ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/BDB/BDB.pm
(Generate patch)

Comparing BDB/BDB.pm (file contents):
Revision 1.4 by root, Mon Feb 5 23:46:15 2007 UTC vs.
Revision 1.53 by root, Tue Oct 21 02:21:25 2008 UTC

4 4
5=head1 SYNOPSIS 5=head1 SYNOPSIS
6 6
7 use BDB; 7 use BDB;
8 8
9 my $env = db_env_create;
10
11 mkdir "bdtest", 0700;
12 db_env_open
13 $env,
14 "bdtest",
15 BDB::INIT_LOCK | BDB::INIT_LOG | BDB::INIT_MPOOL
16 | BDB::INIT_TXN | BDB::RECOVER | BDB::USE_ENVIRON | BDB::CREATE,
17 0600;
18
19 $env->set_flags (BDB::AUTO_COMMIT | BDB::TXN_NOSYNC, 1);
20
21 my $db = db_create $env;
22 db_open $db, undef, "table", undef, BDB::BTREE, BDB::AUTO_COMMIT | BDB::CREATE
23 | BDB::READ_UNCOMMITTED, 0600;
24 db_put $db, undef, "key", "data", 0, sub {
25 db_del $db, undef, "key";
26 };
27 db_sync $db;
28
29 # when you also use Coro, management is easy:
30 use Coro::BDB;
31
32 # automatic event loop intergration with AnyEvent:
33 use AnyEvent::BDB;
34
35 # automatic result processing with EV:
36 my $WATCHER = EV::io BDB::poll_fileno, EV::READ, \&BDB::poll_cb;
37
38 # with Glib:
39 add_watch Glib::IO BDB::poll_fileno,
40 in => sub { BDB::poll_cb; 1 };
41
42 # or simply flush manually
43 BDB::flush;
44
45
9=head1 DESCRIPTION 46=head1 DESCRIPTION
10 47
11=head2 EXAMPLE 48See the BerkeleyDB documentation (L<http://www.oracle.com/technology/documentation/berkeley-db/db/index.html>).
49The BDB API is very similar to the C API (the translation has been very faithful).
50
51See also the example sections in the document below and possibly the eg/
52subdirectory of the BDB distribution. Last not least see the IO::AIO
53documentation, as that module uses almost the same asynchronous request
54model as this module.
55
56I know this is woefully inadequate documentation. Send a patch!
57
12 58
13=head1 REQUEST ANATOMY AND LIFETIME 59=head1 REQUEST ANATOMY AND LIFETIME
14 60
15Every request method creates a request. which is a C data structure not 61Every request method creates a request. which is a C data structure not
16directly visible to Perl. 62directly visible to Perl.
62no warnings; 108no warnings;
63use strict 'vars'; 109use strict 'vars';
64 110
65use base 'Exporter'; 111use base 'Exporter';
66 112
113our $VERSION;
114
67BEGIN { 115BEGIN {
68 our $VERSION = '0.1'; 116 $VERSION = '1.81';
69 117
70 our @BDB_REQ = qw( 118 our @BDB_REQ = qw(
71 db_env_create db_env_open db_env_close 119 db_env_open db_env_close db_env_txn_checkpoint db_env_lock_detect
120 db_env_memp_sync db_env_memp_trickle db_env_dbrename db_env_dbremove
121 db_env_log_archive
72 db_create db_open db_close db_compact db_sync db_put db_get db_pget 122 db_open db_close db_compact db_sync db_upgrade
123 db_put db_exists db_get db_pget db_del db_key_range
73 db_txn_commit db_txn_abort 124 db_txn_commit db_txn_abort db_txn_finish
125 db_c_close db_c_count db_c_put db_c_get db_c_pget db_c_del
126 db_sequence_open db_sequence_close
127 db_sequence_get db_sequence_remove
74 ); 128 );
75 our @EXPORT = (@BDB_REQ, qw(dbreq_pri dbreq_nice)); 129 our @EXPORT = (@BDB_REQ, qw(dbreq_pri dbreq_nice db_env_create db_create));
130 our @EXPORT_OK = qw(
76 our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush 131 poll_fileno poll_cb poll_wait flush
77 min_parallel max_parallel max_idle 132 min_parallel max_parallel max_idle
78 nreqs nready npending nthreads 133 nreqs nready npending nthreads
79 max_poll_time max_poll_reqs); 134 max_poll_time max_poll_reqs
135 );
80 136
81 require XSLoader; 137 require XSLoader;
82 XSLoader::load ("BDB", $VERSION); 138 XSLoader::load ("BDB", $VERSION);
83} 139}
84 140
141=head1 BERKELEYDB FUNCTIONS
142
143All of these are functions. The create functions simply return a new
144object and never block. All the remaining functions take an optional
145callback as last argument. If it is missing, then the function will be
146executed synchronously. In both cases, C<$!> will reflect the return value
147of the function.
148
149BDB functions that cannot block (mostly functions that manipulate
150settings) are method calls on the relevant objects, so the rule of thumb
151is: if it's a method, it's not blocking, if it's a function, it takes a
152callback as last argument.
153
154In the following, C<$int> signifies an integer return value,
155C<bdb_filename> is a "filename" (octets on unix, madness on windows),
156C<U32> is an unsigned 32 bit integer, C<int> is some integer, C<NV> is a
157floating point value.
158
159Most C<SV *> types are generic perl scalars (for input and output of data
160values).
161
162The various C<DB_ENV> etc. arguments are handles return by
163C<db_env_create>, C<db_create>, C<txn_begin> and so on. If they have an
164appended C<_ornull> this means they are optional and you can pass C<undef>
165for them, resulting a NULL pointer on the C level.
166
167The C<SV *callback> is the optional callback function to call when the
168request is completed. This last callback argument is special: the callback
169is simply the last argument passed. If there are "optional" arguments
170before the callback they can be left out. The callback itself can be left
171out or specified as C<undef>, in which case the function will be executed
172synchronously.
173
174For example, C<db_env_txn_checkpoint> usually is called with all integer
175arguments zero. These can be left out, so all of these specify a call
176to C<< DB_ENV->txn_checkpoint >>, to be executed asynchronously with a
177callback to be called:
178
179 db_env_txn_checkpoint $db_env, 0, 0, 0, sub { };
180 db_env_txn_checkpoint $db_env, 0, 0, sub { };
181 db_env_txn_checkpoint $db_env, sub { };
182
183While these all specify a call to C<< DB_ENV->txn_checkpoint >> to be
184executed synchronously:
185
186 db_env_txn_checkpoint $db_env, 0, 0, 0, undef;
187 db_env_txn_checkpoint $db_env, 0, 0, 0;
188 db_env_txn_checkpoint $db_env, 0;
189
190=head2 BDB functions
191
192Functions in the BDB namespace, exported by default:
193
194 $env = db_env_create (U32 env_flags = 0)
195 flags: RPCCLIENT
196
197 db_env_open (DB_ENV *env, bdb_filename db_home, U32 open_flags, int mode, SV *callback = &PL_sv_undef)
198 open_flags: INIT_CDB INIT_LOCK INIT_LOG INIT_MPOOL INIT_REP INIT_TXN RECOVER RECOVER_FATAL USE_ENVIRON USE_ENVIRON_ROOT CREATE LOCKDOWN PRIVATE REGISTER SYSTEM_MEM
199 db_env_close (DB_ENV *env, U32 flags = 0, SV *callback = &PL_sv_undef)
200 db_env_txn_checkpoint (DB_ENV *env, U32 kbyte = 0, U32 min = 0, U32 flags = 0, SV *callback = &PL_sv_undef)
201 flags: FORCE
202 db_env_lock_detect (DB_ENV *env, U32 flags = 0, U32 atype = DB_LOCK_DEFAULT, SV *dummy = 0, SV *callback = &PL_sv_undef)
203 atype: LOCK_DEFAULT LOCK_EXPIRE LOCK_MAXLOCKS LOCK_MAXWRITE LOCK_MINLOCKS LOCK_MINWRITE LOCK_OLDEST LOCK_RANDOM LOCK_YOUNGEST
204 db_env_memp_sync (DB_ENV *env, SV *dummy = 0, SV *callback = &PL_sv_undef)
205 db_env_memp_trickle (DB_ENV *env, int percent, SV *dummy = 0, SV *callback = &PL_sv_undef)
206 db_env_dbremove (DB_ENV *env, DB_TXN_ornull *txnid, bdb_filename file, bdb_filename database, U32 flags = 0, SV *callback = &PL_sv_undef)
207 db_env_dbrename (DB_ENV *env, DB_TXN_ornull *txnid, bdb_filename file, bdb_filename database, bdb_filename newname, U32 flags = 0, SV *callback = &PL_sv_undef)
208 db_env_log_archive (DB_ENV *env, SV *listp, U32 flags = 0, SV *callback = &PL_sv_undef)
209
210 $db = db_create (DB_ENV *env = 0, U32 flags = 0)
211 flags: XA_CREATE
212
213 db_open (DB *db, DB_TXN_ornull *txnid, bdb_filename file, bdb_filename database, int type, U32 flags, int mode, SV *callback = &PL_sv_undef)
214 flags: AUTO_COMMIT CREATE EXCL MULTIVERSION NOMMAP RDONLY READ_UNCOMMITTED THREAD TRUNCATE
215 db_close (DB *db, U32 flags = 0, SV *callback = &PL_sv_undef)
216 flags: DB_NOSYNC
217 db_upgrade (DB *db, bdb_filename file, U32 flags = 0, SV *callback = &PL_sv_undef)
218 db_compact (DB *db, DB_TXN_ornull *txn = 0, SV *start = 0, SV *stop = 0, SV *unused1 = 0, U32 flags = DB_FREE_SPACE, SV *unused2 = 0, SV *callback = &PL_sv_undef)
219 flags: FREELIST_ONLY FREE_SPACE
220 db_sync (DB *db, U32 flags = 0, SV *callback = &PL_sv_undef)
221 db_key_range (DB *db, DB_TXN_ornull *txn, SV *key, SV *key_range, U32 flags = 0, SV *callback = &PL_sv_undef)
222 db_put (DB *db, DB_TXN_ornull *txn, SV *key, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef)
223 flags: APPEND NODUPDATA NOOVERWRITE
224 db_exists (DB *db, DB_TXN_ornull *txn, SV *key, U32 flags = 0, SV *callback = 0) (v4.6)
225 db_get (DB *db, DB_TXN_ornull *txn, SV *key, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef)
226 flags: CONSUME CONSUME_WAIT GET_BOTH SET_RECNO MULTIPLE READ_COMMITTED READ_UNCOMMITTED RMW
227 db_pget (DB *db, DB_TXN_ornull *txn, SV *key, SV *pkey, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef)
228 flags: CONSUME CONSUME_WAIT GET_BOTH SET_RECNO MULTIPLE READ_COMMITTED READ_UNCOMMITTED RMW
229 db_del (DB *db, DB_TXN_ornull *txn, SV *key, U32 flags = 0, SV *callback = &PL_sv_undef)
230 db_txn_commit (DB_TXN *txn, U32 flags = 0, SV *callback = &PL_sv_undef)
231 flags: TXN_NOSYNC TXN_SYNC
232 db_txn_abort (DB_TXN *txn, SV *callback = &PL_sv_undef)
233
234 db_c_close (DBC *dbc, SV *callback = &PL_sv_undef)
235 db_c_count (DBC *dbc, SV *count, U32 flags = 0, SV *callback = &PL_sv_undef)
236 db_c_put (DBC *dbc, SV *key, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef)
237 flags: AFTER BEFORE CURRENT KEYFIRST KEYLAST NODUPDATA
238 db_c_get (DBC *dbc, SV *key, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef)
239 flags: CURRENT FIRST GET_BOTH GET_BOTH_RANGE GET_RECNO JOIN_ITEM LAST NEXT NEXT_DUP NEXT_NODUP PREV PREV_DUP PREV_NODUP SET SET_RANGE SET_RECNO READ_UNCOMMITTED MULTIPLE MULTIPLE_KEY RMW
240 db_c_pget (DBC *dbc, SV *key, SV *pkey, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef)
241 db_c_del (DBC *dbc, U32 flags = 0, SV *callback = &PL_sv_undef)
242
243 db_sequence_open (DB_SEQUENCE *seq, DB_TXN_ornull *txnid, SV *key, U32 flags = 0, SV *callback = &PL_sv_undef)
244 flags: CREATE EXCL
245 db_sequence_close (DB_SEQUENCE *seq, U32 flags = 0, SV *callback = &PL_sv_undef)
246 db_sequence_get (DB_SEQUENCE *seq, DB_TXN_ornull *txnid, int delta, SV *seq_value, U32 flags = DB_TXN_NOSYNC, SV *callback = &PL_sv_undef)
247 flags: TXN_NOSYNC
248 db_sequence_remove (DB_SEQUENCE *seq, DB_TXN_ornull *txnid = 0, U32 flags = 0, SV *callback = &PL_sv_undef)
249 flags: TXN_NOSYNC
250
251=head3 db_txn_finish (DB_TXN *txn, U32 flags = 0, SV *callback = &PL_sv_undef)
252
253This is not actually a Berkeley DB function but a BDB module
254extension. The background for this exytension is: It is very annoying to
255have to check every single BDB function for error returns and provide a
256codepath out of your transaction. While the BDB module still makes this
257possible, it contains the following extensions:
258
259When a transaction-protected function returns any operating system
260error (errno > 0), BDB will set the C<TXN_DEADLOCK> flag on the
261transaction. This flag is also set by Berkeley DB functions themselves
262when an operation fails with LOCK_DEADLOCK, and it causes all further
263operations on that transaction (including C<db_txn_commit>) to fail.
264
265The C<db_txn_finish> request will look at this flag, and, if it is set,
266will automatically call C<db_txn_abort> (setting errno to C<LOCK_DEADLOCK>
267if it isn't set to something else yet). If it isn't set, it will call
268C<db_txn_commit> and return the error normally.
269
270How to use this? Easy: just write your transaction normally:
271
272 my $txn = $db_env->txn_begin;
273 db_get $db, $txn, "key", my $data;
274 db_put $db, $txn, "key", $data + 1 unless $! == BDB::NOTFOUND;
275 db_txn_finish $txn;
276 die "transaction failed" if $!;
277
278That is, handle only the expected errors. If something unexpected happens
279(EIO, LOCK_NOTGRANTED or a deadlock in either db_get or db_put), then the remaining
280requests (db_put in this case) will simply be skipped (they will fail with
281LOCK_DEADLOCK) and the transaction will be aborted.
282
283You can use the C<< $txn->failed >> method to check wether a transaction
284has failed in this way and abort further processing (excluding
285C<db_txn_finish>).
286
287=head2 DB_ENV/database environment methods
288
289Methods available on DB_ENV/$env handles:
290
291 DESTROY (DB_ENV_ornull *env)
292 CODE:
293 if (env)
294 env->close (env, 0);
295
296 $int = $env->set_data_dir (const char *dir)
297 $int = $env->set_tmp_dir (const char *dir)
298 $int = $env->set_lg_dir (const char *dir)
299 $int = $env->set_shm_key (long shm_key)
300 $int = $env->set_cachesize (U32 gbytes, U32 bytes, int ncache = 0)
301 $int = $env->set_flags (U32 flags, int onoff = 1)
302 $int = $env->log_set_config (U32 flags, int onoff = 1) (v4.7)
303 $int = $env->set_intermediate_dir_mode (const char *modestring) (v4.7)
304 $env->set_errfile (FILE *errfile = 0)
305 $env->set_msgfile (FILE *msgfile = 0)
306 $int = $env->set_verbose (U32 which, int onoff = 1)
307 $int = $env->set_encrypt (const char *password, U32 flags = 0)
308 $int = $env->set_timeout (NV timeout_seconds, U32 flags = SET_TXN_TIMEOUT)
309 $int = $env->set_mp_max_openfd (int maxopenfd);
310 $int = $env->set_mp_max_write (int maxwrite, int maxwrite_sleep);
311 $int = $env->set_mp_mmapsize (int mmapsize_mb)
312 $int = $env->set_lk_detect (U32 detect = DB_LOCK_DEFAULT)
313 $int = $env->set_lk_max_lockers (U32 max)
314 $int = $env->set_lk_max_locks (U32 max)
315 $int = $env->set_lk_max_objects (U32 max)
316 $int = $env->set_lg_bsize (U32 max)
317 $int = $env->set_lg_max (U32 max)
318 $int = $env->mutex_set_increment (U32 increment)
319 $int = $env->mutex_set_tas_spins (U32 tas_spins)
320 $int = $env->mutex_set_max (U32 max)
321 $int = $env->mutex_set_align (U32 align)
322
323 $txn = $env->txn_begin (DB_TXN_ornull *parent = 0, U32 flags = 0)
324 flags: READ_COMMITTED READ_UNCOMMITTED TXN_NOSYNC TXN_NOWAIT TXN_SNAPSHOT TXN_SYNC TXN_WAIT TXN_WRITE_NOSYNC
325 $txn = $env->cdsgroup_begin; (v4.5)
326
327=head3 Example:
328
329 use AnyEvent;
330 use BDB;
331
332 our $FH; open $FH, "<&=" . BDB::poll_fileno;
333 our $WATCHER = AnyEvent->io (fh => $FH, poll => 'r', cb => \&BDB::poll_cb);
334
335 BDB::min_parallel 8;
336
337 my $env = db_env_create;
338
339 mkdir "bdtest", 0700;
340 db_env_open
341 $env,
342 "bdtest",
343 BDB::INIT_LOCK | BDB::INIT_LOG | BDB::INIT_MPOOL | BDB::INIT_TXN | BDB::RECOVER | BDB::USE_ENVIRON | BDB::CREATE,
344 0600;
345
346 $env->set_flags (BDB::AUTO_COMMIT | BDB::TXN_NOSYNC, 1);
347
348
349=head2 DB/database methods
350
351Methods available on DB/$db handles:
352
353 DESTROY (DB_ornull *db)
354 CODE:
355 if (db)
356 {
357 SV *env = (SV *)db->app_private;
358 db->close (db, 0);
359 SvREFCNT_dec (env);
360 }
361
362 $int = $db->set_cachesize (U32 gbytes, U32 bytes, int ncache = 0)
363 $int = $db->set_flags (U32 flags)
364 flags: CHKSUM ENCRYPT TXN_NOT_DURABLE
365 Btree: DUP DUPSORT RECNUM REVSPLITOFF
366 Hash: DUP DUPSORT
367 Queue: INORDER
368 Recno: RENUMBER SNAPSHOT
369
370 $int = $db->set_encrypt (const char *password, U32 flags)
371 $int = $db->set_lorder (int lorder)
372 $int = $db->set_bt_minkey (U32 minkey)
373 $int = $db->set_re_delim (int delim)
374 $int = $db->set_re_pad (int re_pad)
375 $int = $db->set_re_source (char *source)
376 $int = $db->set_re_len (U32 re_len)
377 $int = $db->set_h_ffactor (U32 h_ffactor)
378 $int = $db->set_h_nelem (U32 h_nelem)
379 $int = $db->set_q_extentsize (U32 extentsize)
380
381 $dbc = $db->cursor (DB_TXN_ornull *txn = 0, U32 flags = 0)
382 flags: READ_COMMITTED READ_UNCOMMITTED WRITECURSOR TXN_SNAPSHOT
383 $seq = $db->sequence (U32 flags = 0)
384
385=head3 Example:
386
387 my $db = db_create $env;
388 db_open $db, undef, "table", undef, BDB::BTREE, BDB::AUTO_COMMIT | BDB::CREATE | BDB::READ_UNCOMMITTED, 0600;
389
390 for (1..1000) {
391 db_put $db, undef, "key $_", "data $_";
392
393 db_key_range $db, undef, "key $_", my $keyrange;
394 my ($lt, $eq, $gt) = @$keyrange;
395 }
396
397 db_del $db, undef, "key $_" for 1..1000;
398
399 db_sync $db;
400
401
402=head2 DB_TXN/transaction methods
403
404Methods available on DB_TXN/$txn handles:
405
406 DESTROY (DB_TXN_ornull *txn)
407 CODE:
408 if (txn)
409 txn->abort (txn);
410
411 $int = $txn->set_timeout (NV timeout_seconds, U32 flags = SET_TXN_TIMEOUT)
412 flags: SET_LOCK_TIMEOUT SET_TXN_TIMEOUT
413
414 $bool = $txn->failed
415 # see db_txn_finish documentation, above
416
417
418=head2 DBC/cursor methods
419
420Methods available on DBC/$dbc handles:
421
422 DESTROY (DBC_ornull *dbc)
423 CODE:
424 if (dbc)
425 dbc->c_close (dbc);
426
427 $int = $cursor->set_priority ($priority = PRIORITY_*) (v4.6)
428
429=head3 Example:
430
431 my $c = $db->cursor;
432
433 for (;;) {
434 db_c_get $c, my $key, my $data, BDB::NEXT;
435 warn "<$!,$key,$data>";
436 last if $!;
437 }
438
439 db_c_close $c;
440
441
442=head2 DB_SEQUENCE/sequence methods
443
444Methods available on DB_SEQUENCE/$seq handles:
445
446 DESTROY (DB_SEQUENCE_ornull *seq)
447 CODE:
448 if (seq)
449 seq->close (seq, 0);
450
451 $int = $seq->initial_value (db_seq_t value)
452 $int = $seq->set_cachesize (U32 size)
453 $int = $seq->set_flags (U32 flags)
454 flags: SEQ_DEC SEQ_INC SEQ_WRAP
455 $int = $seq->set_range (db_seq_t min, db_seq_t max)
456
457=head3 Example:
458
459 my $seq = $db->sequence;
460
461 db_sequence_open $seq, undef, "seq", BDB::CREATE;
462 db_sequence_get $seq, undef, 1, my $value;
463
464
85=head2 SUPPORT FUNCTIONS 465=head1 SUPPORT FUNCTIONS
86 466
87=head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION 467=head2 EVENT PROCESSING AND EVENT LOOP INTEGRATION
88 468
89=over 4 469=over 4
470
471=item $msg = BDB::strerror [$errno]
472
473Returns the string corresponding to the given errno value. If no argument
474is given, use C<$!>.
475
476Note that the BDB module also patches the C<$!> variable directly, so you
477should be able to get a bdb error string by simply stringifying C<$!>.
90 478
91=item $fileno = BDB::poll_fileno 479=item $fileno = BDB::poll_fileno
92 480
93Return the I<request result pipe file descriptor>. This filehandle must be 481Return the I<request result pipe file descriptor>. This filehandle must be
94polled for reading by some mechanism outside this module (e.g. Event or 482polled for reading by some mechanism outside this module (e.g. Event or
133interactiveness when perl is not fast enough to process all requests in 521interactiveness when perl is not fast enough to process all requests in
134time. 522time.
135 523
136For interactive programs, values such as C<0.01> to C<0.1> should be fine. 524For interactive programs, values such as C<0.01> to C<0.1> should be fine.
137 525
138Example: Install an Event watcher that automatically calls 526Example: Install an EV watcher that automatically calls
139BDB::poll_cb with low priority, to ensure that other parts of the 527BDB::poll_cb with low priority, to ensure that other parts of the
140program get the CPU sometimes even under high AIO load. 528program get the CPU sometimes even under high load.
141 529
142 # try not to spend much more than 0.1s in poll_cb 530 # try not to spend much more than 0.1s in poll_cb
143 BDB::max_poll_time 0.1; 531 BDB::max_poll_time 0.1;
144 532
145 # use a low priority so other tasks have priority 533 my $bdb_poll = EV::io BDB::poll_fileno, EV::READ, \&BDB::poll_cb);
146 Event->io (fd => BDB::poll_fileno,
147 poll => 'r', nice => 1,
148 cb => &BDB::poll_cb);
149 534
150=item BDB::poll_wait 535=item BDB::poll_wait
151 536
152If there are any outstanding requests and none of them in the result 537If there are any outstanding requests and none of them in the result
153phase, wait till the result filehandle becomes ready for reading (simply 538phase, wait till the result filehandle becomes ready for reading (simply
165 550
166 BDB::poll_wait, BDB::poll_cb 551 BDB::poll_wait, BDB::poll_cb
167 552
168=item BDB::flush 553=item BDB::flush
169 554
170Wait till all outstanding AIO requests have been handled. 555Wait till all outstanding BDB requests have been handled.
171 556
172Strictly equivalent to: 557Strictly equivalent to:
173 558
174 BDB::poll_wait, BDB::poll_cb 559 BDB::poll_wait, BDB::poll_cb
175 while BDB::nreqs; 560 while BDB::nreqs;
176 561
562=back
563
564=head2 VERSION CHECKING
565
566BerkeleyDB comes in various versions, many of them have minor
567incompatibilities. This means that traditional "at least version x.x"
568checks are often not sufficient.
569
570Example: set the log_autoremove option in a way compatible with <v4.7 and
571v4.7. Note the use of & on the constants to avoid triggering a compiletime
572bug when the symbol isn't available.
573
574 $DB_ENV->set_flags (&BDB::LOG_AUTOREMOVE ) if BDB::VERSION v0, v4.7;
575 $DB_ENV->log_set_config (&BDB::LOG_AUTO_REMOVE) if BDB::VERSION v4.7;
576
577=over 4
578
579=item BDB::VERSION
580
581The C<BDB::VERSION> function, when called without arguments, returns the
582Berkeley DB version as a v-string (usually with 3 components). You should
583use C<lt> and C<ge> operators exclusively to make comparisons.
584
585Example: check for at least version 4.7.
586
587 BDB::VERSION ge v4.7 or die;
588
589=item BDB::VERSION min-version
590
591Returns true if the BDB version is at least the given version (specified
592as a v-string), false otherwise.
593
594Example: check for at least version 4.5.
595
596 BDB::VERSION v4.7 or die;
597
598=item BDB::VERSION min-version, max-version
599
600Returns true of the BDB version is at least version C<min-version> (specify C<undef> or C<v0> for any minimum version)
601and less then C<max-version>.
602
603Example: check wether version is strictly less then v4.7.
604
605 BDB::VERSION v0, v4.7
606 or die "version 4.7 is not yet supported";
607
608=back
609
610=cut
611
612sub VERSION {
613 # I was dumb enough to override the VERSION method here, so let's try
614 # to fix it up.
615
616 if ($_[0] eq __PACKAGE__) {
617 $VERSION
618 } else {
619 if (@_ > 0) {
620 return undef if VERSION_v lt $_[0];
621 if (@_ > 1) {
622 return undef if VERSION_v ge $_[1];
623 }
624 }
625
626 VERSION_v
627 }
628}
629
177=head3 CONTROLLING THE NUMBER OF THREADS 630=head2 CONTROLLING THE NUMBER OF THREADS
631
632=over 4
178 633
179=item BDB::min_parallel $nthreads 634=item BDB::min_parallel $nthreads
180 635
181Set the minimum number of AIO threads to C<$nthreads>. The current 636Set the minimum number of BDB threads to C<$nthreads>. The current
182default is C<8>, which means eight asynchronous operations can execute 637default is C<8>, which means eight asynchronous operations can execute
183concurrently at any one time (the number of outstanding requests, 638concurrently at any one time (the number of outstanding requests,
184however, is unlimited). 639however, is unlimited).
185 640
186BDB starts threads only on demand, when an AIO request is queued and 641BDB starts threads only on demand, when an BDB request is queued and
187no free thread exists. Please note that queueing up a hundred requests can 642no free thread exists. Please note that queueing up a hundred requests can
188create demand for a hundred threads, even if it turns out that everything 643create demand for a hundred threads, even if it turns out that everything
189is in the cache and could have been processed faster by a single thread. 644is in the cache and could have been processed faster by a single thread.
190 645
191It is recommended to keep the number of threads relatively low, as some 646It is recommended to keep the number of threads relatively low, as some
196Under most circumstances you don't need to call this function, as the 651Under most circumstances you don't need to call this function, as the
197module selects a default that is suitable for low to moderate load. 652module selects a default that is suitable for low to moderate load.
198 653
199=item BDB::max_parallel $nthreads 654=item BDB::max_parallel $nthreads
200 655
201Sets the maximum number of AIO threads to C<$nthreads>. If more than the 656Sets the maximum number of BDB threads to C<$nthreads>. If more than the
202specified number of threads are currently running, this function kills 657specified number of threads are currently running, this function kills
203them. This function blocks until the limit is reached. 658them. This function blocks until the limit is reached.
204 659
205While C<$nthreads> are zero, aio requests get queued but not executed 660While C<$nthreads> are zero, aio requests get queued but not executed
206until the number of threads has been increased again. 661until the number of threads has been increased again.
241 696
242You can still queue as many requests as you want. Therefore, 697You can still queue as many requests as you want. Therefore,
243C<max_oustsanding> is mainly useful in simple scripts (with low values) or 698C<max_oustsanding> is mainly useful in simple scripts (with low values) or
244as a stop gap to shield against fatal memory overflow (with large values). 699as a stop gap to shield against fatal memory overflow (with large values).
245 700
246=item BDB::set_sync_prepare $cb 701=item $old_cb = BDB::set_sync_prepare $cb
247 702
248Sets a callback that is called whenever a request is created without an 703Sets a callback that is called whenever a request is created without an
249explicit callback. It has to return two code references. The first is used 704explicit callback. It has to return two code references. The first is used
250as the request callback, and the second is called to wait until the first 705as the request callback (it should save the return status), and the second
251callback has been called. The default implementation works like this: 706is called to wait until the first callback has been called (it must set
707C<$!> to the return status).
708
709This mechanism can be used to include BDB into other event mechanisms,
710such as L<Coro::BDB>.
711
712To allow other, callback-based, events to be executed while callback-less
713ones are run, you could use this sync prepare function:
252 714
253 sub { 715 sub {
254 my $status; 716 my $status;
255 ( 717 (
256 sub { $status = $! }, 718 sub { $status = $! },
257 sub { BDB::poll while !defined $status; $! = $status }, 719 sub { BDB::poll while !defined $status; $! = $status },
258 ) 720 )
259 } 721 }
260 722
723It works by polling for results till the request has finished and then
724sets C<$!> to the return value. This means that if you don't use a
725callback, BDB would simply fall back to synchronous operations.
726
727By default, or if the sync prepare function is set to C<undef>, is to
728execute callback-less BDB requests in the foreground thread, setting C<$!>
729to the return value, without polling for other events.
730
261=back 731=back
262 732
263=head3 STATISTICAL INFORMATION 733=head2 STATISTICAL INFORMATION
264 734
265=over 4 735=over 4
266 736
267=item BDB::nreqs 737=item BDB::nreqs
268 738
286 756
287=back 757=back
288 758
289=cut 759=cut
290 760
291set_sync_prepare { 761set_sync_prepare (undef);
292 my $status;
293 (
294 sub {
295 $status = $!;
296 },
297 sub {
298 BDB::poll while !defined $status;
299 $! = $status;
300 },
301 )
302};
303 762
304min_parallel 8; 763min_parallel 8;
305 764
306END { flush } 765END { flush }
307 766
3081; 7671;
309 768
769=head1 COMMON PITFALLS
770
771=head2 Unexpected Crashes
772
773Remember that, by default, BDB will execute requests in parallel, in
774somewhat random order. That means that it is easy to run a C<db_get>
775request on thesa me database as a concurrent C<db_close> request, leading
776to a crash, silent data corruption, eventually the next world war on
777terrorism.
778
779If you only ever use foreground requests (without a callback), this will
780not be an issue.
781
782=head2 Unexpected Freezes or Deadlocks
783
784Remember that, by default, BDB will execute requests in parallel, which
785easily leads to deadlocks (even concurrent put's on the same database can
786deadlock).
787
788You either need to run deadlock detection (and handle the resulting
789errors), or make sure only one process ever updates the database, ine one
790thread, e.g. by using only foreground requests (without a callback).
791
310=head2 FORK BEHAVIOUR 792=head1 FORK BEHAVIOUR
311 793
312This module should do "the right thing" when the process using it forks: 794This module should do "the right thing" when the process using it forks:
313 795
314Before the fork, IO::AIO enters a quiescent state where no requests 796Before the fork, BDB enters a quiescent state where no requests
315can be added in other threads and no results will be processed. After 797can be added in other threads and no results will be processed. After
316the fork the parent simply leaves the quiescent state and continues 798the fork the parent simply leaves the quiescent state and continues
317request/result processing, while the child frees the request/result queue 799request/result processing, while the child frees the request/result queue
318(so that the requests started before the fork will only be handled in the 800(so that the requests started before the fork will only be handled in the
319parent). Threads will be started on demand until the limit set in the 801parent). Threads will be started on demand until the limit set in the
320parent process has been reached again. 802parent process has been reached again.
321 803
322In short: the parent will, after a short pause, continue as if fork had 804In short: the parent will, after a short pause, continue as if fork had
323not been called, while the child will act as if IO::AIO has not been used 805not been called, while the child will act as if BDB has not been used
324yet. 806yet.
325 807
808Win32 note: there is no fork on win32, and perls emulation of it is too
809broken to be supported, so do not use BDB in a windows pseudo-fork, better
810yet, switch to a more capable platform.
811
326=head2 MEMORY USAGE 812=head1 MEMORY USAGE
327 813
328Per-request usage: 814Per-request usage:
329 815
330Each aio request uses - depending on your architecture - around 100-200 816Each aio request uses - depending on your architecture - around 100-200
331bytes of memory. In addition, stat requests need a stat buffer (possibly 817bytes of memory. In addition, stat requests need a stat buffer (possibly
332a few hundred bytes), readdir requires a result buffer and so on. Perl 818a few hundred bytes), readdir requires a result buffer and so on. Perl
333scalars and other data passed into aio requests will also be locked and 819scalars and other data passed into aio requests will also be locked and
334will consume memory till the request has entered the done state. 820will consume memory till the request has entered the done state.
335 821
336This is now awfully much, so queuing lots of requests is not usually a 822This is not awfully much, so queuing lots of requests is not usually a
337problem. 823problem.
338 824
339Per-thread usage: 825Per-thread usage:
340 826
341In the execution phase, some aio requests require more memory for 827In the execution phase, some aio requests require more memory for
342temporary buffers, and each thread requires a stack and other data 828temporary buffers, and each thread requires a stack and other data
343structures (usually around 16k-128k, depending on the OS). 829structures (usually around 16k-128k, depending on the OS).
344 830
831=head1 WIN32 FILENAMES/DATABASE NAME MESS
832
833Perl on Win32 supports only ASCII filenames (the reason is that it abuses
834an internal flag to store wether a filename is Unicode or ANSI, but that
835flag is used for somethign else in the perl core, so there is no way to
836detect wether a filename is ANSI or Unicode-encoded). The BDB module
837tries to work around this issue by assuming that the filename is an ANSI
838filename and BDB was built for unicode support.
839
345=head1 KNOWN BUGS 840=head1 KNOWN BUGS
346 841
347Known bugs will be fixed in the next release. 842Known bugs will be fixed in the next release, except:
843
844 If you use a transaction in any request, and the request returns
845 with an operating system error or DB_LOCK_NOTGRANTED, the internal
846 TXN_DEADLOCK flag will be set on the transaction. See C<db_txn_finish>,
847 above.
348 848
349=head1 SEE ALSO 849=head1 SEE ALSO
350 850
351L<Coro::AIO>. 851L<AnyEvent::BDB> (event loop integration), L<Coro::BDB> (more natural
852syntax), L<IO::AIO> (nice to have).
352 853
353=head1 AUTHOR 854=head1 AUTHOR
354 855
355 Marc Lehmann <schmorp@schmorp.de> 856 Marc Lehmann <schmorp@schmorp.de>
356 http://home.schmorp.de/ 857 http://home.schmorp.de/

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines