 - Move all SQL statements into a single location.
55 - Cleanup db_update_media and db_update_pool
56 - Add UA rc and history files.
57 - put termcap (used by console) in ./configure and
58 allow -with-termcap-dir.
59 - Remove JobMediaId it is not used.
60 - Enhance time and size scanning routines.
61 - Fix Autoprune for Volumes to respect need for full save.
62 - DateWritten may be wrong.
63 - Fix Win32 config file definition name on /install
64 - When we are at EOM, we must ask each job to write JobMedia
65 record (update_volume_info).
66 - No READLINE_SRC if found in alternate directory.
67 - Add Client FS/OS id (Linux, Win95/98, ...).
68 - Put Windows files in Windows stream?
70 ====== 31 May 2002 ========
71 Now that Bacula 1.20 is released, virtually all the
72 basic features are implemented (some are still quite
73 primitive though). Over the next month or two, I'm
74 planning to focus on the following items:
77 - Fix any bugs I find or you report.
78 - Finish the implementation of automatic pruning
79 (add pruning of Restore and Verify jobs).
80 - Make sure pruning of Volumes won't prune the
81 only backup of a FileSet
84 - Improve the Restore capabilities of Bacula
85 - Restore to most recent system state (i.e.
86 figure out what tapes need to be mounted and
88 - Restore to a particular time (perhaps several
89 variations -- e.g. before date, after date).
90 - Interactive Restore where you get to select
91 what files are to be restored (much like the Unix
92 "restore" program permits). Now that we have a
93 catalog of all files saved, it would be nice to
95 - Restore options (overwrite, overwrite if older,
96 overwrite if newer, never overwrite, ...)
97 - Improve the standalone programs (bls and bextract)
98 to have pattern matching capabilities (e.g. restore
99 by FileSet, Job, JobType, JobLevel, ...).
100 - Ideally after each Job, Bacula could write out a
101 set of commands to a file that if later feed to
102 bextract would restore your system to the current
103 state (at least for the saved FileSet). This would
104 provide a simple disaster recovery that could be
105 initiated from a "floppy" and one simple ASCII control
106 file. I'm not exactly sure how to do this, but it
107 shouldn't be too hard and I'll
108 be trying to go in this direction.
111 - Implement tape verification to ensure that the data
112 written for a particular Job can really be read.
113 - Compare tape File attributes to Catalog.
114 (File attributes are size, dates, MD5, but not
116 - Compare tape to Client files (attributes, or
120 - With the current Bacula 1.21 (not yet in the CVS) I
121 expect there is about 95% chance that running multiple
122 simultaneous Jobs will actually work without stepping
123 on each other. I'm planning to try this sometime soon.
128 - Rework Storage daemon with new rwl_lock routines.
129 - Implement Label templates
130 - Pass JCR to database routines permitting better error printing.
133 - Verify against Full.
136 mysqldump -f --opt bacula >bacula
140 - Probably add End of Data tape records (this would make
141 the tape format incompatible with the previous version).
142 - I'll most likely enhance the current tape format
143 in the way that I previously described, which will make
144 some of the labels incompatible, but the change will
145 not affect the current restore code since it does not
146 look at the details of the labels.
147 - I may add a few more waiting conditions in the Storage
148 daemon where it will current immediately aborts a
149 Job if the necessary resources are not available (e.g.
150 tape is being written and a read request arrives).
151 - Write an applet for Linux.
153 - Remove PoolId from Job table, it exists in Media.
154 - Allow commands to detach or run in background.
155 - Write better dump of Messages resource.
156 - Fix status delay on storage daemon during rewind.
157 - Add VerNo to each Session label record.
158 - Add Job to Session records.
159 - Add VOLUME_CAT_INFO to the EOS tape record (as
160 well as to the EOD record).
161 - Add SD message variables to control operator wait time
162 - Maximum Operator Wait
163 - Minimum Message Interval
164 - Maximum Message Interval
165 - Add EOM handling variables
167 - Require EOD records
168 - Send Operator message when cannot read tape label.
169 - Think about how to handle I/O error on MTEOM.
170 - If Storage daemon aborts a job, ensure that this
171 is printed in the error message.
172 - Verify level=Volume (scan only), level=Data (compare of data to file).
173 Verify level=Catalog, level=InitCatalog
174 - Scan tape contents into database.
176 - Cold start full restore (restore catalog then
177 user selects what to restore). Write summary file containing only
178 Job, Media, and Catalog information. Store on another machine.
179 - Dump/Restore database
182 - Implement first cut of Catalog Retention period (remove old
183 entries from database).
184 - Add SessionTime/Id filters to bextract.
186 - Ensure that Start/End File/Block are correct.
187 - Add keyword search to show command in Console.
188 - If MySQL database is not running, job terminates with
189 wierd type and wierd error code.
190 - Write a regression script
191 - Report bad status from smtp or mail program.
192 - Fix Win2000 error with no messages during startup.
193 - Add estimate to Console
194 - Events : tape has more than xxx bytes.
195 - In Storage daemon, status should include job cancelled.
196 - Write general list maintenance subroutines.
197 - Implement immortal format with EDOs.
198 - Restrict characters permitted in a Resource name.
199 - Restore file xx or files xx, yy to their most recent values.
200 - Provide definitive identification of type in backup.
201 - Complete code in Bacula Resources -- this will permit
202 reading a new config file at any time.
203 - Document new Console
204 - Handle ctl-c in Console
205 - Test restore of Windows backup
206 - Implement LabelTemplate (at least first cut).
207 - Implement script driven addition of File daemon to
210 - Bug: anonymous Volumes requires mount in some cases.
211 - see setgroup and user for Bacula p4-5 of stunnel.c
212 - Implement new serialize subroutines
213 send(socket, "string", &Vol, "uint32", &i, NULL)
214 - Add save type to Session label.
215 - Correct date on Session label.
216 - On I/O error, write EOF, then try to write again.
217 - Audit all UA commands to ensure that we always prompt where
219 - If ./btape is called without /dev, assume argument is
220 a Storage resource name.
221 - Put memory utilization in Status output of each daemon
222 if full status requested or if some level of debug on.
223 - Make database type selectable by .conf files i.e. at runtime
224 - gethostbyname failure in bnet_connect() continues
225 generating errors -- should stop.
226 - Don't create a volume that is already written. I.e. create only once.
227 - If error at end of tape, implement some way to kill waiting processes.
228 - Get correct block/file information in Catalog, pay attention
230 - Add HOST to Volume label.
231 - Set flag for uname -a. Add to Volume label.
232 - Implement throttled work queue.
233 - Write bscan program that will syncronize the DB Media record with
234 the contents of the Volume -- for use after a crash.
235 - Check for EOT at ENOSPC or EIO or ENXIO (unix Pc)
236 - Allow multiple Storage specifications (or multiple names on
237 a single Storage specification) in the Job record. Thus a job
238 can be backed up to a number of storage devices.
239 - Implement full MediaLabel code.
240 - Implement dump label to UA
241 - Copy volume using single drive.
242 - Copy volume with multiple driven (same or different block size).
243 - Add block size (min, max) to Vol label.
244 - Concept of VolumeSet during restore which is a list
245 of Volume names needed.
246 - Restore files modified after date
247 - Restore file modified before date
248 - Emergency restore info:
250 - Backup working directory
252 - Restore options (do not overwrite)
253 - Restore -- do nothing but show what would happend
254 - Authentication between SD and FD
255 - SET LD_RUN_PATH=$HOME/mysql/lib/mysql
256 - Send Volumes needed during restore to Console
257 - Put Job statistics in End Session Label (files saved,
258 total bytes, start time, ...).
259 - Put FileSet name in the SOS label.
260 - Implement Restore FileSet=
261 - Write a scanner for the UA (keyword, scan-routine, result, prompt).
262 - Create a protocol.h and protocol.c where all protocol messages
264 - If SD cannot open a drive, make it periodically retry.
265 - Put Bacula version somewhere in Job stream, probably Start Session
267 - Remove duplicate fields from jcr (e.g. jcr.level and
269 - Timout a job or terminate if link goes down, or reopen link and query.
270 - Define how we handle times to avoid problem with Unix dates (2049 ?).
271 - The daemons should know when one is already
272 running and refuse to run a second copy.
273 - Fill all fields in Vol/Job Header -- ensure that everything
274 needed is written to tape. Think about restore to Catalog
275 from tape. Client record needs improving.
276 - Find general solution for sscanf size problems (as well
277 as sprintf. Do at run time?
279 - Concept of precious tapes (cannot be reused).
280 - Allow FD to run from inetd ???
281 - Preprocessing command per file.
282 - Postprocessing command per file (when restoring).
284 - Restore should get Device and Pool information from
285 job record rather than from config.
286 - Make SD send attribute stream to DR but first
287 buffering to file, then sending only when the
288 files are written to tape.
289 - Autolabel should be specified by DR instead of SD.
290 - Ability to recreate the catalog from a tape.
291 - Find out how to get the system tape block limits, e.g.:
292 Apr 22 21:22:10 polymatou kernel: st1: Block limits 1 - 245760 bytes.
293 Apr 22 21:22:10 polymatou kernel: st0: Block limits 2 - 16777214 bytes.
296 - AutoScan (check checksum of tape)
297 - Format command = "format /dev/nst0"
301 - Seek resolution (usually corresponds to buffer size)
302 - EODErrorCode=ENOSPC or code
303 - Partial Read error code
304 - Partial write error code
305 - Nonformatted read error
306 - Nonformatted write error
307 - WriteProtected error
311 - IgnoreCloseErrors=yes
321 - Eliminate duplicate File records to shrink database.
322 - FD sends unsaved file list to Director at end of job.
323 - Implement InsertUniqueDB.
324 - Write a Storage daemon that uses pipes and
325 standard Unix programs to write to the tape.
327 - Need something that monitors the JCR queue and
328 times out jobs by asking the deamons where they are.
329 - Add daemon JCR JobId=0 to have a daemon context
333 - Client (list of clients to force client)
334 - Devices (list of devices to force device)
339 - Recycle from other pools: Yes, No
340 - Recycle to other pools: Yes, no
343 - Optional MediaType to force media?
354 if(connect(sockfd, (struct sockaddr * ) (& addr), sizeof(addr)) .lt. 0){
360 linger.l_linger = 60;
361 i = setsockopt(sockfd, SOL_SOCKET, SO_LINGER, (char *) &linger,
364 fl = fcntl(sockfd, F_GETFL);
365 fcntl(sockfd, F_SETFL, fl & (~ O_NONBLOCK) & (~ O_NDELAY));
367 - Add "0nnn" in front of all sscanf %s fields
368 to prevent field overflow.
370 What: jobid or file list
371 From: tape, file, ...
372 Where: original location, another path
373 How: Always replace, Replace if newer, Never replace
374 Report: files restored; files not restored; errors; warnings
376 - Enhance Jmsg code to permit buffering and saving to disk.
377 - Probably create a jcr with JobId=0 as a master
378 catchall if jcr not found or if operation involves
380 - device driver = "xxxx" for drives.
381 - restart: paranoid: read label fsf to
382 eom read append block, and go
383 super-paranoid: read label, read all files
384 in between, read append block, and go
385 verify: backspace, read append block, and go
386 permissive: same as above but frees drive
387 if tape is not valid.
389 - Ensure that /dev/null works
390 - File daemon should build list of files skipped, and then
391 at end of save retry and report any errors.
392 - Need report class for messages. Perhaps
393 report resource where report=group of messages
394 - Extract what=(session_id|file_list); where
396 - enhance scan_attrib and rename scan_jobtype, and
397 fill in code for "since" option
398 - dir_config: get rid of all printfs
399 - To buffer messages, we need associated jobid and Director name.
400 - Need to save contents of FileSet to tape?
401 - Director needs a time after which the report status is sent
402 anyway -- or better yet, a retry time for the job.
403 Don't reschedule a job if previous incarnation is still running.
404 - Figure out how to do a "full" restore from catalog
405 - Figure out how to save the catalog (possibly a special FileSet).
406 - Figure out how to restore the catalog.
407 - Figure out how to put a Volume into the catalog (from the tape)
408 - Figure out how to do a restore from a Volume
409 - Some way to automatically backup everything is needed????
410 - Need a structure for pending actions:
412 - termination status (part of buffered msgs?)
413 - Concept of grouping Storage devices and job can use
414 any of a number of devices
416 Read, Write, Clean, Delete
417 - Login to Bacula; Bacula users with different permissions:
419 - Tape recycle destination
420 - Job Schedule Status
424 - File daemon should pass Director the operating system info
425 to be stored in the Client Record (or verified that it has
427 - Store info on each file system type (probably in the job header on tape.
428 This could be the output of df; or perhaps some sort of /etc/mtab record.
431 - Use media 1 time (so that we can do 6 days of incremental
432 backups before switching to another tape) (already)
433 specify # times (jobs)
434 specify bytes (already)
435 specify time (seconds, hours, days)
436 - Implement FSM (File System Modules).
437 - Identify unchanged or "system" files and save them to a
438 special tape thus removing them from the standard
439 backup FileSet -- BASE backup.
440 - Turn virutally all sprintfs into snprintfs.
441 - Heartbeat between daemons.
442 - Audit M_ error codes to ensure they are correct and
444 - Add variable break characters to lex analyzer.
445 Either a bit mask or a string of chars so that
446 the caller can change the break characters.
447 - Make a single T_BREAK to replace T_COMMA, etc.
448 - Ensure that File daemon and Storage daemon can
449 continue a save if the Director goes down (this
450 is NOT currently the case). Must detect socket error,
451 buffer messages for later.
454 Done: (see kernsdone for more)
455 --the console script is broken as installed and has to be hand-massaged with
456 paths, config files etc.
457 - Termination status in FD for Verify = C -- incorrect.
458 - Implement alter_sqlite_tables
459 - Fix scheduler -- see "Hourly cycle". It doesn't do both each
460 hour, rather it alternates between 0:05 and 0:35.
461 - Create Counter DB records.