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

Comparing BDB/BDB.pm (file contents):
Revision 1.1 by root, Mon Feb 5 18:40:55 2007 UTC vs.
Revision 1.40 by root, Tue Jul 8 08:35:12 2008 UTC

1=head1 NAME 1=head1 NAME
2 2
3BDB::AIO - Asynchronous Berkeley DB access 3BDB - Asynchronous Berkeley DB access
4 4
5=head1 SYNOPSIS 5=head1 SYNOPSIS
6 6
7 use BDB::AIO; 7 use BDB;
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
8 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.
55 101
56=back 102=back
57 103
58=cut 104=cut
59 105
60package BDB::AIO; 106package BDB;
61 107
62no warnings; 108no warnings;
63use strict 'vars'; 109use strict 'vars';
64 110
65use base 'Exporter'; 111use base 'Exporter';
66 112
67BEGIN { 113BEGIN {
68 our $VERSION = '0.1'; 114 our $VERSION = '1.6';
69 115
70 our @BDB_REQ = qw(); 116 our @BDB_REQ = qw(
117 db_env_open db_env_close db_env_txn_checkpoint db_env_lock_detect
118 db_env_memp_sync db_env_memp_trickle db_env_dbrename db_env_dbremove
119 db_open db_close db_compact db_sync db_upgrade
120 db_put db_get db_pget db_del db_key_range
121 db_txn_commit db_txn_abort db_txn_finish
122 db_c_close db_c_count db_c_put db_c_get db_c_pget db_c_del
123 db_sequence_open db_sequence_close
124 db_sequence_get db_sequence_remove
125 );
126 our @EXPORT = (@BDB_REQ, qw(dbreq_pri dbreq_nice db_env_create db_create));
127 our @EXPORT_OK = qw(
71 our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush 128 poll_fileno poll_cb poll_wait flush
72 min_parallel max_parallel max_idle 129 min_parallel max_parallel max_idle
73 nreqs nready npending nthreads 130 nreqs nready npending nthreads
74 max_poll_time max_poll_reqs); 131 max_poll_time max_poll_reqs
132 );
75 133
76 require XSLoader; 134 require XSLoader;
77 XSLoader::load ("BDB::AIO", $VERSION); 135 XSLoader::load ("BDB", $VERSION);
78} 136}
79 137
138=head2 WIN32 FILENAMES/DATABASE NAME MESS
139
140Perl on Win32 supports only ASCII filenames (the reason is that it abuses
141an internal flag to store wether a filename is Unicode or ANSI, but that
142flag is used for somethign else in the perl core, so there is no way to
143detect wether a filename is ANSI or Unicode-encoded). The BDB module
144tries to work around this issue by assuming that the filename is an ANSI
145filename and BDB was built for unicode support.
146
147=head2 BERKELEYDB FUNCTIONS
148
149All of these are functions. The create functions simply return a new
150object and never block. All the remaining functions take an optional
151callback as last argument. If it is missing, then the function will be
152executed synchronously. In both cases, C<$!> will reflect the return value
153of the function.
154
155BDB functions that cannot block (mostly functions that manipulate
156settings) are method calls on the relevant objects, so the rule of thumb
157is: if it's a method, it's not blocking, if it's a function, it takes a
158callback as last argument.
159
160In the following, C<$int> signifies an integer return value,
161C<bdb_filename> is a "filename" (octets on unix, madness on windows),
162C<U32> is an unsigned 32 bit integer, C<int> is some integer, C<NV> is a
163floating point value.
164
165The C<SV *> types are generic perl scalars (for input and output of data
166values), and the C<SV *callback> is the optional callback function to call
167when the request is completed.
168
169The various C<DB_ENV> etc. arguments are handles return by
170C<db_env_create>, C<db_create>, C<txn_begin> and so on. If they have an
171appended C<_ornull> this means they are optional and you can pass C<undef>
172for them, resulting a NULL pointer on the C level.
173
174=head3 BDB functions
175
176Functions in the BDB namespace, exported by default:
177
178 $env = db_env_create (U32 env_flags = 0)
179 flags: RPCCLIENT
180
181 db_env_open (DB_ENV *env, bdb_filename db_home, U32 open_flags, int mode, SV *callback = &PL_sv_undef)
182 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
183 db_env_close (DB_ENV *env, U32 flags = 0, SV *callback = &PL_sv_undef)
184 db_env_txn_checkpoint (DB_ENV *env, U32 kbyte = 0, U32 min = 0, U32 flags = 0, SV *callback = &PL_sv_undef)
185 flags: FORCE
186 db_env_lock_detect (DB_ENV *env, U32 flags = 0, U32 atype = DB_LOCK_DEFAULT, SV *dummy = 0, SV *callback = &PL_sv_undef)
187 atype: LOCK_DEFAULT LOCK_EXPIRE LOCK_MAXLOCKS LOCK_MAXWRITE LOCK_MINLOCKS LOCK_MINWRITE LOCK_OLDEST LOCK_RANDOM LOCK_YOUNGEST
188 db_env_memp_sync (DB_ENV *env, SV *dummy = 0, SV *callback = &PL_sv_undef)
189 db_env_memp_trickle (DB_ENV *env, int percent, SV *dummy = 0, SV *callback = &PL_sv_undef)
190 db_env_dbremove (DB_ENV *env, DB_TXN_ornull *txnid, bdb_filename file, bdb_filename database, U32 flags = 0, SV *callback = &PL_sv_undef)
191 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)
192
193 $db = db_create (DB_ENV *env = 0, U32 flags = 0)
194 flags: XA_CREATE
195
196 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)
197 flags: AUTO_COMMIT CREATE EXCL MULTIVERSION NOMMAP RDONLY READ_UNCOMMITTED THREAD TRUNCATE
198 db_close (DB *db, U32 flags = 0, SV *callback = &PL_sv_undef)
199 flags: DB_NOSYNC
200 db_upgrade (DB *db, bdb_filename file, U32 flags = 0, SV *callback = &PL_sv_undef)
201 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)
202 flags: FREELIST_ONLY FREE_SPACE
203 db_sync (DB *db, U32 flags = 0, SV *callback = &PL_sv_undef)
204 db_key_range (DB *db, DB_TXN_ornull *txn, SV *key, SV *key_range, U32 flags = 0, SV *callback = &PL_sv_undef)
205 db_put (DB *db, DB_TXN_ornull *txn, SV *key, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef)
206 flags: APPEND NODUPDATA NOOVERWRITE
207 db_get (DB *db, DB_TXN_ornull *txn, SV *key, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef)
208 flags: CONSUME CONSUME_WAIT GET_BOTH SET_RECNO MULTIPLE READ_COMMITTED READ_UNCOMMITTED RMW
209 db_pget (DB *db, DB_TXN_ornull *txn, SV *key, SV *pkey, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef)
210 flags: CONSUME CONSUME_WAIT GET_BOTH SET_RECNO MULTIPLE READ_COMMITTED READ_UNCOMMITTED RMW
211 db_del (DB *db, DB_TXN_ornull *txn, SV *key, U32 flags = 0, SV *callback = &PL_sv_undef)
212 db_txn_commit (DB_TXN *txn, U32 flags = 0, SV *callback = &PL_sv_undef)
213 flags: TXN_NOSYNC TXN_SYNC
214 db_txn_abort (DB_TXN *txn, SV *callback = &PL_sv_undef)
215
216 db_c_close (DBC *dbc, SV *callback = &PL_sv_undef)
217 db_c_count (DBC *dbc, SV *count, U32 flags = 0, SV *callback = &PL_sv_undef)
218 db_c_put (DBC *dbc, SV *key, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef)
219 flags: AFTER BEFORE CURRENT KEYFIRST KEYLAST NODUPDATA
220 db_c_get (DBC *dbc, SV *key, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef)
221 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
222 db_c_pget (DBC *dbc, SV *key, SV *pkey, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef)
223 db_c_del (DBC *dbc, U32 flags = 0, SV *callback = &PL_sv_undef)
224
225 db_sequence_open (DB_SEQUENCE *seq, DB_TXN_ornull *txnid, SV *key, U32 flags = 0, SV *callback = &PL_sv_undef)
226 flags: CREATE EXCL
227 db_sequence_close (DB_SEQUENCE *seq, U32 flags = 0, SV *callback = &PL_sv_undef)
228 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)
229 flags: TXN_NOSYNC
230 db_sequence_remove (DB_SEQUENCE *seq, DB_TXN_ornull *txnid = 0, U32 flags = 0, SV *callback = &PL_sv_undef)
231 flags: TXN_NOSYNC
232
233=head4 db_txn_finish (DB_TXN *txn, U32 flags = 0, SV *callback = &PL_sv_undef)
234
235This is not actually a Berkeley DB function but a BDB module
236extension. The background for this exytension is: It is very annoying to
237have to check every single BDB function for error returns and provide a
238codepath out of your transaction. While the BDB module still makes this
239possible, it contains the following extensions:
240
241When a transaction-protected function returns any operating system
242error (errno > 0), BDB will set the C<TXN_DEADLOCK> flag on the
243transaction. This flag is also set by Berkeley DB functions themselves
244when an operation fails with LOCK_DEADLOCK, and it causes all further
245operations on that transaction (including C<db_txn_commit>) to fail.
246
247The C<db_txn_finish> request will look at this flag, and, if it is set,
248will automatically call C<db_txn_abort> (setting errno to C<LOCK_DEADLOCK>
249if it isn't set to something else yet). If it isn't set, it will call
250C<db_txn_commit> and return the error normally.
251
252How to use this? Easy: just write your transaction normally:
253
254 my $txn = $db_env->txn_begin;
255 db_get $db, $txn, "key", my $data;
256 db_put $db, $txn, "key", $data + 1 unless $! == BDB::NOTFOUND;
257 db_txn_finish $txn;
258 die "transaction failed" if $!;
259
260That is, handle only the expected errors. If something unexpected happens
261(EIO, LOCK_NOTGRANTED or a deadlock in either db_get or db_put), then the remaining
262requests (db_put in this case) will simply be skipped (they will fail with
263LOCK_DEADLOCK) and the transaction will be aborted.
264
265You can use the C<< $txn->failed >> method to check wether a transaction
266has failed in this way and abort further processing (excluding
267C<db_txn_finish>).
268
269=head3 DB_ENV/database environment methods
270
271Methods available on DB_ENV/$env handles:
272
273 DESTROY (DB_ENV_ornull *env)
274 CODE:
275 if (env)
276 env->close (env, 0);
277
278 $int = $env->set_data_dir (const char *dir)
279 $int = $env->set_tmp_dir (const char *dir)
280 $int = $env->set_lg_dir (const char *dir)
281 $int = $env->set_shm_key (long shm_key)
282 $int = $env->set_cachesize (U32 gbytes, U32 bytes, int ncache = 0)
283 $int = $env->set_flags (U32 flags, int onoff = 1)
284 $int = $env->log_set_config (U32 flags, int onoff = 1) [v4.7]
285 $int = $env->set_intermediate_dir_mode (const char *modestring) [v4.7]
286 $env->set_errfile (FILE *errfile = 0)
287 $env->set_msgfile (FILE *msgfile = 0)
288 $int = $env->set_verbose (U32 which, int onoff = 1)
289 $int = $env->set_encrypt (const char *password, U32 flags = 0)
290 $int = $env->set_timeout (NV timeout_seconds, U32 flags = SET_TXN_TIMEOUT)
291 $int = $env->set_mp_max_openfd (int maxopenfd);
292 $int = $env->set_mp_max_write (int maxwrite, int maxwrite_sleep);
293 $int = $env->set_mp_mmapsize (int mmapsize_mb)
294 $int = $env->set_lk_detect (U32 detect = DB_LOCK_DEFAULT)
295 $int = $env->set_lk_max_lockers (U32 max)
296 $int = $env->set_lk_max_locks (U32 max)
297 $int = $env->set_lk_max_objects (U32 max)
298 $int = $env->set_lg_bsize (U32 max)
299 $int = $env->set_lg_max (U32 max)
300 $int = $env->mutex_set_increment (U32 increment)
301 $int = $env->mutex_set_tas_spins (U32 tas_spins)
302 $int = $env->mutex_set_max (U32 max)
303 $int = $env->mutex_set_align (U32 align)
304
305 $txn = $env->txn_begin (DB_TXN_ornull *parent = 0, U32 flags = 0)
306 flags: READ_COMMITTED READ_UNCOMMITTED TXN_NOSYNC TXN_NOWAIT TXN_SNAPSHOT TXN_SYNC TXN_WAIT TXN_WRITE_NOSYNC
307
308=head4 Example:
309
310 use AnyEvent;
311 use BDB;
312
313 our $FH; open $FH, "<&=" . BDB::poll_fileno;
314 our $WATCHER = AnyEvent->io (fh => $FH, poll => 'r', cb => \&BDB::poll_cb);
315
316 BDB::min_parallel 8;
317
318 my $env = db_env_create;
319
320 mkdir "bdtest", 0700;
321 db_env_open
322 $env,
323 "bdtest",
324 BDB::INIT_LOCK | BDB::INIT_LOG | BDB::INIT_MPOOL | BDB::INIT_TXN | BDB::RECOVER | BDB::USE_ENVIRON | BDB::CREATE,
325 0600;
326
327 $env->set_flags (BDB::AUTO_COMMIT | BDB::TXN_NOSYNC, 1);
328
329
330=head3 DB/database methods
331
332Methods available on DB/$db handles:
333
334 DESTROY (DB_ornull *db)
335 CODE:
336 if (db)
337 {
338 SV *env = (SV *)db->app_private;
339 db->close (db, 0);
340 SvREFCNT_dec (env);
341 }
342
343 $int = $db->set_cachesize (U32 gbytes, U32 bytes, int ncache = 0)
344 $int = $db->set_flags (U32 flags)
345 flags: CHKSUM ENCRYPT TXN_NOT_DURABLE
346 Btree: DUP DUPSORT RECNUM REVSPLITOFF
347 Hash: DUP DUPSORT
348 Queue: INORDER
349 Recno: RENUMBER SNAPSHOT
350
351 $int = $db->set_encrypt (const char *password, U32 flags)
352 $int = $db->set_lorder (int lorder)
353 $int = $db->set_bt_minkey (U32 minkey)
354 $int = $db->set_re_delim (int delim)
355 $int = $db->set_re_pad (int re_pad)
356 $int = $db->set_re_source (char *source)
357 $int = $db->set_re_len (U32 re_len)
358 $int = $db->set_h_ffactor (U32 h_ffactor)
359 $int = $db->set_h_nelem (U32 h_nelem)
360 $int = $db->set_q_extentsize (U32 extentsize)
361
362 $dbc = $db->cursor (DB_TXN_ornull *txn = 0, U32 flags = 0)
363 flags: READ_COMMITTED READ_UNCOMMITTED WRITECURSOR TXN_SNAPSHOT
364 $seq = $db->sequence (U32 flags = 0)
365
366=head4 Example:
367
368 my $db = db_create $env;
369 db_open $db, undef, "table", undef, BDB::BTREE, BDB::AUTO_COMMIT | BDB::CREATE | BDB::READ_UNCOMMITTED, 0600;
370
371 for (1..1000) {
372 db_put $db, undef, "key $_", "data $_";
373
374 db_key_range $db, undef, "key $_", my $keyrange;
375 my ($lt, $eq, $gt) = @$keyrange;
376 }
377
378 db_del $db, undef, "key $_" for 1..1000;
379
380 db_sync $db;
381
382
383=head3 DB_TXN/transaction methods
384
385Methods available on DB_TXN/$txn handles:
386
387 DESTROY (DB_TXN_ornull *txn)
388 CODE:
389 if (txn)
390 txn->abort (txn);
391
392 $int = $txn->set_timeout (NV timeout_seconds, U32 flags = SET_TXN_TIMEOUT)
393 flags: SET_LOCK_TIMEOUT SET_TXN_TIMEOUT
394
395 $bool = $txn->failed
396 # see db_txn_finish documentation, above
397
398
399=head3 DBC/cursor methods
400
401Methods available on DBC/$dbc handles:
402
403 DESTROY (DBC_ornull *dbc)
404 CODE:
405 if (dbc)
406 dbc->c_close (dbc);
407
408 $int = $cursor->set_priority ($priority = PRIORITY_*)
409
410=head4 Example:
411
412 my $c = $db->cursor;
413
414 for (;;) {
415 db_c_get $c, my $key, my $data, BDB::NEXT;
416 warn "<$!,$key,$data>";
417 last if $!;
418 }
419
420 db_c_close $c;
421
422
423=head3 DB_SEQUENCE/sequence methods
424
425Methods available on DB_SEQUENCE/$seq handles:
426
427 DESTROY (DB_SEQUENCE_ornull *seq)
428 CODE:
429 if (seq)
430 seq->close (seq, 0);
431
432 $int = $seq->initial_value (db_seq_t value)
433 $int = $seq->set_cachesize (U32 size)
434 $int = $seq->set_flags (U32 flags)
435 flags: SEQ_DEC SEQ_INC SEQ_WRAP
436 $int = $seq->set_range (db_seq_t min, db_seq_t max)
437
438=head4 Example:
439
440 my $seq = $db->sequence;
441
442 db_sequence_open $seq, undef, "seq", BDB::CREATE;
443 db_sequence_get $seq, undef, 1, my $value;
444
445
80=head2 SUPPORT FUNCTIONS 446=head2 SUPPORT FUNCTIONS
81 447
82=head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION 448=head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION
83 449
84=over 4 450=over 4
85 451
452=item $msg = BDB::strerror [$errno]
453
454Returns the string corresponding to the given errno value. If no argument
455is given, use C<$!>.
456
457Note that the BDB module also patches the C<$!> variable directly, so you
458should be able to get a bdb error string by simply stringifying C<$!>.
459
86=item $fileno = BDB::AIO::poll_fileno 460=item $fileno = BDB::poll_fileno
87 461
88Return the I<request result pipe file descriptor>. This filehandle must be 462Return the I<request result pipe file descriptor>. This filehandle must be
89polled for reading by some mechanism outside this module (e.g. Event or 463polled for reading by some mechanism outside this module (e.g. Event or
90select, see below or the SYNOPSIS). If the pipe becomes readable you have 464select, see below or the SYNOPSIS). If the pipe becomes readable you have
91to call C<poll_cb> to check the results. 465to call C<poll_cb> to check the results.
92 466
93See C<poll_cb> for an example. 467See C<poll_cb> for an example.
94 468
95=item BDB::AIO::poll_cb 469=item BDB::poll_cb
96 470
97Process some outstanding events on the result pipe. You have to call this 471Process some outstanding events on the result pipe. You have to call this
98regularly. Returns the number of events processed. Returns immediately 472regularly. Returns the number of events processed. Returns immediately
99when no events are outstanding. The amount of events processed depends on 473when no events are outstanding. The amount of events processed depends on
100the settings of C<BDB::AIO::max_poll_req> and C<BDB::AIO::max_poll_time>. 474the settings of C<BDB::max_poll_req> and C<BDB::max_poll_time>.
101 475
102If not all requests were processed for whatever reason, the filehandle 476If not all requests were processed for whatever reason, the filehandle
103will still be ready when C<poll_cb> returns. 477will still be ready when C<poll_cb> returns.
104 478
105Example: Install an Event watcher that automatically calls 479Example: Install an Event watcher that automatically calls
106BDB::AIO::poll_cb with high priority: 480BDB::poll_cb with high priority:
107 481
108 Event->io (fd => BDB::AIO::poll_fileno, 482 Event->io (fd => BDB::poll_fileno,
109 poll => 'r', async => 1, 483 poll => 'r', async => 1,
110 cb => \&BDB::AIO::poll_cb); 484 cb => \&BDB::poll_cb);
111 485
112=item BDB::AIO::max_poll_reqs $nreqs 486=item BDB::max_poll_reqs $nreqs
113 487
114=item BDB::AIO::max_poll_time $seconds 488=item BDB::max_poll_time $seconds
115 489
116These set the maximum number of requests (default C<0>, meaning infinity) 490These set the maximum number of requests (default C<0>, meaning infinity)
117that are being processed by C<BDB::AIO::poll_cb> in one call, respectively 491that are being processed by C<BDB::poll_cb> in one call, respectively
118the maximum amount of time (default C<0>, meaning infinity) spent in 492the maximum amount of time (default C<0>, meaning infinity) spent in
119C<BDB::AIO::poll_cb> to process requests (more correctly the mininum amount 493C<BDB::poll_cb> to process requests (more correctly the mininum amount
120of time C<poll_cb> is allowed to use). 494of time C<poll_cb> is allowed to use).
121 495
122Setting C<max_poll_time> to a non-zero value creates an overhead of one 496Setting C<max_poll_time> to a non-zero value creates an overhead of one
123syscall per request processed, which is not normally a problem unless your 497syscall per request processed, which is not normally a problem unless your
124callbacks are really really fast or your OS is really really slow (I am 498callbacks are really really fast or your OS is really really slow (I am
128interactiveness when perl is not fast enough to process all requests in 502interactiveness when perl is not fast enough to process all requests in
129time. 503time.
130 504
131For interactive programs, values such as C<0.01> to C<0.1> should be fine. 505For interactive programs, values such as C<0.01> to C<0.1> should be fine.
132 506
133Example: Install an Event watcher that automatically calls 507Example: Install an EV watcher that automatically calls
134BDB::AIO::poll_cb with low priority, to ensure that other parts of the 508BDB::poll_cb with low priority, to ensure that other parts of the
135program get the CPU sometimes even under high AIO load. 509program get the CPU sometimes even under high load.
136 510
137 # try not to spend much more than 0.1s in poll_cb 511 # try not to spend much more than 0.1s in poll_cb
138 BDB::AIO::max_poll_time 0.1; 512 BDB::max_poll_time 0.1;
139 513
140 # use a low priority so other tasks have priority 514 my $bdb_poll = EV::io BDB::poll_fileno, EV::READ, \&BDB::poll_cb);
141 Event->io (fd => BDB::AIO::poll_fileno,
142 poll => 'r', nice => 1,
143 cb => &BDB::AIO::poll_cb);
144 515
145=item BDB::AIO::poll_wait 516=item BDB::poll_wait
146 517
147If there are any outstanding requests and none of them in the result 518If there are any outstanding requests and none of them in the result
148phase, wait till the result filehandle becomes ready for reading (simply 519phase, wait till the result filehandle becomes ready for reading (simply
149does a C<select> on the filehandle. This is useful if you want to 520does a C<select> on the filehandle. This is useful if you want to
150synchronously wait for some requests to finish). 521synchronously wait for some requests to finish).
151 522
152See C<nreqs> for an example. 523See C<nreqs> for an example.
153 524
154=item BDB::AIO::poll 525=item BDB::poll
155 526
156Waits until some requests have been handled. 527Waits until some requests have been handled.
157 528
158Returns the number of requests processed, but is otherwise strictly 529Returns the number of requests processed, but is otherwise strictly
159equivalent to: 530equivalent to:
160 531
161 BDB::AIO::poll_wait, BDB::AIO::poll_cb 532 BDB::poll_wait, BDB::poll_cb
162 533
163=item BDB::AIO::flush 534=item BDB::flush
164 535
165Wait till all outstanding AIO requests have been handled. 536Wait till all outstanding BDB requests have been handled.
166 537
167Strictly equivalent to: 538Strictly equivalent to:
168 539
169 BDB::AIO::poll_wait, BDB::AIO::poll_cb 540 BDB::poll_wait, BDB::poll_cb
170 while BDB::AIO::nreqs; 541 while BDB::nreqs;
542
543=back
544
545=head3 VERSION CHECKING
546
547BerkeleyDB comes in various versions, many of them have minor
548incompatibilities. This means that traditional "at least version x.x"
549checks are often not sufficient.
550
551Example: set the log_autoremove option in a way compatible with <v.47 and
552v4.7. Note the use of & on the constants to avoid triggering a compiletime
553bug when the symbol isn't available.
554
555 $DB_ENV->set_flags (&BDB::LOG_AUTOREMOVE ) if BDB::VERSION v0, v4.7;
556 $DB_ENV->log_set_config (&BDB::LOG_AUTO_REMOVE) if BDB::VERSION v4.7;
557
558=over 4
559
560=item BDB::VERSION
561
562The C<BDB::VERSION> function, when called without arguments, returns the
563Berkeley DB version as a v-string (usually with 3 components). You should
564use C<lt> and C<ge> operators exclusively to make comparisons.
565
566Example: check for at least version 4.7.
567
568 BDB::VERSION ge v4.7 or die;
569
570=item BDB::VERSION min-version
571
572Returns true if the BDB version is at least the given version (specified
573as a v-string), false otherwise.
574
575Example: check for at least version 4.5.
576
577 BDB::VERSION v4.7 or die;
578
579=item BDB::VERSION min-version, max-version
580
581Returns true of the BDB version is at least version C<min-version> (specify C<undef> or C<v0> for any minimum version)
582and less then C<max-version>.
583
584Example: check wether version is strictly less then v4.7.
585
586 BDB::VERSION v0, v4.7
587 or die "version 4.7 is not yet supported";
588
589=back
590
591=cut
592
593sub VERSION {
594 if (@_ > 0) {
595 return undef if VERSION_v lt $_[0];
596 if (@_ > 1) {
597 return undef if VERSION_v ge $_[1];
598 }
599 }
600
601 VERSION_v
602}
171 603
172=head3 CONTROLLING THE NUMBER OF THREADS 604=head3 CONTROLLING THE NUMBER OF THREADS
173 605
606=over 4
607
174=item BDB::AIO::min_parallel $nthreads 608=item BDB::min_parallel $nthreads
175 609
176Set the minimum number of AIO threads to C<$nthreads>. The current 610Set the minimum number of BDB threads to C<$nthreads>. The current
177default is C<8>, which means eight asynchronous operations can execute 611default is C<8>, which means eight asynchronous operations can execute
178concurrently at any one time (the number of outstanding requests, 612concurrently at any one time (the number of outstanding requests,
179however, is unlimited). 613however, is unlimited).
180 614
181BDB::AIO starts threads only on demand, when an AIO request is queued and 615BDB starts threads only on demand, when an BDB request is queued and
182no free thread exists. Please note that queueing up a hundred requests can 616no free thread exists. Please note that queueing up a hundred requests can
183create demand for a hundred threads, even if it turns out that everything 617create demand for a hundred threads, even if it turns out that everything
184is in the cache and could have been processed faster by a single thread. 618is in the cache and could have been processed faster by a single thread.
185 619
186It is recommended to keep the number of threads relatively low, as some 620It is recommended to keep the number of threads relatively low, as some
189versions, 4-32 threads should be fine. 623versions, 4-32 threads should be fine.
190 624
191Under most circumstances you don't need to call this function, as the 625Under most circumstances you don't need to call this function, as the
192module selects a default that is suitable for low to moderate load. 626module selects a default that is suitable for low to moderate load.
193 627
194=item BDB::AIO::max_parallel $nthreads 628=item BDB::max_parallel $nthreads
195 629
196Sets the maximum number of AIO threads to C<$nthreads>. If more than the 630Sets the maximum number of BDB threads to C<$nthreads>. If more than the
197specified number of threads are currently running, this function kills 631specified number of threads are currently running, this function kills
198them. This function blocks until the limit is reached. 632them. This function blocks until the limit is reached.
199 633
200While C<$nthreads> are zero, aio requests get queued but not executed 634While C<$nthreads> are zero, aio requests get queued but not executed
201until the number of threads has been increased again. 635until the number of threads has been increased again.
203This module automatically runs C<max_parallel 0> at program end, to ensure 637This module automatically runs C<max_parallel 0> at program end, to ensure
204that all threads are killed and that there are no outstanding requests. 638that all threads are killed and that there are no outstanding requests.
205 639
206Under normal circumstances you don't need to call this function. 640Under normal circumstances you don't need to call this function.
207 641
208=item BDB::AIO::max_idle $nthreads 642=item BDB::max_idle $nthreads
209 643
210Limit the number of threads (default: 4) that are allowed to idle (i.e., 644Limit the number of threads (default: 4) that are allowed to idle (i.e.,
211threads that did not get a request to process within 10 seconds). That 645threads that did not get a request to process within 10 seconds). That
212means if a thread becomes idle while C<$nthreads> other threads are also 646means if a thread becomes idle while C<$nthreads> other threads are also
213idle, it will free its resources and exit. 647idle, it will free its resources and exit.
218 652
219The default is probably ok in most situations, especially if thread 653The default is probably ok in most situations, especially if thread
220creation is fast. If thread creation is very slow on your system you might 654creation is fast. If thread creation is very slow on your system you might
221want to use larger values. 655want to use larger values.
222 656
223=item $oldmaxreqs = BDB::AIO::max_outstanding $maxreqs 657=item $oldmaxreqs = BDB::max_outstanding $maxreqs
224 658
225This is a very bad function to use in interactive programs because it 659This is a very bad function to use in interactive programs because it
226blocks, and a bad way to reduce concurrency because it is inexact: Better 660blocks, and a bad way to reduce concurrency because it is inexact: Better
227use an C<aio_group> together with a feed callback. 661use an C<aio_group> together with a feed callback.
228 662
236 670
237You can still queue as many requests as you want. Therefore, 671You can still queue as many requests as you want. Therefore,
238C<max_oustsanding> is mainly useful in simple scripts (with low values) or 672C<max_oustsanding> is mainly useful in simple scripts (with low values) or
239as a stop gap to shield against fatal memory overflow (with large values). 673as a stop gap to shield against fatal memory overflow (with large values).
240 674
675=item BDB::set_sync_prepare $cb
676
677Sets a callback that is called whenever a request is created without an
678explicit callback. It has to return two code references. The first is used
679as the request callback (it should save the return status), and the second
680is called to wait until the first callback has been called (it must set
681C<$!> to the return status).
682
683This mechanism can be used to include BDB into other event mechanisms,
684such as L<AnyEvent::BDB> or L<Coro::BDB>.
685
686The default implementation works like this:
687
688 sub {
689 my $status;
690 (
691 sub { $status = $! },
692 sub { BDB::poll while !defined $status; $! = $status },
693 )
694 }
695
696It simply blocks the process till the request has finished and then sets
697C<$!> to the return value. This means that if you don't use a callback,
698BDB will simply fall back to synchronous operations.
699
700=back
701
241=head3 STATISTICAL INFORMATION 702=head3 STATISTICAL INFORMATION
242 703
704=over 4
705
243=item BDB::AIO::nreqs 706=item BDB::nreqs
244 707
245Returns the number of requests currently in the ready, execute or pending 708Returns the number of requests currently in the ready, execute or pending
246states (i.e. for which their callback has not been invoked yet). 709states (i.e. for which their callback has not been invoked yet).
247 710
248Example: wait till there are no outstanding requests anymore: 711Example: wait till there are no outstanding requests anymore:
249 712
250 BDB::AIO::poll_wait, BDB::AIO::poll_cb 713 BDB::poll_wait, BDB::poll_cb
251 while BDB::AIO::nreqs; 714 while BDB::nreqs;
252 715
253=item BDB::AIO::nready 716=item BDB::nready
254 717
255Returns the number of requests currently in the ready state (not yet 718Returns the number of requests currently in the ready state (not yet
256executed). 719executed).
257 720
258=item BDB::AIO::npending 721=item BDB::npending
259 722
260Returns the number of requests currently in the pending state (executed, 723Returns the number of requests currently in the pending state (executed,
261but not yet processed by poll_cb). 724but not yet processed by poll_cb).
262 725
263=back 726=back
264 727
265=cut 728=cut
266 729
267# support function to convert a fd into a perl filehandle 730set_sync_prepare {
268sub _fd2fh { 731 my $status;
269 return undef if $_[0] < 0; 732 (
270 733 sub {
271 # try to generate nice filehandles 734 $status = $!;
272 my $sym = "BDB::AIO::fd#$_[0]"; 735 },
273 local *$sym; 736 sub {
274 737 BDB::poll while !defined $status;
275 open *$sym, "+<&=$_[0]" # usually works under any unix 738 $! = $status;
276 or open *$sym, "<&=$_[0]" # cygwin needs this 739 },
277 or open *$sym, ">&=$_[0]" # or this 740 )
278 or return undef; 741};
279
280 *$sym
281}
282 742
283min_parallel 8; 743min_parallel 8;
284 744
285END { flush } 745END { flush }
286 746
288 748
289=head2 FORK BEHAVIOUR 749=head2 FORK BEHAVIOUR
290 750
291This module should do "the right thing" when the process using it forks: 751This module should do "the right thing" when the process using it forks:
292 752
293Before the fork, IO::AIO enters a quiescent state where no requests 753Before the fork, BDB enters a quiescent state where no requests
294can be added in other threads and no results will be processed. After 754can be added in other threads and no results will be processed. After
295the fork the parent simply leaves the quiescent state and continues 755the fork the parent simply leaves the quiescent state and continues
296request/result processing, while the child frees the request/result queue 756request/result processing, while the child frees the request/result queue
297(so that the requests started before the fork will only be handled in the 757(so that the requests started before the fork will only be handled in the
298parent). Threads will be started on demand until the limit set in the 758parent). Threads will be started on demand until the limit set in the
299parent process has been reached again. 759parent process has been reached again.
300 760
301In short: the parent will, after a short pause, continue as if fork had 761In short: the parent will, after a short pause, continue as if fork had
302not been called, while the child will act as if IO::AIO has not been used 762not been called, while the child will act as if BDB has not been used
303yet. 763yet.
764
765Win32 note: there is no fork on win32, and perls emulation of it is too
766broken to be supported, so do not use BDB in a windows pseudo-fork, better
767yet, switch to a more capable platform.
304 768
305=head2 MEMORY USAGE 769=head2 MEMORY USAGE
306 770
307Per-request usage: 771Per-request usage:
308 772
310bytes of memory. In addition, stat requests need a stat buffer (possibly 774bytes of memory. In addition, stat requests need a stat buffer (possibly
311a few hundred bytes), readdir requires a result buffer and so on. Perl 775a few hundred bytes), readdir requires a result buffer and so on. Perl
312scalars and other data passed into aio requests will also be locked and 776scalars and other data passed into aio requests will also be locked and
313will consume memory till the request has entered the done state. 777will consume memory till the request has entered the done state.
314 778
315This is now awfully much, so queuing lots of requests is not usually a 779This is not awfully much, so queuing lots of requests is not usually a
316problem. 780problem.
317 781
318Per-thread usage: 782Per-thread usage:
319 783
320In the execution phase, some aio requests require more memory for 784In the execution phase, some aio requests require more memory for
321temporary buffers, and each thread requires a stack and other data 785temporary buffers, and each thread requires a stack and other data
322structures (usually around 16k-128k, depending on the OS). 786structures (usually around 16k-128k, depending on the OS).
323 787
324=head1 KNOWN BUGS 788=head1 KNOWN BUGS
325 789
326Known bugs will be fixed in the next release. 790Known bugs will be fixed in the next release, except:
791
792 If you use a transaction in any request, and the request returns
793 with an operating system error or DB_LOCK_NOTGRANTED, the internal
794 TXN_DEADLOCK flag will be set on the transaction. See C<db_txn_finish>,
795 above.
327 796
328=head1 SEE ALSO 797=head1 SEE ALSO
329 798
330L<Coro::AIO>. 799L<AnyEvent::BDB> (event loop integration), L<Coro::BDB> (more natural
800syntax), L<IO::AIO> (nice to have).
331 801
332=head1 AUTHOR 802=head1 AUTHOR
333 803
334 Marc Lehmann <schmorp@schmorp.de> 804 Marc Lehmann <schmorp@schmorp.de>
335 http://home.schmorp.de/ 805 http://home.schmorp.de/

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines