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

Comparing BDB/BDB.pm (file contents):
Revision 1.7 by root, Mon Mar 5 19:47:01 2007 UTC vs.
Revision 1.35 by root, Sun Mar 30 08:01:58 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 result processing with AnyEvent:
33 our $FH; open $FH, "<&=" . BDB::poll_fileno;
34 our $WATCHER = AnyEvent->io (fh => $FH, poll => 'r', cb => \&BDB::poll_cb);
35
36 # automatic result processing with EV:
37 my $WATCHER = EV::io BDB::poll_fileno, EV::READ, \&BDB::poll_cb;
38
39 # with Glib:
40 add_watch Glib::IO BDB::poll_fileno,
41 in => sub { BDB::poll_cb; 1 };
42
43 # or simply flush manually
44 BDB::flush;
45
46
9=head1 DESCRIPTION 47=head1 DESCRIPTION
10 48
11See the eg/ directory in the distribution and the berkeleydb C 49See 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 50The BDB API is very similar to the C API (the translation has been very faithful).
13known for this module so far.
14 51
15=head2 EXAMPLE 52See also the example sections in the document below and possibly the eg/
53subdirectory of the BDB distribution. Last not least see the IO::AIO
54documentation, as that module uses almost the same asynchronous request
55model as this module.
56
57I know this is woefully inadequate documentation. Send a patch!
58
16 59
17=head1 REQUEST ANATOMY AND LIFETIME 60=head1 REQUEST ANATOMY AND LIFETIME
18 61
19Every request method creates a request. which is a C data structure not 62Every request method creates a request. which is a C data structure not
20directly visible to Perl. 63directly visible to Perl.
67use strict 'vars'; 110use strict 'vars';
68 111
69use base 'Exporter'; 112use base 'Exporter';
70 113
71BEGIN { 114BEGIN {
72 our $VERSION = '0.1'; 115 our $VERSION = '1.45';
73 116
74 our @BDB_REQ = qw( 117 our @BDB_REQ = qw(
75 db_env_open db_env_close db_env_txn_checkpoint db_env_lock_detect 118 db_env_open db_env_close db_env_txn_checkpoint db_env_lock_detect
76 db_env_memp_sync db_env_memp_trickle 119 db_env_memp_sync db_env_memp_trickle
77 db_open db_close db_compact db_sync db_put db_get db_pget db_del db_key_range 120 db_open db_close db_compact db_sync db_upgrade
121 db_put db_get db_pget db_del db_key_range
78 db_txn_commit db_txn_abort 122 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 123 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 124 db_sequence_open db_sequence_close
81 db_sequence_get db_sequence_remove 125 db_sequence_get db_sequence_remove
82 ); 126 );
83 our @EXPORT = (@BDB_REQ, qw(dbreq_pri dbreq_nice db_env_create db_create)); 127 our @EXPORT = (@BDB_REQ, qw(dbreq_pri dbreq_nice db_env_create db_create));
90 134
91 require XSLoader; 135 require XSLoader;
92 XSLoader::load ("BDB", $VERSION); 136 XSLoader::load ("BDB", $VERSION);
93} 137}
94 138
139=head2 WIN32 FILENAMES/DATABASE NAME MESS
140
141Perl on Win32 supports only ASCII filenames (the reason is that it abuses
142an internal flag to store wether a filename is Unicode or ANSI, but that
143flag is used for somethign else in the perl core, so there is no way to
144detect wether a filename is ANSI or Unicode-encoded). The BDB module
145tries to work around this issue by assuming that the filename is an ANSI
146filename and BDB was built for unicode support.
147
148=head2 BERKELEYDB FUNCTIONS
149
150All of these are functions. The create functions simply return a new
151object and never block. All the remaining functions take an optional
152callback as last argument. If it is missing, then the function will be
153executed synchronously. In both cases, C<$!> will reflect the return value
154of the function.
155
156BDB functions that cannot block (mostly functions that manipulate
157settings) are method calls on the relevant objects, so the rule of thumb
158is: if it's a method, it's not blocking, if it's a function, it takes a
159callback as last argument.
160
161In the following, C<$int> signifies an integer return value,
162C<octetstring> is a "binary string" (i.e. a perl string with no character
163indices >255), C<U32> is an unsigned 32 bit integer, C<int> is some
164integer, C<NV> is a floating point value.
165
166The C<SV *> types are generic perl scalars (for input and output of data
167values), and the C<SV *callback> is the optional callback function to call
168when the request is completed.
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
175=head3 BDB functions
176
177Functions in the BDB namespace, exported by default:
178
179 $env = db_env_create (U32 env_flags = 0)
180 flags: RPCCLIENT
181
182 db_env_open (DB_ENV *env, octetstring db_home, U32 open_flags, int mode, SV *callback = &PL_sv_undef)
183 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
184 db_env_close (DB_ENV *env, U32 flags = 0, SV *callback = &PL_sv_undef)
185 db_env_txn_checkpoint (DB_ENV *env, U32 kbyte = 0, U32 min = 0, U32 flags = 0, SV *callback = &PL_sv_undef)
186 flags: FORCE
187 db_env_lock_detect (DB_ENV *env, U32 flags = 0, U32 atype = DB_LOCK_DEFAULT, SV *dummy = 0, SV *callback = &PL_sv_undef)
188 atype: LOCK_DEFAULT LOCK_EXPIRE LOCK_MAXLOCKS LOCK_MAXWRITE LOCK_MINLOCKS LOCK_MINWRITE LOCK_OLDEST LOCK_RANDOM LOCK_YOUNGEST
189 db_env_memp_sync (DB_ENV *env, SV *dummy = 0, SV *callback = &PL_sv_undef)
190 db_env_memp_trickle (DB_ENV *env, int percent, SV *dummy = 0, SV *callback = &PL_sv_undef)
191
192 $db = db_create (DB_ENV *env = 0, U32 flags = 0)
193 flags: XA_CREATE
194
195 db_open (DB *db, DB_TXN_ornull *txnid, octetstring file, octetstring database, int type, U32 flags, int mode, SV *callback = &PL_sv_undef)
196 flags: AUTO_COMMIT CREATE EXCL MULTIVERSION NOMMAP RDONLY READ_UNCOMMITTED THREAD TRUNCATE
197 db_close (DB *db, U32 flags = 0, SV *callback = &PL_sv_undef)
198 flags: DB_NOSYNC
199 db_upgrade (DB *db, octetstring file, U32 flags = 0, SV *callback = &PL_sv_undef)
200 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)
201 flags: FREELIST_ONLY FREE_SPACE
202 db_sync (DB *db, U32 flags = 0, SV *callback = &PL_sv_undef)
203 db_key_range (DB *db, DB_TXN_ornull *txn, SV *key, SV *key_range, U32 flags = 0, SV *callback = &PL_sv_undef)
204 db_put (DB *db, DB_TXN_ornull *txn, SV *key, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef)
205 flags: APPEND NODUPDATA NOOVERWRITE
206 db_get (DB *db, DB_TXN_ornull *txn, SV *key, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef)
207 flags: CONSUME CONSUME_WAIT GET_BOTH SET_RECNO MULTIPLE READ_COMMITTED READ_UNCOMMITTED RMW
208 db_pget (DB *db, DB_TXN_ornull *txn, SV *key, SV *pkey, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef)
209 flags: CONSUME CONSUME_WAIT GET_BOTH SET_RECNO MULTIPLE READ_COMMITTED READ_UNCOMMITTED RMW
210 db_del (DB *db, DB_TXN_ornull *txn, SV *key, U32 flags = 0, SV *callback = &PL_sv_undef)
211 db_txn_commit (DB_TXN *txn, U32 flags = 0, SV *callback = &PL_sv_undef)
212 flags: TXN_NOSYNC TXN_SYNC
213 db_txn_abort (DB_TXN *txn, SV *callback = &PL_sv_undef)
214
215 db_c_close (DBC *dbc, SV *callback = &PL_sv_undef)
216 db_c_count (DBC *dbc, SV *count, U32 flags = 0, SV *callback = &PL_sv_undef)
217 db_c_put (DBC *dbc, SV *key, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef)
218 flags: AFTER BEFORE CURRENT KEYFIRST KEYLAST NODUPDATA
219 db_c_get (DBC *dbc, SV *key, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef)
220 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
221 db_c_pget (DBC *dbc, SV *key, SV *pkey, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef)
222 db_c_del (DBC *dbc, U32 flags = 0, SV *callback = &PL_sv_undef)
223
224 db_sequence_open (DB_SEQUENCE *seq, DB_TXN_ornull *txnid, SV *key, U32 flags = 0, SV *callback = &PL_sv_undef)
225 flags: CREATE EXCL
226 db_sequence_close (DB_SEQUENCE *seq, U32 flags = 0, SV *callback = &PL_sv_undef)
227 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)
228 flags: TXN_NOSYNC
229 db_sequence_remove (DB_SEQUENCE *seq, DB_TXN_ornull *txnid = 0, U32 flags = 0, SV *callback = &PL_sv_undef)
230 flags: TXN_NOSYNC
231
232=head4 db_txn_finish (DB_TXN *txn, U32 flags = 0, SV *callback = &PL_sv_undef)
233
234This is not actually a Berkeley DB function but a BDB module
235extension. The background for this exytension is: It is very annoying to
236have to check every single BDB function for error returns and provide a
237codepath out of your transaction. While the BDB module still makes this
238possible, it contains the following extensions:
239
240When a transaction-protected function returns any operating system
241error (errno > 0), BDB will set the C<TXN_DEADLOCK> flag on the
242transaction. This flag is also set by Berkeley DB functions themselves
243when an operation fails with LOCK_DEADLOCK, and it causes all further
244operations on that transaction (including C<db_txn_commit>) to fail.
245
246The C<db_txn_finish> request will look at this flag, and, if it is set,
247will automatically call C<db_txn_abort> (setting errno to C<LOCK_DEADLOCK>
248if it isn't set to something else yet). If it isn't set, it will call
249C<db_txn_commit> and return the error normally.
250
251How to use this? Easy: just write your transaction normally:
252
253 my $txn = $db_env->txn_begin;
254 db_get $db, $txn, "key", my $data;
255 db_put $db, $txn, "key", $data + 1 unless $! == BDB::NOTFOUND;
256 db_txn_finish $txn;
257 die "transaction failed" if $!;
258
259That is, handle only the expected errors. If something unexpected happens
260(EIO, LOCK_NOTGRANTED or a deadlock in either db_get or db_put), then the remaining
261requests (db_put in this case) will simply be skipped (they will fail with
262LOCK_DEADLOCK) and the transaction will be aborted.
263
264You can use the C<< $txn->failed >> method to check wether a transaction
265has failed in this way and abort further processing (excluding
266C<db_txn_finish>).
267
268=head3 DB_ENV/database environment methods
269
270Methods available on DB_ENV/$env handles:
271
272 DESTROY (DB_ENV_ornull *env)
273 CODE:
274 if (env)
275 env->close (env, 0);
276
277 $int = $env->set_data_dir (const char *dir)
278 $int = $env->set_tmp_dir (const char *dir)
279 $int = $env->set_lg_dir (const char *dir)
280 $int = $env->set_shm_key (long shm_key)
281 $int = $env->set_cachesize (U32 gbytes, U32 bytes, int ncache = 0)
282 $int = $env->set_flags (U32 flags, int onoff)
283 $env->set_errfile (FILE *errfile = 0)
284 $env->set_msgfile (FILE *msgfile = 0)
285 $int = $env->set_verbose (U32 which, int onoff = 1)
286 $int = $env->set_encrypt (const char *password, U32 flags = 0)
287 $int = $env->set_timeout (NV timeout_seconds, U32 flags = SET_TXN_TIMEOUT)
288 $int = $env->set_mp_max_openfd (int maxopenfd);
289 $int = $env->set_mp_max_write (int maxwrite, int maxwrite_sleep);
290 $int = $env->set_mp_mmapsize (int mmapsize_mb)
291 $int = $env->set_lk_detect (U32 detect = DB_LOCK_DEFAULT)
292 $int = $env->set_lk_max_lockers (U32 max)
293 $int = $env->set_lk_max_locks (U32 max)
294 $int = $env->set_lk_max_objects (U32 max)
295 $int = $env->set_lg_bsize (U32 max)
296 $int = $env->set_lg_max (U32 max)
297 $int = $env->mutex_set_increment (U32 increment)
298 $int = $env->mutex_set_tas_spins (U32 tas_spins)
299 $int = $env->mutex_set_max (U32 max)
300 $int = $env->mutex_set_align (U32 align)
301
302 $txn = $env->txn_begin (DB_TXN_ornull *parent = 0, U32 flags = 0)
303 flags: READ_COMMITTED READ_UNCOMMITTED TXN_NOSYNC TXN_NOWAIT TXN_SNAPSHOT TXN_SYNC TXN_WAIT TXN_WRITE_NOSYNC
304
305=head4 Example:
306
307 use AnyEvent;
308 use BDB;
309
310 our $FH; open $FH, "<&=" . BDB::poll_fileno;
311 our $WATCHER = AnyEvent->io (fh => $FH, poll => 'r', cb => \&BDB::poll_cb);
312
313 BDB::min_parallel 8;
314
315 my $env = db_env_create;
316
317 mkdir "bdtest", 0700;
318 db_env_open
319 $env,
320 "bdtest",
321 BDB::INIT_LOCK | BDB::INIT_LOG | BDB::INIT_MPOOL | BDB::INIT_TXN | BDB::RECOVER | BDB::USE_ENVIRON | BDB::CREATE,
322 0600;
323
324 $env->set_flags (BDB::AUTO_COMMIT | BDB::TXN_NOSYNC, 1);
325
326
327=head3 DB/database methods
328
329Methods available on DB/$db handles:
330
331 DESTROY (DB_ornull *db)
332 CODE:
333 if (db)
334 {
335 SV *env = (SV *)db->app_private;
336 db->close (db, 0);
337 SvREFCNT_dec (env);
338 }
339
340 $int = $db->set_cachesize (U32 gbytes, U32 bytes, int ncache = 0)
341 $int = $db->set_flags (U32 flags)
342 flags: CHKSUM ENCRYPT TXN_NOT_DURABLE
343 Btree: DUP DUPSORT RECNUM REVSPLITOFF
344 Hash: DUP DUPSORT
345 Queue: INORDER
346 Recno: RENUMBER SNAPSHOT
347
348 $int = $db->set_encrypt (const char *password, U32 flags)
349 $int = $db->set_lorder (int lorder)
350 $int = $db->set_bt_minkey (U32 minkey)
351 $int = $db->set_re_delim (int delim)
352 $int = $db->set_re_pad (int re_pad)
353 $int = $db->set_re_source (char *source)
354 $int = $db->set_re_len (U32 re_len)
355 $int = $db->set_h_ffactor (U32 h_ffactor)
356 $int = $db->set_h_nelem (U32 h_nelem)
357 $int = $db->set_q_extentsize (U32 extentsize)
358
359 $dbc = $db->cursor (DB_TXN_ornull *txn = 0, U32 flags = 0)
360 flags: READ_COMMITTED READ_UNCOMMITTED WRITECURSOR TXN_SNAPSHOT
361 $seq = $db->sequence (U32 flags = 0)
362
363=head4 Example:
364
365 my $db = db_create $env;
366 db_open $db, undef, "table", undef, BDB::BTREE, BDB::AUTO_COMMIT | BDB::CREATE | BDB::READ_UNCOMMITTED, 0600;
367
368 for (1..1000) {
369 db_put $db, undef, "key $_", "data $_";
370
371 db_key_range $db, undef, "key $_", my $keyrange;
372 my ($lt, $eq, $gt) = @$keyrange;
373 }
374
375 db_del $db, undef, "key $_" for 1..1000;
376
377 db_sync $db;
378
379
380=head3 DB_TXN/transaction methods
381
382Methods available on DB_TXN/$txn handles:
383
384 DESTROY (DB_TXN_ornull *txn)
385 CODE:
386 if (txn)
387 txn->abort (txn);
388
389 $int = $txn->set_timeout (NV timeout_seconds, U32 flags = SET_TXN_TIMEOUT)
390 flags: SET_LOCK_TIMEOUT SET_TXN_TIMEOUT
391
392 $bool = $txn->failed
393 # see db_txn_finish documentation, above
394
395
396=head3 DBC/cursor methods
397
398Methods available on DBC/$dbc handles:
399
400 DESTROY (DBC_ornull *dbc)
401 CODE:
402 if (dbc)
403 dbc->c_close (dbc);
404
405 $int = $cursor->set_priority ($priority = PRIORITY_*)
406
407=head4 Example:
408
409 my $c = $db->cursor;
410
411 for (;;) {
412 db_c_get $c, my $key, my $data, BDB::NEXT;
413 warn "<$!,$key,$data>";
414 last if $!;
415 }
416
417 db_c_close $c;
418
419
420=head3 DB_SEQUENCE/sequence methods
421
422Methods available on DB_SEQUENCE/$seq handles:
423
424 DESTROY (DB_SEQUENCE_ornull *seq)
425 CODE:
426 if (seq)
427 seq->close (seq, 0);
428
429 $int = $seq->initial_value (db_seq_t value)
430 $int = $seq->set_cachesize (U32 size)
431 $int = $seq->set_flags (U32 flags)
432 flags: SEQ_DEC SEQ_INC SEQ_WRAP
433 $int = $seq->set_range (db_seq_t min, db_seq_t max)
434
435=head4 Example:
436
437 my $seq = $db->sequence;
438
439 db_sequence_open $seq, undef, "seq", BDB::CREATE;
440 db_sequence_get $seq, undef, 1, my $value;
441
442
95=head2 SUPPORT FUNCTIONS 443=head2 SUPPORT FUNCTIONS
96 444
97=head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION 445=head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION
98 446
99=over 4 447=over 4
448
449=item $msg = BDB::strerror [$errno]
450
451Returns the string corresponding to the given errno value. If no argument
452is given, use C<$!>.
453
454Note that the BDB module also patches the C<$!> variable directly, so you
455should be able to get a bdb error string by simply stringifying C<$!>.
100 456
101=item $fileno = BDB::poll_fileno 457=item $fileno = BDB::poll_fileno
102 458
103Return the I<request result pipe file descriptor>. This filehandle must be 459Return the I<request result pipe file descriptor>. This filehandle must be
104polled for reading by some mechanism outside this module (e.g. Event or 460polled for reading by some mechanism outside this module (e.g. Event or
143interactiveness when perl is not fast enough to process all requests in 499interactiveness when perl is not fast enough to process all requests in
144time. 500time.
145 501
146For interactive programs, values such as C<0.01> to C<0.1> should be fine. 502For interactive programs, values such as C<0.01> to C<0.1> should be fine.
147 503
148Example: Install an Event watcher that automatically calls 504Example: Install an EV watcher that automatically calls
149BDB::poll_cb with low priority, to ensure that other parts of the 505BDB::poll_cb with low priority, to ensure that other parts of the
150program get the CPU sometimes even under high AIO load. 506program get the CPU sometimes even under high load.
151 507
152 # try not to spend much more than 0.1s in poll_cb 508 # try not to spend much more than 0.1s in poll_cb
153 BDB::max_poll_time 0.1; 509 BDB::max_poll_time 0.1;
154 510
155 # use a low priority so other tasks have priority 511 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);
159 512
160=item BDB::poll_wait 513=item BDB::poll_wait
161 514
162If there are any outstanding requests and none of them in the result 515If there are any outstanding requests and none of them in the result
163phase, wait till the result filehandle becomes ready for reading (simply 516phase, wait till the result filehandle becomes ready for reading (simply
175 528
176 BDB::poll_wait, BDB::poll_cb 529 BDB::poll_wait, BDB::poll_cb
177 530
178=item BDB::flush 531=item BDB::flush
179 532
180Wait till all outstanding AIO requests have been handled. 533Wait till all outstanding BDB requests have been handled.
181 534
182Strictly equivalent to: 535Strictly equivalent to:
183 536
184 BDB::poll_wait, BDB::poll_cb 537 BDB::poll_wait, BDB::poll_cb
185 while BDB::nreqs; 538 while BDB::nreqs;
186 539
540=back
541
187=head3 CONTROLLING THE NUMBER OF THREADS 542=head3 CONTROLLING THE NUMBER OF THREADS
188 543
544=over 4
545
189=item BDB::min_parallel $nthreads 546=item BDB::min_parallel $nthreads
190 547
191Set the minimum number of AIO threads to C<$nthreads>. The current 548Set the minimum number of BDB threads to C<$nthreads>. The current
192default is C<8>, which means eight asynchronous operations can execute 549default is C<8>, which means eight asynchronous operations can execute
193concurrently at any one time (the number of outstanding requests, 550concurrently at any one time (the number of outstanding requests,
194however, is unlimited). 551however, is unlimited).
195 552
196BDB starts threads only on demand, when an AIO request is queued and 553BDB starts threads only on demand, when an BDB request is queued and
197no free thread exists. Please note that queueing up a hundred requests can 554no free thread exists. Please note that queueing up a hundred requests can
198create demand for a hundred threads, even if it turns out that everything 555create demand for a hundred threads, even if it turns out that everything
199is in the cache and could have been processed faster by a single thread. 556is in the cache and could have been processed faster by a single thread.
200 557
201It is recommended to keep the number of threads relatively low, as some 558It is recommended to keep the number of threads relatively low, as some
206Under most circumstances you don't need to call this function, as the 563Under most circumstances you don't need to call this function, as the
207module selects a default that is suitable for low to moderate load. 564module selects a default that is suitable for low to moderate load.
208 565
209=item BDB::max_parallel $nthreads 566=item BDB::max_parallel $nthreads
210 567
211Sets the maximum number of AIO threads to C<$nthreads>. If more than the 568Sets the maximum number of BDB threads to C<$nthreads>. If more than the
212specified number of threads are currently running, this function kills 569specified number of threads are currently running, this function kills
213them. This function blocks until the limit is reached. 570them. This function blocks until the limit is reached.
214 571
215While C<$nthreads> are zero, aio requests get queued but not executed 572While C<$nthreads> are zero, aio requests get queued but not executed
216until the number of threads has been increased again. 573until the number of threads has been increased again.
319 676
320=head2 FORK BEHAVIOUR 677=head2 FORK BEHAVIOUR
321 678
322This module should do "the right thing" when the process using it forks: 679This module should do "the right thing" when the process using it forks:
323 680
324Before the fork, IO::AIO enters a quiescent state where no requests 681Before the fork, BDB enters a quiescent state where no requests
325can be added in other threads and no results will be processed. After 682can be added in other threads and no results will be processed. After
326the fork the parent simply leaves the quiescent state and continues 683the fork the parent simply leaves the quiescent state and continues
327request/result processing, while the child frees the request/result queue 684request/result processing, while the child frees the request/result queue
328(so that the requests started before the fork will only be handled in the 685(so that the requests started before the fork will only be handled in the
329parent). Threads will be started on demand until the limit set in the 686parent). Threads will be started on demand until the limit set in the
330parent process has been reached again. 687parent process has been reached again.
331 688
332In short: the parent will, after a short pause, continue as if fork had 689In short: the parent will, after a short pause, continue as if fork had
333not been called, while the child will act as if IO::AIO has not been used 690not been called, while the child will act as if BDB has not been used
334yet. 691yet.
692
693Win32 note: there is no fork on win32, and perls emulation of it is too
694broken to be supported, so do not use BDB in a windows pseudo-fork, better
695yet, switch to a more capable platform.
335 696
336=head2 MEMORY USAGE 697=head2 MEMORY USAGE
337 698
338Per-request usage: 699Per-request usage:
339 700
341bytes of memory. In addition, stat requests need a stat buffer (possibly 702bytes of memory. In addition, stat requests need a stat buffer (possibly
342a few hundred bytes), readdir requires a result buffer and so on. Perl 703a few hundred bytes), readdir requires a result buffer and so on. Perl
343scalars and other data passed into aio requests will also be locked and 704scalars and other data passed into aio requests will also be locked and
344will consume memory till the request has entered the done state. 705will consume memory till the request has entered the done state.
345 706
346This is now awfully much, so queuing lots of requests is not usually a 707This is not awfully much, so queuing lots of requests is not usually a
347problem. 708problem.
348 709
349Per-thread usage: 710Per-thread usage:
350 711
351In the execution phase, some aio requests require more memory for 712In the execution phase, some aio requests require more memory for
352temporary buffers, and each thread requires a stack and other data 713temporary buffers, and each thread requires a stack and other data
353structures (usually around 16k-128k, depending on the OS). 714structures (usually around 16k-128k, depending on the OS).
354 715
355=head1 KNOWN BUGS 716=head1 KNOWN BUGS
356 717
357Known bugs will be fixed in the next release. 718Known bugs will be fixed in the next release, except:
719
720 If you use a transaction in any request, and the request returns
721 with an operating system error or DB_LOCK_NOTGRANTED, the internal
722 TXN_DEADLOCK flag will be set on the transaction. See C<db_txn_finish>,
723 above.
358 724
359=head1 SEE ALSO 725=head1 SEE ALSO
360 726
361L<Coro::AIO>. 727L<Coro::BDB>, L<IO::AIO>.
362 728
363=head1 AUTHOR 729=head1 AUTHOR
364 730
365 Marc Lehmann <schmorp@schmorp.de> 731 Marc Lehmann <schmorp@schmorp.de>
366 http://home.schmorp.de/ 732 http://home.schmorp.de/

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines