… | |
… | |
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 integration with AnyEvent: |
|
|
33 | use AnyEvent::BDB; |
|
|
34 | |
|
|
35 | # automatic result processing with EV: |
|
|
36 | my $WATCHER = EV::io BDB::poll_fileno, EV::READ, \&BDB::poll_cb; |
|
|
37 | |
|
|
38 | # with Glib: |
|
|
39 | add_watch Glib::IO BDB::poll_fileno, |
|
|
40 | in => sub { BDB::poll_cb; 1 }; |
|
|
41 | |
|
|
42 | # or simply flush manually |
|
|
43 | BDB::flush; |
|
|
44 | |
|
|
45 | |
9 | =head1 DESCRIPTION |
46 | =head1 DESCRIPTION |
10 | |
47 | |
11 | See the eg/ directory in the distribution and the berkeleydb C |
48 | See the BerkeleyDB documentation (L<http://www.oracle.com/technology/documentation/berkeley-db/db/index.html>). |
12 | documentation. This is inadequate, but the only sources of documentation |
49 | The BDB API is very similar to the C API (the translation has been very faithful). |
13 | known for this module so far. |
|
|
14 | |
50 | |
15 | =head2 EXAMPLE |
51 | See also the example sections in the document below and possibly the eg/ |
|
|
52 | subdirectory of the BDB distribution. Last not least see the IO::AIO |
|
|
53 | documentation, as that module uses almost the same asynchronous request |
|
|
54 | model as this module. |
|
|
55 | |
|
|
56 | I 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 | |
19 | Every request method creates a request. which is a C data structure not |
61 | Every request method creates a request. which is a C data structure not |
20 | directly visible to Perl. |
62 | directly visible to Perl. |
… | |
… | |
61 | |
103 | |
62 | =cut |
104 | =cut |
63 | |
105 | |
64 | package BDB; |
106 | package BDB; |
65 | |
107 | |
66 | no warnings; |
108 | use common::sense; |
67 | use strict 'vars'; |
|
|
68 | |
109 | |
69 | use base 'Exporter'; |
110 | use base 'Exporter'; |
70 | |
111 | |
|
|
112 | our $VERSION; |
|
|
113 | |
71 | BEGIN { |
114 | BEGIN { |
72 | our $VERSION = '0.1'; |
115 | $VERSION = '1.91'; |
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 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 |
120 | db_env_log_archive |
|
|
121 | db_open db_close db_compact db_sync db_verify 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", $VERSION); |
93 | } |
138 | } |
94 | |
139 | |
|
|
140 | =head1 BERKELEYDB FUNCTIONS |
|
|
141 | |
|
|
142 | All of these are functions. The create functions simply return a new |
|
|
143 | object and never block. All the remaining functions take an optional |
|
|
144 | callback as last argument. If it is missing, then the function will be |
|
|
145 | executed synchronously. In both cases, C<$!> will reflect the return value |
|
|
146 | of the function. |
|
|
147 | |
|
|
148 | BDB functions that cannot block (mostly functions that manipulate |
|
|
149 | settings) are method calls on the relevant objects, so the rule of thumb |
|
|
150 | is: if it's a method, it's not blocking, if it's a function, it takes a |
|
|
151 | callback as last argument. |
|
|
152 | |
|
|
153 | In the following, C<$int> signifies an integer return value, |
|
|
154 | C<bdb_filename> is a "filename" (octets on unix, madness on windows), |
|
|
155 | C<U32> is an unsigned 32 bit integer, C<int> is some integer, C<NV> is a |
|
|
156 | floating point value. |
|
|
157 | |
|
|
158 | Most C<SV *> types are generic perl scalars (for input and output of data |
|
|
159 | values). |
|
|
160 | |
|
|
161 | The various C<DB_ENV> etc. arguments are handles return by |
|
|
162 | C<db_env_create>, C<db_create>, C<txn_begin> and so on. If they have an |
|
|
163 | appended C<_ornull> this means they are optional and you can pass C<undef> |
|
|
164 | for them, resulting a NULL pointer on the C level. |
|
|
165 | |
|
|
166 | The C<SV *callback> is the optional callback function to call when the |
|
|
167 | request is completed. This last callback argument is special: the callback |
|
|
168 | is simply the last argument passed. If there are "optional" arguments |
|
|
169 | before the callback they can be left out. The callback itself can be left |
|
|
170 | out or specified as C<undef>, in which case the function will be executed |
|
|
171 | synchronously. |
|
|
172 | |
|
|
173 | For example, C<db_env_txn_checkpoint> usually is called with all integer |
|
|
174 | arguments zero. These can be left out, so all of these specify a call |
|
|
175 | to C<< DB_ENV->txn_checkpoint >>, to be executed asynchronously with a |
|
|
176 | callback to be called: |
|
|
177 | |
|
|
178 | db_env_txn_checkpoint $db_env, 0, 0, 0, sub { }; |
|
|
179 | db_env_txn_checkpoint $db_env, 0, 0, sub { }; |
|
|
180 | db_env_txn_checkpoint $db_env, sub { }; |
|
|
181 | |
|
|
182 | While these all specify a call to C<< DB_ENV->txn_checkpoint >> to be |
|
|
183 | executed synchronously: |
|
|
184 | |
|
|
185 | db_env_txn_checkpoint $db_env, 0, 0, 0, undef; |
|
|
186 | db_env_txn_checkpoint $db_env, 0, 0, 0; |
|
|
187 | db_env_txn_checkpoint $db_env, 0; |
|
|
188 | |
|
|
189 | =head2 BDB functions |
|
|
190 | |
|
|
191 | Functions in the BDB namespace, exported by default: |
|
|
192 | |
|
|
193 | $env = db_env_create (U32 env_flags = 0) |
|
|
194 | flags: RPCCLIENT |
|
|
195 | |
|
|
196 | db_env_open (DB_ENV *env, bdb_filename db_home, U32 open_flags, int mode, SV *callback = 0) |
|
|
197 | 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 |
|
|
198 | db_env_close (DB_ENV *env, U32 flags = 0, SV *callback = 0) |
|
|
199 | db_env_txn_checkpoint (DB_ENV *env, U32 kbyte = 0, U32 min = 0, U32 flags = 0, SV *callback = 0) |
|
|
200 | flags: FORCE |
|
|
201 | db_env_lock_detect (DB_ENV *env, U32 flags = 0, U32 atype = DB_LOCK_DEFAULT, SV *dummy = 0, SV *callback = 0) |
|
|
202 | atype: LOCK_DEFAULT LOCK_EXPIRE LOCK_MAXLOCKS LOCK_MAXWRITE LOCK_MINLOCKS LOCK_MINWRITE LOCK_OLDEST LOCK_RANDOM LOCK_YOUNGEST |
|
|
203 | db_env_memp_sync (DB_ENV *env, SV *dummy = 0, SV *callback = 0) |
|
|
204 | db_env_memp_trickle (DB_ENV *env, int percent, SV *dummy = 0, SV *callback = 0) |
|
|
205 | db_env_dbremove (DB_ENV *env, DB_TXN_ornull *txnid, bdb_filename file, bdb_filename database, U32 flags = 0, SV *callback = 0) |
|
|
206 | db_env_dbrename (DB_ENV *env, DB_TXN_ornull *txnid, bdb_filename file, bdb_filename database, bdb_filename newname, U32 flags = 0, SV *callback = 0) |
|
|
207 | db_env_log_archive (DB_ENV *env, SV *listp, U32 flags = 0, SV *callback = 0) |
|
|
208 | |
|
|
209 | $db = db_create (DB_ENV *env = 0, U32 flags = 0) |
|
|
210 | flags: XA_CREATE |
|
|
211 | |
|
|
212 | db_open (DB *db, DB_TXN_ornull *txnid, bdb_filename file, bdb_filename database, int type, U32 flags, int mode, SV *callback = 0) |
|
|
213 | flags: AUTO_COMMIT CREATE EXCL MULTIVERSION NOMMAP RDONLY READ_UNCOMMITTED THREAD TRUNCATE |
|
|
214 | db_close (DB *db, U32 flags = 0, SV *callback = 0) |
|
|
215 | flags: DB_NOSYNC |
|
|
216 | db_verify (DB *db, bdb_filename file, bdb_filename database = 0, SV *dummy = 0, U32 flags = 0, SV *callback = 0) |
|
|
217 | db_upgrade (DB *db, bdb_filename file, U32 flags = 0, SV *callback = 0) |
|
|
218 | db_compact (DB *db, DB_TXN_ornull *txn = 0, SV *start = 0, SV *stop = 0, SV *unused1 = 0, U32 flags = DB_FREE_SPACE, SV *unused2 = 0, SV *callback = 0) |
|
|
219 | flags: FREELIST_ONLY FREE_SPACE |
|
|
220 | db_sync (DB *db, U32 flags = 0, SV *callback = 0) |
|
|
221 | db_key_range (DB *db, DB_TXN_ornull *txn, SV *key, SV *key_range, U32 flags = 0, SV *callback = 0) |
|
|
222 | db_put (DB *db, DB_TXN_ornull *txn, SV *key, SV *data, U32 flags = 0, SV *callback = 0) |
|
|
223 | flags: APPEND NODUPDATA NOOVERWRITE |
|
|
224 | db_exists (DB *db, DB_TXN_ornull *txn, SV *key, U32 flags = 0, SV *callback = 0) (v4.6) |
|
|
225 | db_get (DB *db, DB_TXN_ornull *txn, SV *key, SV *data, U32 flags = 0, SV *callback = 0) |
|
|
226 | flags: CONSUME CONSUME_WAIT GET_BOTH SET_RECNO MULTIPLE READ_COMMITTED READ_UNCOMMITTED RMW |
|
|
227 | db_pget (DB *db, DB_TXN_ornull *txn, SV *key, SV *pkey, SV *data, U32 flags = 0, SV *callback = 0) |
|
|
228 | flags: CONSUME CONSUME_WAIT GET_BOTH SET_RECNO MULTIPLE READ_COMMITTED READ_UNCOMMITTED RMW |
|
|
229 | db_del (DB *db, DB_TXN_ornull *txn, SV *key, U32 flags = 0, SV *callback = 0) |
|
|
230 | db_txn_commit (DB_TXN *txn, U32 flags = 0, SV *callback = 0) |
|
|
231 | flags: TXN_NOSYNC TXN_SYNC |
|
|
232 | db_txn_abort (DB_TXN *txn, SV *callback = 0) |
|
|
233 | |
|
|
234 | db_c_close (DBC *dbc, SV *callback = 0) |
|
|
235 | db_c_count (DBC *dbc, SV *count, U32 flags = 0, SV *callback = 0) |
|
|
236 | db_c_put (DBC *dbc, SV *key, SV *data, U32 flags = 0, SV *callback = 0) |
|
|
237 | flags: AFTER BEFORE CURRENT KEYFIRST KEYLAST NODUPDATA |
|
|
238 | db_c_get (DBC *dbc, SV *key, SV *data, U32 flags = 0, SV *callback = 0) |
|
|
239 | flags: CURRENT FIRST GET_BOTH GET_BOTH_RANGE GET_RECNO JOIN_ITEM LAST NEXT NEXT_DUP NEXT_NODUP PREV PREV_DUP PREV_NODUP SET SET_RANGE SET_RECNO READ_UNCOMMITTED MULTIPLE MULTIPLE_KEY RMW |
|
|
240 | db_c_pget (DBC *dbc, SV *key, SV *pkey, SV *data, U32 flags = 0, SV *callback = 0) |
|
|
241 | db_c_del (DBC *dbc, U32 flags = 0, SV *callback = 0) |
|
|
242 | |
|
|
243 | db_sequence_open (DB_SEQUENCE *seq, DB_TXN_ornull *txnid, SV *key, U32 flags = 0, SV *callback = 0) |
|
|
244 | flags: CREATE EXCL |
|
|
245 | db_sequence_close (DB_SEQUENCE *seq, U32 flags = 0, SV *callback = 0) |
|
|
246 | db_sequence_get (DB_SEQUENCE *seq, DB_TXN_ornull *txnid, int delta, SV *seq_value, U32 flags = DB_TXN_NOSYNC, SV *callback = 0) |
|
|
247 | flags: TXN_NOSYNC |
|
|
248 | db_sequence_remove (DB_SEQUENCE *seq, DB_TXN_ornull *txnid = 0, U32 flags = 0, SV *callback = 0) |
|
|
249 | flags: TXN_NOSYNC |
|
|
250 | |
|
|
251 | =head3 db_txn_finish (DB_TXN *txn, U32 flags = 0, SV *callback = 0) |
|
|
252 | |
|
|
253 | This is not actually a Berkeley DB function but a BDB module |
|
|
254 | extension. The background for this exytension is: It is very annoying to |
|
|
255 | have to check every single BDB function for error returns and provide a |
|
|
256 | codepath out of your transaction. While the BDB module still makes this |
|
|
257 | possible, it contains the following extensions: |
|
|
258 | |
|
|
259 | When a transaction-protected function returns any operating system |
|
|
260 | error (errno > 0), BDB will set the C<TXN_DEADLOCK> flag on the |
|
|
261 | transaction. This flag is also set by Berkeley DB functions themselves |
|
|
262 | when an operation fails with LOCK_DEADLOCK, and it causes all further |
|
|
263 | operations on that transaction (including C<db_txn_commit>) to fail. |
|
|
264 | |
|
|
265 | The C<db_txn_finish> request will look at this flag, and, if it is set, |
|
|
266 | will automatically call C<db_txn_abort> (setting errno to C<LOCK_DEADLOCK> |
|
|
267 | if it isn't set to something else yet). If it isn't set, it will call |
|
|
268 | C<db_txn_commit> and return the error normally. |
|
|
269 | |
|
|
270 | How to use this? Easy: just write your transaction normally: |
|
|
271 | |
|
|
272 | my $txn = $db_env->txn_begin; |
|
|
273 | db_get $db, $txn, "key", my $data; |
|
|
274 | db_put $db, $txn, "key", $data + 1 unless $! == BDB::NOTFOUND; |
|
|
275 | db_txn_finish $txn; |
|
|
276 | die "transaction failed" if $!; |
|
|
277 | |
|
|
278 | That is, handle only the expected errors. If something unexpected happens |
|
|
279 | (EIO, LOCK_NOTGRANTED or a deadlock in either db_get or db_put), then the remaining |
|
|
280 | requests (db_put in this case) will simply be skipped (they will fail with |
|
|
281 | LOCK_DEADLOCK) and the transaction will be aborted. |
|
|
282 | |
|
|
283 | You can use the C<< $txn->failed >> method to check wether a transaction |
|
|
284 | has failed in this way and abort further processing (excluding |
|
|
285 | C<db_txn_finish>). |
|
|
286 | |
|
|
287 | |
|
|
288 | =head2 DB_ENV/database environment methods |
|
|
289 | |
|
|
290 | Methods available on DB_ENV/$env handles: |
|
|
291 | |
|
|
292 | DESTROY (DB_ENV_ornull *env) |
|
|
293 | CODE: |
|
|
294 | if (env) |
|
|
295 | env->close (env, 0); |
|
|
296 | |
|
|
297 | $int = $env->set_data_dir (const char *dir) |
|
|
298 | $int = $env->set_tmp_dir (const char *dir) |
|
|
299 | $int = $env->set_lg_dir (const char *dir) |
|
|
300 | $int = $env->set_shm_key (long shm_key) |
|
|
301 | $int = $env->set_cachesize (U32 gbytes, U32 bytes, int ncache = 0) |
|
|
302 | $int = $env->set_flags (U32 flags, int onoff = 1) |
|
|
303 | $int = $env->log_set_config (U32 flags, int onoff = 1) (v4.7) |
|
|
304 | $int = $env->set_intermediate_dir_mode (const char *modestring) (v4.7) |
|
|
305 | $env->set_errfile (FILE *errfile = 0) |
|
|
306 | $env->set_msgfile (FILE *msgfile = 0) |
|
|
307 | $int = $env->set_verbose (U32 which, int onoff = 1) |
|
|
308 | $int = $env->set_encrypt (const char *password, U32 flags = 0) |
|
|
309 | $int = $env->set_timeout (NV timeout_seconds, U32 flags = SET_TXN_TIMEOUT) |
|
|
310 | $int = $env->set_mp_max_openfd (int maxopenfd); |
|
|
311 | $int = $env->set_mp_max_write (int maxwrite, int maxwrite_sleep); |
|
|
312 | $int = $env->set_mp_mmapsize (int mmapsize_mb) |
|
|
313 | $int = $env->set_lk_detect (U32 detect = DB_LOCK_DEFAULT) |
|
|
314 | $int = $env->set_lk_max_lockers (U32 max) |
|
|
315 | $int = $env->set_lk_max_locks (U32 max) |
|
|
316 | $int = $env->set_lk_max_objects (U32 max) |
|
|
317 | $int = $env->set_lg_bsize (U32 max) |
|
|
318 | $int = $env->set_lg_max (U32 max) |
|
|
319 | $int = $env->mutex_set_increment (U32 increment) |
|
|
320 | $int = $env->mutex_set_tas_spins (U32 tas_spins) |
|
|
321 | $int = $env->mutex_set_max (U32 max) |
|
|
322 | $int = $env->mutex_set_align (U32 align) |
|
|
323 | |
|
|
324 | $txn = $env->txn_begin (DB_TXN_ornull *parent = 0, U32 flags = 0) |
|
|
325 | flags: READ_COMMITTED READ_UNCOMMITTED TXN_NOSYNC TXN_NOWAIT TXN_SNAPSHOT TXN_SYNC TXN_WAIT TXN_WRITE_NOSYNC |
|
|
326 | $txn = $env->cdsgroup_begin; (v4.5) |
|
|
327 | |
|
|
328 | =head3 Example: |
|
|
329 | |
|
|
330 | use AnyEvent; |
|
|
331 | use BDB; |
|
|
332 | |
|
|
333 | our $FH; open $FH, "<&=" . BDB::poll_fileno; |
|
|
334 | our $WATCHER = AnyEvent->io (fh => $FH, poll => 'r', cb => \&BDB::poll_cb); |
|
|
335 | |
|
|
336 | BDB::min_parallel 8; |
|
|
337 | |
|
|
338 | my $env = db_env_create; |
|
|
339 | |
|
|
340 | mkdir "bdtest", 0700; |
|
|
341 | db_env_open |
|
|
342 | $env, |
|
|
343 | "bdtest", |
|
|
344 | BDB::INIT_LOCK | BDB::INIT_LOG | BDB::INIT_MPOOL | BDB::INIT_TXN | BDB::RECOVER | BDB::USE_ENVIRON | BDB::CREATE, |
|
|
345 | 0600; |
|
|
346 | |
|
|
347 | $env->set_flags (BDB::AUTO_COMMIT | BDB::TXN_NOSYNC, 1); |
|
|
348 | |
|
|
349 | |
|
|
350 | =head2 DB/database methods |
|
|
351 | |
|
|
352 | Methods available on DB/$db handles: |
|
|
353 | |
|
|
354 | DESTROY (DB_ornull *db) |
|
|
355 | CODE: |
|
|
356 | if (db) |
|
|
357 | { |
|
|
358 | SV *env = (SV *)db->app_private; |
|
|
359 | db->close (db, 0); |
|
|
360 | SvREFCNT_dec (env); |
|
|
361 | } |
|
|
362 | |
|
|
363 | $int = $db->set_cachesize (U32 gbytes, U32 bytes, int ncache = 0) |
|
|
364 | $int = $db->set_flags (U32 flags) |
|
|
365 | flags: CHKSUM ENCRYPT TXN_NOT_DURABLE |
|
|
366 | Btree: DUP DUPSORT RECNUM REVSPLITOFF |
|
|
367 | Hash: DUP DUPSORT |
|
|
368 | Queue: INORDER |
|
|
369 | Recno: RENUMBER SNAPSHOT |
|
|
370 | |
|
|
371 | $int = $db->set_encrypt (const char *password, U32 flags) |
|
|
372 | $int = $db->set_lorder (int lorder) |
|
|
373 | $int = $db->set_bt_minkey (U32 minkey) |
|
|
374 | $int = $db->set_re_delim (int delim) |
|
|
375 | $int = $db->set_re_pad (int re_pad) |
|
|
376 | $int = $db->set_re_source (char *source) |
|
|
377 | $int = $db->set_re_len (U32 re_len) |
|
|
378 | $int = $db->set_h_ffactor (U32 h_ffactor) |
|
|
379 | $int = $db->set_h_nelem (U32 h_nelem) |
|
|
380 | $int = $db->set_q_extentsize (U32 extentsize) |
|
|
381 | |
|
|
382 | $dbc = $db->cursor (DB_TXN_ornull *txn = 0, U32 flags = 0) |
|
|
383 | flags: READ_COMMITTED READ_UNCOMMITTED WRITECURSOR TXN_SNAPSHOT |
|
|
384 | $seq = $db->sequence (U32 flags = 0) |
|
|
385 | |
|
|
386 | =head3 Example: |
|
|
387 | |
|
|
388 | my $db = db_create $env; |
|
|
389 | db_open $db, undef, "table", undef, BDB::BTREE, BDB::AUTO_COMMIT | BDB::CREATE | BDB::READ_UNCOMMITTED, 0600; |
|
|
390 | |
|
|
391 | for (1..1000) { |
|
|
392 | db_put $db, undef, "key $_", "data $_"; |
|
|
393 | |
|
|
394 | db_key_range $db, undef, "key $_", my $keyrange; |
|
|
395 | my ($lt, $eq, $gt) = @$keyrange; |
|
|
396 | } |
|
|
397 | |
|
|
398 | db_del $db, undef, "key $_" for 1..1000; |
|
|
399 | |
|
|
400 | db_sync $db; |
|
|
401 | |
|
|
402 | |
|
|
403 | =head2 DB_TXN/transaction methods |
|
|
404 | |
|
|
405 | Methods available on DB_TXN/$txn handles: |
|
|
406 | |
|
|
407 | DESTROY (DB_TXN_ornull *txn) |
|
|
408 | CODE: |
|
|
409 | if (txn) |
|
|
410 | txn->abort (txn); |
|
|
411 | |
|
|
412 | $int = $txn->set_timeout (NV timeout_seconds, U32 flags = SET_TXN_TIMEOUT) |
|
|
413 | flags: SET_LOCK_TIMEOUT SET_TXN_TIMEOUT |
|
|
414 | |
|
|
415 | $bool = $txn->failed |
|
|
416 | # see db_txn_finish documentation, above |
|
|
417 | |
|
|
418 | |
|
|
419 | =head2 DBC/cursor methods |
|
|
420 | |
|
|
421 | Methods available on DBC/$dbc handles: |
|
|
422 | |
|
|
423 | DESTROY (DBC_ornull *dbc) |
|
|
424 | CODE: |
|
|
425 | if (dbc) |
|
|
426 | dbc->c_close (dbc); |
|
|
427 | |
|
|
428 | $int = $cursor->set_priority ($priority = PRIORITY_*) (v4.6) |
|
|
429 | |
|
|
430 | =head3 Example: |
|
|
431 | |
|
|
432 | my $c = $db->cursor; |
|
|
433 | |
|
|
434 | for (;;) { |
|
|
435 | db_c_get $c, my $key, my $data, BDB::NEXT; |
|
|
436 | warn "<$!,$key,$data>"; |
|
|
437 | last if $!; |
|
|
438 | } |
|
|
439 | |
|
|
440 | db_c_close $c; |
|
|
441 | |
|
|
442 | |
|
|
443 | =head2 DB_SEQUENCE/sequence methods |
|
|
444 | |
|
|
445 | Methods available on DB_SEQUENCE/$seq handles: |
|
|
446 | |
|
|
447 | DESTROY (DB_SEQUENCE_ornull *seq) |
|
|
448 | CODE: |
|
|
449 | if (seq) |
|
|
450 | seq->close (seq, 0); |
|
|
451 | |
|
|
452 | $int = $seq->initial_value (db_seq_t value) |
|
|
453 | $int = $seq->set_cachesize (U32 size) |
|
|
454 | $int = $seq->set_flags (U32 flags) |
|
|
455 | flags: SEQ_DEC SEQ_INC SEQ_WRAP |
|
|
456 | $int = $seq->set_range (db_seq_t min, db_seq_t max) |
|
|
457 | |
|
|
458 | =head3 Example: |
|
|
459 | |
|
|
460 | my $seq = $db->sequence; |
|
|
461 | |
|
|
462 | db_sequence_open $seq, undef, "seq", BDB::CREATE; |
|
|
463 | db_sequence_get $seq, undef, 1, my $value; |
|
|
464 | |
|
|
465 | |
95 | =head2 SUPPORT FUNCTIONS |
466 | =head1 SUPPORT FUNCTIONS |
96 | |
467 | |
97 | =head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION |
468 | =head2 EVENT PROCESSING AND EVENT LOOP INTEGRATION |
98 | |
469 | |
99 | =over 4 |
470 | =over 4 |
|
|
471 | |
|
|
472 | =item $msg = BDB::strerror [$errno] |
|
|
473 | |
|
|
474 | Returns the string corresponding to the given errno value. If no argument |
|
|
475 | is given, use C<$!>. |
|
|
476 | |
|
|
477 | Note that the BDB module also patches the C<$!> variable directly, so you |
|
|
478 | should be able to get a bdb error string by simply stringifying C<$!>. |
100 | |
479 | |
101 | =item $fileno = BDB::poll_fileno |
480 | =item $fileno = BDB::poll_fileno |
102 | |
481 | |
103 | Return the I<request result pipe file descriptor>. This filehandle must be |
482 | Return the I<request result pipe file descriptor>. This filehandle must be |
104 | polled for reading by some mechanism outside this module (e.g. Event or |
483 | polled for reading by some mechanism outside this module (e.g. Event or |
… | |
… | |
143 | interactiveness when perl is not fast enough to process all requests in |
522 | interactiveness when perl is not fast enough to process all requests in |
144 | time. |
523 | time. |
145 | |
524 | |
146 | For interactive programs, values such as C<0.01> to C<0.1> should be fine. |
525 | For interactive programs, values such as C<0.01> to C<0.1> should be fine. |
147 | |
526 | |
148 | Example: Install an Event watcher that automatically calls |
527 | Example: Install an EV watcher that automatically calls |
149 | BDB::poll_cb with low priority, to ensure that other parts of the |
528 | BDB::poll_cb with low priority, to ensure that other parts of the |
150 | program get the CPU sometimes even under high AIO load. |
529 | program get the CPU sometimes even under high load. |
151 | |
530 | |
152 | # try not to spend much more than 0.1s in poll_cb |
531 | # try not to spend much more than 0.1s in poll_cb |
153 | BDB::max_poll_time 0.1; |
532 | BDB::max_poll_time 0.1; |
154 | |
533 | |
155 | # use a low priority so other tasks have priority |
534 | 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 | |
535 | |
160 | =item BDB::poll_wait |
536 | =item BDB::poll_wait |
161 | |
537 | |
162 | If there are any outstanding requests and none of them in the result |
538 | If there are any outstanding requests and none of them in the result |
163 | phase, wait till the result filehandle becomes ready for reading (simply |
539 | phase, wait till the result filehandle becomes ready for reading (simply |
… | |
… | |
175 | |
551 | |
176 | BDB::poll_wait, BDB::poll_cb |
552 | BDB::poll_wait, BDB::poll_cb |
177 | |
553 | |
178 | =item BDB::flush |
554 | =item BDB::flush |
179 | |
555 | |
180 | Wait till all outstanding AIO requests have been handled. |
556 | Wait till all outstanding BDB requests have been handled. |
181 | |
557 | |
182 | Strictly equivalent to: |
558 | Strictly equivalent to: |
183 | |
559 | |
184 | BDB::poll_wait, BDB::poll_cb |
560 | BDB::poll_wait, BDB::poll_cb |
185 | while BDB::nreqs; |
561 | while BDB::nreqs; |
186 | |
562 | |
|
|
563 | =back |
|
|
564 | |
|
|
565 | =head2 VERSION CHECKING |
|
|
566 | |
|
|
567 | BerkeleyDB comes in various versions, many of them have minor |
|
|
568 | incompatibilities. This means that traditional "at least version x.x" |
|
|
569 | checks are often not sufficient. |
|
|
570 | |
|
|
571 | Example: set the log_autoremove option in a way compatible with <v4.7 and |
|
|
572 | v4.7. Note the use of & on the constants to avoid triggering a compiletime |
|
|
573 | bug when the symbol isn't available. |
|
|
574 | |
|
|
575 | $DB_ENV->set_flags (&BDB::LOG_AUTOREMOVE ) if BDB::VERSION v0, v4.7; |
|
|
576 | $DB_ENV->log_set_config (&BDB::LOG_AUTO_REMOVE) if BDB::VERSION v4.7; |
|
|
577 | |
|
|
578 | =over 4 |
|
|
579 | |
|
|
580 | =item BDB::VERSION |
|
|
581 | |
|
|
582 | The C<BDB::VERSION> function, when called without arguments, returns the |
|
|
583 | Berkeley DB version as a v-string (usually with 3 components). You should |
|
|
584 | use C<lt> and C<ge> operators exclusively to make comparisons. |
|
|
585 | |
|
|
586 | Example: check for at least version 4.7. |
|
|
587 | |
|
|
588 | BDB::VERSION ge v4.7 or die; |
|
|
589 | |
|
|
590 | =item BDB::VERSION min-version |
|
|
591 | |
|
|
592 | Returns true if the BDB version is at least the given version (specified |
|
|
593 | as a v-string), false otherwise. |
|
|
594 | |
|
|
595 | Example: check for at least version 4.5. |
|
|
596 | |
|
|
597 | BDB::VERSION v4.7 or die; |
|
|
598 | |
|
|
599 | =item BDB::VERSION min-version, max-version |
|
|
600 | |
|
|
601 | Returns true of the BDB version is at least version C<min-version> (specify C<undef> or C<v0> for any minimum version) |
|
|
602 | and less then C<max-version>. |
|
|
603 | |
|
|
604 | Example: check wether version is strictly less then v4.7. |
|
|
605 | |
|
|
606 | BDB::VERSION v0, v4.7 |
|
|
607 | or die "version 4.7 is not yet supported"; |
|
|
608 | |
|
|
609 | =back |
|
|
610 | |
|
|
611 | =cut |
|
|
612 | |
|
|
613 | sub VERSION { |
|
|
614 | # I was dumb enough to override the VERSION method here, so let's try |
|
|
615 | # to fix it up. |
|
|
616 | |
|
|
617 | if ($_[0] eq __PACKAGE__) { |
|
|
618 | $VERSION |
|
|
619 | } else { |
|
|
620 | if (@_ > 0) { |
|
|
621 | return undef if VERSION_v lt $_[0]; |
|
|
622 | if (@_ > 1) { |
|
|
623 | return undef if VERSION_v ge $_[1]; |
|
|
624 | } |
|
|
625 | } |
|
|
626 | |
|
|
627 | VERSION_v |
|
|
628 | } |
|
|
629 | } |
|
|
630 | |
187 | =head3 CONTROLLING THE NUMBER OF THREADS |
631 | =head2 CONTROLLING THE NUMBER OF THREADS |
|
|
632 | |
|
|
633 | =over 4 |
188 | |
634 | |
189 | =item BDB::min_parallel $nthreads |
635 | =item BDB::min_parallel $nthreads |
190 | |
636 | |
191 | Set the minimum number of AIO threads to C<$nthreads>. The current |
637 | Set the minimum number of BDB threads to C<$nthreads>. The current |
192 | default is C<8>, which means eight asynchronous operations can execute |
638 | default is C<8>, which means eight asynchronous operations can execute |
193 | concurrently at any one time (the number of outstanding requests, |
639 | concurrently at any one time (the number of outstanding requests, |
194 | however, is unlimited). |
640 | however, is unlimited). |
195 | |
641 | |
196 | BDB starts threads only on demand, when an AIO request is queued and |
642 | BDB starts threads only on demand, when an BDB request is queued and |
197 | no free thread exists. Please note that queueing up a hundred requests can |
643 | no free thread exists. Please note that queueing up a hundred requests can |
198 | create demand for a hundred threads, even if it turns out that everything |
644 | create demand for a hundred threads, even if it turns out that everything |
199 | is in the cache and could have been processed faster by a single thread. |
645 | is in the cache and could have been processed faster by a single thread. |
200 | |
646 | |
201 | It is recommended to keep the number of threads relatively low, as some |
647 | It is recommended to keep the number of threads relatively low, as some |
… | |
… | |
206 | Under most circumstances you don't need to call this function, as the |
652 | Under most circumstances you don't need to call this function, as the |
207 | module selects a default that is suitable for low to moderate load. |
653 | module selects a default that is suitable for low to moderate load. |
208 | |
654 | |
209 | =item BDB::max_parallel $nthreads |
655 | =item BDB::max_parallel $nthreads |
210 | |
656 | |
211 | Sets the maximum number of AIO threads to C<$nthreads>. If more than the |
657 | Sets the maximum number of BDB threads to C<$nthreads>. If more than the |
212 | specified number of threads are currently running, this function kills |
658 | specified number of threads are currently running, this function kills |
213 | them. This function blocks until the limit is reached. |
659 | them. This function blocks until the limit is reached. |
214 | |
660 | |
215 | While C<$nthreads> are zero, aio requests get queued but not executed |
661 | While C<$nthreads> are zero, aio requests get queued but not executed |
216 | until the number of threads has been increased again. |
662 | until the number of threads has been increased again. |
… | |
… | |
251 | |
697 | |
252 | You can still queue as many requests as you want. Therefore, |
698 | You can still queue as many requests as you want. Therefore, |
253 | C<max_oustsanding> is mainly useful in simple scripts (with low values) or |
699 | C<max_oustsanding> is mainly useful in simple scripts (with low values) or |
254 | as a stop gap to shield against fatal memory overflow (with large values). |
700 | as a stop gap to shield against fatal memory overflow (with large values). |
255 | |
701 | |
256 | =item BDB::set_sync_prepare $cb |
702 | =item $old_cb = BDB::set_sync_prepare $cb |
257 | |
703 | |
258 | Sets a callback that is called whenever a request is created without an |
704 | Sets a callback that is called whenever a request is created without an |
259 | explicit callback. It has to return two code references. The first is used |
705 | explicit callback. It has to return two code references. The first is used |
260 | as the request callback, and the second is called to wait until the first |
706 | as the request callback (it should save the return status), and the second |
261 | callback has been called. The default implementation works like this: |
707 | is called to wait until the first callback has been called (it must set |
|
|
708 | C<$!> to the return status). |
|
|
709 | |
|
|
710 | This mechanism can be used to include BDB into other event mechanisms, |
|
|
711 | such as L<Coro::BDB>. |
|
|
712 | |
|
|
713 | To allow other, callback-based, events to be executed while callback-less |
|
|
714 | ones are run, you could use this sync prepare function: |
262 | |
715 | |
263 | sub { |
716 | sub { |
264 | my $status; |
717 | my $status; |
265 | ( |
718 | ( |
266 | sub { $status = $! }, |
719 | sub { $status = $! }, |
267 | sub { BDB::poll while !defined $status; $! = $status }, |
720 | sub { BDB::poll while !defined $status; $! = $status }, |
268 | ) |
721 | ) |
269 | } |
722 | } |
270 | |
723 | |
|
|
724 | It works by polling for results till the request has finished and then |
|
|
725 | sets C<$!> to the return value. This means that if you don't use a |
|
|
726 | callback, BDB would simply fall back to synchronous operations. |
|
|
727 | |
|
|
728 | By default, or if the sync prepare function is set to C<undef>, is to |
|
|
729 | execute callback-less BDB requests in the foreground thread, setting C<$!> |
|
|
730 | to the return value, without polling for other events. |
|
|
731 | |
271 | =back |
732 | =back |
272 | |
733 | |
273 | =head3 STATISTICAL INFORMATION |
734 | =head2 STATISTICAL INFORMATION |
274 | |
735 | |
275 | =over 4 |
736 | =over 4 |
276 | |
737 | |
277 | =item BDB::nreqs |
738 | =item BDB::nreqs |
278 | |
739 | |
… | |
… | |
296 | |
757 | |
297 | =back |
758 | =back |
298 | |
759 | |
299 | =cut |
760 | =cut |
300 | |
761 | |
301 | set_sync_prepare { |
762 | set_sync_prepare (undef); |
302 | my $status; |
|
|
303 | ( |
|
|
304 | sub { |
|
|
305 | $status = $!; |
|
|
306 | }, |
|
|
307 | sub { |
|
|
308 | BDB::poll while !defined $status; |
|
|
309 | $! = $status; |
|
|
310 | }, |
|
|
311 | ) |
|
|
312 | }; |
|
|
313 | |
763 | |
314 | min_parallel 8; |
764 | min_parallel 8; |
315 | |
765 | |
316 | END { flush } |
766 | END { flush } |
317 | |
767 | |
318 | 1; |
768 | 1; |
319 | |
769 | |
|
|
770 | =head1 COMMON PITFALLS |
|
|
771 | |
|
|
772 | =head2 Unexpected Crashes |
|
|
773 | |
|
|
774 | Remember that, by default, BDB will execute requests in parallel, in |
|
|
775 | somewhat random order. That means that it is easy to run a C<db_get> |
|
|
776 | request on the same database as a concurrent C<db_close> request, leading |
|
|
777 | to a crash, silent data corruption, eventually the next world war on |
|
|
778 | terrorism. |
|
|
779 | |
|
|
780 | If you only ever use foreground requests (without a callback), this will |
|
|
781 | not be an issue (unless you use threads). |
|
|
782 | |
|
|
783 | =head2 Unexpected Freezes or Deadlocks |
|
|
784 | |
|
|
785 | Remember that, by default, BDB will execute requests in parallel, which |
|
|
786 | easily leads to deadlocks (even concurrent put's on the same database can |
|
|
787 | deadlock). |
|
|
788 | |
|
|
789 | You either need to run deadlock detection (and handle the resulting |
|
|
790 | errors), or make sure only one process ever updates the database, ine one |
|
|
791 | thread, e.g. by using only foreground requests (without a callback). |
|
|
792 | |
320 | =head2 FORK BEHAVIOUR |
793 | =head1 FORK BEHAVIOUR |
321 | |
794 | |
322 | This module should do "the right thing" when the process using it forks: |
795 | This module should do "the right thing" when the process using it forks: |
323 | |
796 | |
324 | Before the fork, IO::AIO enters a quiescent state where no requests |
797 | Before the fork, BDB enters a quiescent state where no requests |
325 | can be added in other threads and no results will be processed. After |
798 | can be added in other threads and no results will be processed. After |
326 | the fork the parent simply leaves the quiescent state and continues |
799 | the fork the parent simply leaves the quiescent state and continues |
327 | request/result processing, while the child frees the request/result queue |
800 | request/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 |
801 | (so that the requests started before the fork will only be handled in the |
329 | parent). Threads will be started on demand until the limit set in the |
802 | parent). Threads will be started on demand until the limit set in the |
330 | parent process has been reached again. |
803 | parent process has been reached again. |
331 | |
804 | |
332 | In short: the parent will, after a short pause, continue as if fork had |
805 | In short: the parent will, after a short pause, continue as if fork had |
333 | not been called, while the child will act as if IO::AIO has not been used |
806 | not been called, while the child will act as if BDB has not been used |
334 | yet. |
807 | yet. |
335 | |
808 | |
|
|
809 | Win32 note: there is no fork on win32, and perls emulation of it is too |
|
|
810 | broken to be supported, so do not use BDB in a windows pseudo-fork, better |
|
|
811 | yet, switch to a more capable platform. |
|
|
812 | |
336 | =head2 MEMORY USAGE |
813 | =head1 MEMORY USAGE |
337 | |
814 | |
338 | Per-request usage: |
815 | Per-request usage: |
339 | |
816 | |
340 | Each aio request uses - depending on your architecture - around 100-200 |
817 | Each aio request uses - depending on your architecture - around 100-200 |
341 | bytes of memory. In addition, stat requests need a stat buffer (possibly |
818 | bytes of memory. In addition, stat requests need a stat buffer (possibly |
342 | a few hundred bytes), readdir requires a result buffer and so on. Perl |
819 | a few hundred bytes), readdir requires a result buffer and so on. Perl |
343 | scalars and other data passed into aio requests will also be locked and |
820 | scalars and other data passed into aio requests will also be locked and |
344 | will consume memory till the request has entered the done state. |
821 | will consume memory till the request has entered the done state. |
345 | |
822 | |
346 | This is now awfully much, so queuing lots of requests is not usually a |
823 | This is not awfully much, so queuing lots of requests is not usually a |
347 | problem. |
824 | problem. |
348 | |
825 | |
349 | Per-thread usage: |
826 | Per-thread usage: |
350 | |
827 | |
351 | In the execution phase, some aio requests require more memory for |
828 | In the execution phase, some aio requests require more memory for |
352 | temporary buffers, and each thread requires a stack and other data |
829 | temporary buffers, and each thread requires a stack and other data |
353 | structures (usually around 16k-128k, depending on the OS). |
830 | structures (usually around 16k-128k, depending on the OS). |
354 | |
831 | |
|
|
832 | =head1 WIN32 FILENAMES/DATABASE NAME MESS |
|
|
833 | |
|
|
834 | Perl on Win32 supports only ASCII filenames (the reason is that it abuses |
|
|
835 | an internal flag to store wether a filename is Unicode or ANSI, but that |
|
|
836 | flag is used for somethign else in the perl core, so there is no way to |
|
|
837 | detect wether a filename is ANSI or Unicode-encoded). The BDB module |
|
|
838 | tries to work around this issue by assuming that the filename is an ANSI |
|
|
839 | filename and BDB was built for unicode support. |
|
|
840 | |
355 | =head1 KNOWN BUGS |
841 | =head1 KNOWN BUGS |
356 | |
842 | |
357 | Known bugs will be fixed in the next release. |
843 | Known bugs will be fixed in the next release, except: |
|
|
844 | |
|
|
845 | If you use a transaction in any request, and the request returns |
|
|
846 | with an operating system error or DB_LOCK_NOTGRANTED, the internal |
|
|
847 | TXN_DEADLOCK flag will be set on the transaction. See C<db_txn_finish>, |
|
|
848 | above. |
358 | |
849 | |
359 | =head1 SEE ALSO |
850 | =head1 SEE ALSO |
360 | |
851 | |
361 | L<Coro::AIO>. |
852 | L<AnyEvent::BDB> (event loop integration), L<Coro::BDB> (more natural |
|
|
853 | syntax), L<IO::AIO> (nice to have). |
362 | |
854 | |
363 | =head1 AUTHOR |
855 | =head1 AUTHOR |
364 | |
856 | |
365 | Marc Lehmann <schmorp@schmorp.de> |
857 | Marc Lehmann <schmorp@schmorp.de> |
366 | http://home.schmorp.de/ |
858 | http://home.schmorp.de/ |