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

Comparing BDB/BDB.pm (file contents):
Revision 1.8 by root, Wed May 9 06:42:23 2007 UTC vs.
Revision 1.51 by root, Mon Sep 29 02:59:08 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
11See the eg/ directory in the distribution and the berkeleydb C 48See the BerkeleyDB documentation (L<http://www.oracle.com/technology/documentation/berkeley-db/db/index.html>).
12documentation. This is inadequate, but the only sources of documentation 49The BDB API is very similar to the C API (the translation has been very faithful).
13known for this module so far.
14 50
15=head2 EXAMPLE 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
16 58
17=head1 REQUEST ANATOMY AND LIFETIME 59=head1 REQUEST ANATOMY AND LIFETIME
18 60
19Every 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
20directly visible to Perl. 62directly visible to Perl.
66no warnings; 108no warnings;
67use strict 'vars'; 109use strict 'vars';
68 110
69use base 'Exporter'; 111use base 'Exporter';
70 112
113our $VERSION;
114
71BEGIN { 115BEGIN {
72 our $VERSION = '0.1'; 116 $VERSION = '1.801';
73 117
74 our @BDB_REQ = qw( 118 our @BDB_REQ = qw(
75 db_env_open db_env_close db_env_txn_checkpoint db_env_lock_detect 119 db_env_open db_env_close db_env_txn_checkpoint db_env_lock_detect
76 db_env_memp_sync db_env_memp_trickle 120 db_env_memp_sync db_env_memp_trickle db_env_dbrename db_env_dbremove
77 db_open db_close db_compact db_sync db_put db_get db_pget db_del db_key_range 121 db_open db_close db_compact db_sync db_upgrade
122 db_put db_exists db_get db_pget db_del db_key_range
78 db_txn_commit db_txn_abort 123 db_txn_commit db_txn_abort db_txn_finish
79 db_c_close db_c_count db_c_put db_c_get db_c_pget db_c_del 124 db_c_close db_c_count db_c_put db_c_get db_c_pget db_c_del
80 db_sequence_open db_sequence_close 125 db_sequence_open db_sequence_close
81 db_sequence_get db_sequence_remove 126 db_sequence_get db_sequence_remove
82 ); 127 );
83 our @EXPORT = (@BDB_REQ, qw(dbreq_pri dbreq_nice db_env_create db_create)); 128 our @EXPORT = (@BDB_REQ, qw(dbreq_pri dbreq_nice db_env_create db_create));
90 135
91 require XSLoader; 136 require XSLoader;
92 XSLoader::load ("BDB", $VERSION); 137 XSLoader::load ("BDB", $VERSION);
93} 138}
94 139
140=head2 WIN32 FILENAMES/DATABASE NAME MESS
141
142Perl on Win32 supports only ASCII filenames (the reason is that it abuses
143an internal flag to store wether a filename is Unicode or ANSI, but that
144flag is used for somethign else in the perl core, so there is no way to
145detect wether a filename is ANSI or Unicode-encoded). The BDB module
146tries to work around this issue by assuming that the filename is an ANSI
147filename and BDB was built for unicode support.
148
149=head2 BERKELEYDB FUNCTIONS
150
151All of these are functions. The create functions simply return a new
152object and never block. All the remaining functions take an optional
153callback as last argument. If it is missing, then the function will be
154executed synchronously. In both cases, C<$!> will reflect the return value
155of the function.
156
157BDB functions that cannot block (mostly functions that manipulate
158settings) are method calls on the relevant objects, so the rule of thumb
159is: if it's a method, it's not blocking, if it's a function, it takes a
160callback as last argument.
161
162In the following, C<$int> signifies an integer return value,
163C<bdb_filename> is a "filename" (octets on unix, madness on windows),
164C<U32> is an unsigned 32 bit integer, C<int> is some integer, C<NV> is a
165floating point value.
166
167Most C<SV *> types are generic perl scalars (for input and output of data
168values).
169
170The various C<DB_ENV> etc. arguments are handles return by
171C<db_env_create>, C<db_create>, C<txn_begin> and so on. If they have an
172appended C<_ornull> this means they are optional and you can pass C<undef>
173for them, resulting a NULL pointer on the C level.
174
175The C<SV *callback> is the optional callback function to call when the
176request is completed. This last callback argument is special: the callback
177is simply the last argument passed. If there are "optional" arguments
178before the callback they can be left out. The callback itself can be left
179out or specified as C<undef>, in which case the function will be executed
180synchronously.
181
182For example, C<db_env_txn_checkpoint> usually is called with all integer
183arguments zero. These can be left out, so all of these specify a call
184to C<< DB_ENV->txn_checkpoint >>, to be executed asynchronously with a
185callback to be called:
186
187 db_env_txn_checkpoint $db_env, 0, 0, 0, sub { };
188 db_env_txn_checkpoint $db_env, 0, 0, sub { };
189 db_env_txn_checkpoint $db_env, sub { };
190
191While these all specify a call to C<< DB_ENV->txn_checkpoint >> to be
192executed synchronously:
193
194 db_env_txn_checkpoint $db_env, 0, 0, 0, undef;
195 db_env_txn_checkpoint $db_env, 0, 0, 0;
196 db_env_txn_checkpoint $db_env, 0;
197
198=head3 BDB functions
199
200Functions in the BDB namespace, exported by default:
201
202 $env = db_env_create (U32 env_flags = 0)
203 flags: RPCCLIENT
204
205 db_env_open (DB_ENV *env, bdb_filename db_home, U32 open_flags, int mode, SV *callback = &PL_sv_undef)
206 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
207 db_env_close (DB_ENV *env, U32 flags = 0, SV *callback = &PL_sv_undef)
208 db_env_txn_checkpoint (DB_ENV *env, U32 kbyte = 0, U32 min = 0, U32 flags = 0, SV *callback = &PL_sv_undef)
209 flags: FORCE
210 db_env_lock_detect (DB_ENV *env, U32 flags = 0, U32 atype = DB_LOCK_DEFAULT, SV *dummy = 0, SV *callback = &PL_sv_undef)
211 atype: LOCK_DEFAULT LOCK_EXPIRE LOCK_MAXLOCKS LOCK_MAXWRITE LOCK_MINLOCKS LOCK_MINWRITE LOCK_OLDEST LOCK_RANDOM LOCK_YOUNGEST
212 db_env_memp_sync (DB_ENV *env, SV *dummy = 0, SV *callback = &PL_sv_undef)
213 db_env_memp_trickle (DB_ENV *env, int percent, SV *dummy = 0, SV *callback = &PL_sv_undef)
214 db_env_dbremove (DB_ENV *env, DB_TXN_ornull *txnid, bdb_filename file, bdb_filename database, U32 flags = 0, SV *callback = &PL_sv_undef)
215 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)
216
217 $db = db_create (DB_ENV *env = 0, U32 flags = 0)
218 flags: XA_CREATE
219
220 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)
221 flags: AUTO_COMMIT CREATE EXCL MULTIVERSION NOMMAP RDONLY READ_UNCOMMITTED THREAD TRUNCATE
222 db_close (DB *db, U32 flags = 0, SV *callback = &PL_sv_undef)
223 flags: DB_NOSYNC
224 db_upgrade (DB *db, bdb_filename file, U32 flags = 0, SV *callback = &PL_sv_undef)
225 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)
226 flags: FREELIST_ONLY FREE_SPACE
227 db_sync (DB *db, U32 flags = 0, SV *callback = &PL_sv_undef)
228 db_key_range (DB *db, DB_TXN_ornull *txn, SV *key, SV *key_range, U32 flags = 0, SV *callback = &PL_sv_undef)
229 db_put (DB *db, DB_TXN_ornull *txn, SV *key, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef)
230 flags: APPEND NODUPDATA NOOVERWRITE
231 db_exists (DB *db, DB_TXN_ornull *txn, SV *key, U32 flags = 0, SV *callback = 0) (v4.6)
232 db_get (DB *db, DB_TXN_ornull *txn, SV *key, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef)
233 flags: CONSUME CONSUME_WAIT GET_BOTH SET_RECNO MULTIPLE READ_COMMITTED READ_UNCOMMITTED RMW
234 db_pget (DB *db, DB_TXN_ornull *txn, SV *key, SV *pkey, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef)
235 flags: CONSUME CONSUME_WAIT GET_BOTH SET_RECNO MULTIPLE READ_COMMITTED READ_UNCOMMITTED RMW
236 db_del (DB *db, DB_TXN_ornull *txn, SV *key, U32 flags = 0, SV *callback = &PL_sv_undef)
237 db_txn_commit (DB_TXN *txn, U32 flags = 0, SV *callback = &PL_sv_undef)
238 flags: TXN_NOSYNC TXN_SYNC
239 db_txn_abort (DB_TXN *txn, SV *callback = &PL_sv_undef)
240
241 db_c_close (DBC *dbc, SV *callback = &PL_sv_undef)
242 db_c_count (DBC *dbc, SV *count, U32 flags = 0, SV *callback = &PL_sv_undef)
243 db_c_put (DBC *dbc, SV *key, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef)
244 flags: AFTER BEFORE CURRENT KEYFIRST KEYLAST NODUPDATA
245 db_c_get (DBC *dbc, SV *key, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef)
246 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
247 db_c_pget (DBC *dbc, SV *key, SV *pkey, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef)
248 db_c_del (DBC *dbc, U32 flags = 0, SV *callback = &PL_sv_undef)
249
250 db_sequence_open (DB_SEQUENCE *seq, DB_TXN_ornull *txnid, SV *key, U32 flags = 0, SV *callback = &PL_sv_undef)
251 flags: CREATE EXCL
252 db_sequence_close (DB_SEQUENCE *seq, U32 flags = 0, SV *callback = &PL_sv_undef)
253 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)
254 flags: TXN_NOSYNC
255 db_sequence_remove (DB_SEQUENCE *seq, DB_TXN_ornull *txnid = 0, U32 flags = 0, SV *callback = &PL_sv_undef)
256 flags: TXN_NOSYNC
257
258=head4 db_txn_finish (DB_TXN *txn, U32 flags = 0, SV *callback = &PL_sv_undef)
259
260This is not actually a Berkeley DB function but a BDB module
261extension. The background for this exytension is: It is very annoying to
262have to check every single BDB function for error returns and provide a
263codepath out of your transaction. While the BDB module still makes this
264possible, it contains the following extensions:
265
266When a transaction-protected function returns any operating system
267error (errno > 0), BDB will set the C<TXN_DEADLOCK> flag on the
268transaction. This flag is also set by Berkeley DB functions themselves
269when an operation fails with LOCK_DEADLOCK, and it causes all further
270operations on that transaction (including C<db_txn_commit>) to fail.
271
272The C<db_txn_finish> request will look at this flag, and, if it is set,
273will automatically call C<db_txn_abort> (setting errno to C<LOCK_DEADLOCK>
274if it isn't set to something else yet). If it isn't set, it will call
275C<db_txn_commit> and return the error normally.
276
277How to use this? Easy: just write your transaction normally:
278
279 my $txn = $db_env->txn_begin;
280 db_get $db, $txn, "key", my $data;
281 db_put $db, $txn, "key", $data + 1 unless $! == BDB::NOTFOUND;
282 db_txn_finish $txn;
283 die "transaction failed" if $!;
284
285That is, handle only the expected errors. If something unexpected happens
286(EIO, LOCK_NOTGRANTED or a deadlock in either db_get or db_put), then the remaining
287requests (db_put in this case) will simply be skipped (they will fail with
288LOCK_DEADLOCK) and the transaction will be aborted.
289
290You can use the C<< $txn->failed >> method to check wether a transaction
291has failed in this way and abort further processing (excluding
292C<db_txn_finish>).
293
294=head3 DB_ENV/database environment methods
295
296Methods available on DB_ENV/$env handles:
297
298 DESTROY (DB_ENV_ornull *env)
299 CODE:
300 if (env)
301 env->close (env, 0);
302
303 $int = $env->set_data_dir (const char *dir)
304 $int = $env->set_tmp_dir (const char *dir)
305 $int = $env->set_lg_dir (const char *dir)
306 $int = $env->set_shm_key (long shm_key)
307 $int = $env->set_cachesize (U32 gbytes, U32 bytes, int ncache = 0)
308 $int = $env->set_flags (U32 flags, int onoff = 1)
309 $int = $env->log_set_config (U32 flags, int onoff = 1) (v4.7)
310 $int = $env->set_intermediate_dir_mode (const char *modestring) (v4.7)
311 $env->set_errfile (FILE *errfile = 0)
312 $env->set_msgfile (FILE *msgfile = 0)
313 $int = $env->set_verbose (U32 which, int onoff = 1)
314 $int = $env->set_encrypt (const char *password, U32 flags = 0)
315 $int = $env->set_timeout (NV timeout_seconds, U32 flags = SET_TXN_TIMEOUT)
316 $int = $env->set_mp_max_openfd (int maxopenfd);
317 $int = $env->set_mp_max_write (int maxwrite, int maxwrite_sleep);
318 $int = $env->set_mp_mmapsize (int mmapsize_mb)
319 $int = $env->set_lk_detect (U32 detect = DB_LOCK_DEFAULT)
320 $int = $env->set_lk_max_lockers (U32 max)
321 $int = $env->set_lk_max_locks (U32 max)
322 $int = $env->set_lk_max_objects (U32 max)
323 $int = $env->set_lg_bsize (U32 max)
324 $int = $env->set_lg_max (U32 max)
325 $int = $env->mutex_set_increment (U32 increment)
326 $int = $env->mutex_set_tas_spins (U32 tas_spins)
327 $int = $env->mutex_set_max (U32 max)
328 $int = $env->mutex_set_align (U32 align)
329
330 $txn = $env->txn_begin (DB_TXN_ornull *parent = 0, U32 flags = 0)
331 flags: READ_COMMITTED READ_UNCOMMITTED TXN_NOSYNC TXN_NOWAIT TXN_SNAPSHOT TXN_SYNC TXN_WAIT TXN_WRITE_NOSYNC
332 $txn = $env->cdsgroup_begin; (v4.5)
333
334=head4 Example:
335
336 use AnyEvent;
337 use BDB;
338
339 our $FH; open $FH, "<&=" . BDB::poll_fileno;
340 our $WATCHER = AnyEvent->io (fh => $FH, poll => 'r', cb => \&BDB::poll_cb);
341
342 BDB::min_parallel 8;
343
344 my $env = db_env_create;
345
346 mkdir "bdtest", 0700;
347 db_env_open
348 $env,
349 "bdtest",
350 BDB::INIT_LOCK | BDB::INIT_LOG | BDB::INIT_MPOOL | BDB::INIT_TXN | BDB::RECOVER | BDB::USE_ENVIRON | BDB::CREATE,
351 0600;
352
353 $env->set_flags (BDB::AUTO_COMMIT | BDB::TXN_NOSYNC, 1);
354
355
356=head3 DB/database methods
357
358Methods available on DB/$db handles:
359
360 DESTROY (DB_ornull *db)
361 CODE:
362 if (db)
363 {
364 SV *env = (SV *)db->app_private;
365 db->close (db, 0);
366 SvREFCNT_dec (env);
367 }
368
369 $int = $db->set_cachesize (U32 gbytes, U32 bytes, int ncache = 0)
370 $int = $db->set_flags (U32 flags)
371 flags: CHKSUM ENCRYPT TXN_NOT_DURABLE
372 Btree: DUP DUPSORT RECNUM REVSPLITOFF
373 Hash: DUP DUPSORT
374 Queue: INORDER
375 Recno: RENUMBER SNAPSHOT
376
377 $int = $db->set_encrypt (const char *password, U32 flags)
378 $int = $db->set_lorder (int lorder)
379 $int = $db->set_bt_minkey (U32 minkey)
380 $int = $db->set_re_delim (int delim)
381 $int = $db->set_re_pad (int re_pad)
382 $int = $db->set_re_source (char *source)
383 $int = $db->set_re_len (U32 re_len)
384 $int = $db->set_h_ffactor (U32 h_ffactor)
385 $int = $db->set_h_nelem (U32 h_nelem)
386 $int = $db->set_q_extentsize (U32 extentsize)
387
388 $dbc = $db->cursor (DB_TXN_ornull *txn = 0, U32 flags = 0)
389 flags: READ_COMMITTED READ_UNCOMMITTED WRITECURSOR TXN_SNAPSHOT
390 $seq = $db->sequence (U32 flags = 0)
391
392=head4 Example:
393
394 my $db = db_create $env;
395 db_open $db, undef, "table", undef, BDB::BTREE, BDB::AUTO_COMMIT | BDB::CREATE | BDB::READ_UNCOMMITTED, 0600;
396
397 for (1..1000) {
398 db_put $db, undef, "key $_", "data $_";
399
400 db_key_range $db, undef, "key $_", my $keyrange;
401 my ($lt, $eq, $gt) = @$keyrange;
402 }
403
404 db_del $db, undef, "key $_" for 1..1000;
405
406 db_sync $db;
407
408
409=head3 DB_TXN/transaction methods
410
411Methods available on DB_TXN/$txn handles:
412
413 DESTROY (DB_TXN_ornull *txn)
414 CODE:
415 if (txn)
416 txn->abort (txn);
417
418 $int = $txn->set_timeout (NV timeout_seconds, U32 flags = SET_TXN_TIMEOUT)
419 flags: SET_LOCK_TIMEOUT SET_TXN_TIMEOUT
420
421 $bool = $txn->failed
422 # see db_txn_finish documentation, above
423
424
425=head3 DBC/cursor methods
426
427Methods available on DBC/$dbc handles:
428
429 DESTROY (DBC_ornull *dbc)
430 CODE:
431 if (dbc)
432 dbc->c_close (dbc);
433
434 $int = $cursor->set_priority ($priority = PRIORITY_*) (v4.6)
435
436=head4 Example:
437
438 my $c = $db->cursor;
439
440 for (;;) {
441 db_c_get $c, my $key, my $data, BDB::NEXT;
442 warn "<$!,$key,$data>";
443 last if $!;
444 }
445
446 db_c_close $c;
447
448
449=head3 DB_SEQUENCE/sequence methods
450
451Methods available on DB_SEQUENCE/$seq handles:
452
453 DESTROY (DB_SEQUENCE_ornull *seq)
454 CODE:
455 if (seq)
456 seq->close (seq, 0);
457
458 $int = $seq->initial_value (db_seq_t value)
459 $int = $seq->set_cachesize (U32 size)
460 $int = $seq->set_flags (U32 flags)
461 flags: SEQ_DEC SEQ_INC SEQ_WRAP
462 $int = $seq->set_range (db_seq_t min, db_seq_t max)
463
464=head4 Example:
465
466 my $seq = $db->sequence;
467
468 db_sequence_open $seq, undef, "seq", BDB::CREATE;
469 db_sequence_get $seq, undef, 1, my $value;
470
471
95=head2 SUPPORT FUNCTIONS 472=head2 SUPPORT FUNCTIONS
96 473
97=head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION 474=head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION
98 475
99=over 4 476=over 4
477
478=item $msg = BDB::strerror [$errno]
479
480Returns the string corresponding to the given errno value. If no argument
481is given, use C<$!>.
482
483Note that the BDB module also patches the C<$!> variable directly, so you
484should be able to get a bdb error string by simply stringifying C<$!>.
100 485
101=item $fileno = BDB::poll_fileno 486=item $fileno = BDB::poll_fileno
102 487
103Return the I<request result pipe file descriptor>. This filehandle must be 488Return the I<request result pipe file descriptor>. This filehandle must be
104polled for reading by some mechanism outside this module (e.g. Event or 489polled for reading by some mechanism outside this module (e.g. Event or
143interactiveness when perl is not fast enough to process all requests in 528interactiveness when perl is not fast enough to process all requests in
144time. 529time.
145 530
146For interactive programs, values such as C<0.01> to C<0.1> should be fine. 531For interactive programs, values such as C<0.01> to C<0.1> should be fine.
147 532
148Example: Install an Event watcher that automatically calls 533Example: Install an EV watcher that automatically calls
149BDB::poll_cb with low priority, to ensure that other parts of the 534BDB::poll_cb with low priority, to ensure that other parts of the
150program get the CPU sometimes even under high AIO load. 535program get the CPU sometimes even under high load.
151 536
152 # try not to spend much more than 0.1s in poll_cb 537 # try not to spend much more than 0.1s in poll_cb
153 BDB::max_poll_time 0.1; 538 BDB::max_poll_time 0.1;
154 539
155 # use a low priority so other tasks have priority 540 my $bdb_poll = EV::io BDB::poll_fileno, EV::READ, \&BDB::poll_cb);
156 Event->io (fd => BDB::poll_fileno,
157 poll => 'r', nice => 1,
158 cb => &BDB::poll_cb);
159 541
160=item BDB::poll_wait 542=item BDB::poll_wait
161 543
162If there are any outstanding requests and none of them in the result 544If there are any outstanding requests and none of them in the result
163phase, wait till the result filehandle becomes ready for reading (simply 545phase, wait till the result filehandle becomes ready for reading (simply
175 557
176 BDB::poll_wait, BDB::poll_cb 558 BDB::poll_wait, BDB::poll_cb
177 559
178=item BDB::flush 560=item BDB::flush
179 561
180Wait till all outstanding AIO requests have been handled. 562Wait till all outstanding BDB requests have been handled.
181 563
182Strictly equivalent to: 564Strictly equivalent to:
183 565
184 BDB::poll_wait, BDB::poll_cb 566 BDB::poll_wait, BDB::poll_cb
185 while BDB::nreqs; 567 while BDB::nreqs;
186 568
187=back 569=back
188 570
571=head3 VERSION CHECKING
572
573BerkeleyDB comes in various versions, many of them have minor
574incompatibilities. This means that traditional "at least version x.x"
575checks are often not sufficient.
576
577Example: set the log_autoremove option in a way compatible with <v4.7 and
578v4.7. Note the use of & on the constants to avoid triggering a compiletime
579bug when the symbol isn't available.
580
581 $DB_ENV->set_flags (&BDB::LOG_AUTOREMOVE ) if BDB::VERSION v0, v4.7;
582 $DB_ENV->log_set_config (&BDB::LOG_AUTO_REMOVE) if BDB::VERSION v4.7;
583
584=over 4
585
586=item BDB::VERSION
587
588The C<BDB::VERSION> function, when called without arguments, returns the
589Berkeley DB version as a v-string (usually with 3 components). You should
590use C<lt> and C<ge> operators exclusively to make comparisons.
591
592Example: check for at least version 4.7.
593
594 BDB::VERSION ge v4.7 or die;
595
596=item BDB::VERSION min-version
597
598Returns true if the BDB version is at least the given version (specified
599as a v-string), false otherwise.
600
601Example: check for at least version 4.5.
602
603 BDB::VERSION v4.7 or die;
604
605=item BDB::VERSION min-version, max-version
606
607Returns true of the BDB version is at least version C<min-version> (specify C<undef> or C<v0> for any minimum version)
608and less then C<max-version>.
609
610Example: check wether version is strictly less then v4.7.
611
612 BDB::VERSION v0, v4.7
613 or die "version 4.7 is not yet supported";
614
615=back
616
617=cut
618
619sub VERSION {
620 # I was dumb enough to override the VERSION method here, so let's try
621 # to fix it up.
622
623 if ($_[0] eq __PACKAGE__) {
624 $VERSION
625 } else {
626 if (@_ > 0) {
627 return undef if VERSION_v lt $_[0];
628 if (@_ > 1) {
629 return undef if VERSION_v ge $_[1];
630 }
631 }
632
633 VERSION_v
634 }
635}
636
189=head3 CONTROLLING THE NUMBER OF THREADS 637=head3 CONTROLLING THE NUMBER OF THREADS
190 638
191=over 4 639=over 4
192 640
193=item BDB::min_parallel $nthreads 641=item BDB::min_parallel $nthreads
194 642
195Set the minimum number of AIO threads to C<$nthreads>. The current 643Set the minimum number of BDB threads to C<$nthreads>. The current
196default is C<8>, which means eight asynchronous operations can execute 644default is C<8>, which means eight asynchronous operations can execute
197concurrently at any one time (the number of outstanding requests, 645concurrently at any one time (the number of outstanding requests,
198however, is unlimited). 646however, is unlimited).
199 647
200BDB starts threads only on demand, when an AIO request is queued and 648BDB starts threads only on demand, when an BDB request is queued and
201no free thread exists. Please note that queueing up a hundred requests can 649no free thread exists. Please note that queueing up a hundred requests can
202create demand for a hundred threads, even if it turns out that everything 650create demand for a hundred threads, even if it turns out that everything
203is in the cache and could have been processed faster by a single thread. 651is in the cache and could have been processed faster by a single thread.
204 652
205It is recommended to keep the number of threads relatively low, as some 653It is recommended to keep the number of threads relatively low, as some
210Under most circumstances you don't need to call this function, as the 658Under most circumstances you don't need to call this function, as the
211module selects a default that is suitable for low to moderate load. 659module selects a default that is suitable for low to moderate load.
212 660
213=item BDB::max_parallel $nthreads 661=item BDB::max_parallel $nthreads
214 662
215Sets the maximum number of AIO threads to C<$nthreads>. If more than the 663Sets the maximum number of BDB threads to C<$nthreads>. If more than the
216specified number of threads are currently running, this function kills 664specified number of threads are currently running, this function kills
217them. This function blocks until the limit is reached. 665them. This function blocks until the limit is reached.
218 666
219While C<$nthreads> are zero, aio requests get queued but not executed 667While C<$nthreads> are zero, aio requests get queued but not executed
220until the number of threads has been increased again. 668until the number of threads has been increased again.
255 703
256You can still queue as many requests as you want. Therefore, 704You can still queue as many requests as you want. Therefore,
257C<max_oustsanding> is mainly useful in simple scripts (with low values) or 705C<max_oustsanding> is mainly useful in simple scripts (with low values) or
258as a stop gap to shield against fatal memory overflow (with large values). 706as a stop gap to shield against fatal memory overflow (with large values).
259 707
260=item BDB::set_sync_prepare $cb 708=item $old_cb = BDB::set_sync_prepare $cb
261 709
262Sets a callback that is called whenever a request is created without an 710Sets a callback that is called whenever a request is created without an
263explicit callback. It has to return two code references. The first is used 711explicit callback. It has to return two code references. The first is used
264as the request callback, and the second is called to wait until the first 712as the request callback (it should save the return status), and the second
265callback has been called. The default implementation works like this: 713is called to wait until the first callback has been called (it must set
714C<$!> to the return status).
715
716This mechanism can be used to include BDB into other event mechanisms,
717such as L<Coro::BDB>.
718
719To allow other, callback-based, events to be executed while callback-less
720ones are run, you could use this sync prepare function:
266 721
267 sub { 722 sub {
268 my $status; 723 my $status;
269 ( 724 (
270 sub { $status = $! }, 725 sub { $status = $! },
271 sub { BDB::poll while !defined $status; $! = $status }, 726 sub { BDB::poll while !defined $status; $! = $status },
272 ) 727 )
273 } 728 }
274 729
730It works by polling for results till the request has finished and then
731sets C<$!> to the return value. This means that if you don't use a
732callback, BDB would simply fall back to synchronous operations.
733
734By default, or if the sync prepare function is set to C<undef>, is to
735execute callback-less BDB requests in the foreground thread, setting C<$!>
736to the return value, without polling for other events.
737
275=back 738=back
276 739
277=head3 STATISTICAL INFORMATION 740=head3 STATISTICAL INFORMATION
278 741
279=over 4 742=over 4
300 763
301=back 764=back
302 765
303=cut 766=cut
304 767
305set_sync_prepare { 768set_sync_prepare (undef);
306 my $status;
307 (
308 sub {
309 $status = $!;
310 },
311 sub {
312 BDB::poll while !defined $status;
313 $! = $status;
314 },
315 )
316};
317 769
318min_parallel 8; 770min_parallel 8;
319 771
320END { flush } 772END { flush }
321 773
323 775
324=head2 FORK BEHAVIOUR 776=head2 FORK BEHAVIOUR
325 777
326This module should do "the right thing" when the process using it forks: 778This module should do "the right thing" when the process using it forks:
327 779
328Before the fork, IO::AIO enters a quiescent state where no requests 780Before the fork, BDB enters a quiescent state where no requests
329can be added in other threads and no results will be processed. After 781can be added in other threads and no results will be processed. After
330the fork the parent simply leaves the quiescent state and continues 782the fork the parent simply leaves the quiescent state and continues
331request/result processing, while the child frees the request/result queue 783request/result processing, while the child frees the request/result queue
332(so that the requests started before the fork will only be handled in the 784(so that the requests started before the fork will only be handled in the
333parent). Threads will be started on demand until the limit set in the 785parent). Threads will be started on demand until the limit set in the
334parent process has been reached again. 786parent process has been reached again.
335 787
336In short: the parent will, after a short pause, continue as if fork had 788In short: the parent will, after a short pause, continue as if fork had
337not been called, while the child will act as if IO::AIO has not been used 789not been called, while the child will act as if BDB has not been used
338yet. 790yet.
791
792Win32 note: there is no fork on win32, and perls emulation of it is too
793broken to be supported, so do not use BDB in a windows pseudo-fork, better
794yet, switch to a more capable platform.
339 795
340=head2 MEMORY USAGE 796=head2 MEMORY USAGE
341 797
342Per-request usage: 798Per-request usage:
343 799
345bytes of memory. In addition, stat requests need a stat buffer (possibly 801bytes of memory. In addition, stat requests need a stat buffer (possibly
346a few hundred bytes), readdir requires a result buffer and so on. Perl 802a few hundred bytes), readdir requires a result buffer and so on. Perl
347scalars and other data passed into aio requests will also be locked and 803scalars and other data passed into aio requests will also be locked and
348will consume memory till the request has entered the done state. 804will consume memory till the request has entered the done state.
349 805
350This is now awfully much, so queuing lots of requests is not usually a 806This is not awfully much, so queuing lots of requests is not usually a
351problem. 807problem.
352 808
353Per-thread usage: 809Per-thread usage:
354 810
355In the execution phase, some aio requests require more memory for 811In the execution phase, some aio requests require more memory for
356temporary buffers, and each thread requires a stack and other data 812temporary buffers, and each thread requires a stack and other data
357structures (usually around 16k-128k, depending on the OS). 813structures (usually around 16k-128k, depending on the OS).
358 814
359=head1 KNOWN BUGS 815=head1 KNOWN BUGS
360 816
361Known bugs will be fixed in the next release. 817Known bugs will be fixed in the next release, except:
818
819 If you use a transaction in any request, and the request returns
820 with an operating system error or DB_LOCK_NOTGRANTED, the internal
821 TXN_DEADLOCK flag will be set on the transaction. See C<db_txn_finish>,
822 above.
362 823
363=head1 SEE ALSO 824=head1 SEE ALSO
364 825
365L<Coro::AIO>. 826L<AnyEvent::BDB> (event loop integration), L<Coro::BDB> (more natural
827syntax), L<IO::AIO> (nice to have).
366 828
367=head1 AUTHOR 829=head1 AUTHOR
368 830
369 Marc Lehmann <schmorp@schmorp.de> 831 Marc Lehmann <schmorp@schmorp.de>
370 http://home.schmorp.de/ 832 http://home.schmorp.de/

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines