1 |
=head1 NAME |
2 |
|
3 |
BDB - Asynchronous Berkeley DB access |
4 |
|
5 |
=head1 SYNOPSIS |
6 |
|
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 |
|
45 |
|
46 |
=head1 DESCRIPTION |
47 |
|
48 |
See the BerkeleyDB documentation (L<http://www.oracle.com/technology/documentation/berkeley-db/db/index.html>). |
49 |
The BDB API is very similar to the C API (the translation has been very faithful). |
50 |
|
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 |
|
58 |
|
59 |
=head1 REQUEST ANATOMY AND LIFETIME |
60 |
|
61 |
Every request method creates a request. which is a C data structure not |
62 |
directly visible to Perl. |
63 |
|
64 |
During their existance, bdb requests travel through the following states, |
65 |
in order: |
66 |
|
67 |
=over 4 |
68 |
|
69 |
=item ready |
70 |
|
71 |
Immediately after a request is created it is put into the ready state, |
72 |
waiting for a thread to execute it. |
73 |
|
74 |
=item execute |
75 |
|
76 |
A thread has accepted the request for processing and is currently |
77 |
executing it (e.g. blocking in read). |
78 |
|
79 |
=item pending |
80 |
|
81 |
The request has been executed and is waiting for result processing. |
82 |
|
83 |
While request submission and execution is fully asynchronous, result |
84 |
processing is not and relies on the perl interpreter calling C<poll_cb> |
85 |
(or another function with the same effect). |
86 |
|
87 |
=item result |
88 |
|
89 |
The request results are processed synchronously by C<poll_cb>. |
90 |
|
91 |
The C<poll_cb> function will process all outstanding aio requests by |
92 |
calling their callbacks, freeing memory associated with them and managing |
93 |
any groups they are contained in. |
94 |
|
95 |
=item done |
96 |
|
97 |
Request has reached the end of its lifetime and holds no resources anymore |
98 |
(except possibly for the Perl object, but its connection to the actual |
99 |
aio request is severed and calling its methods will either do nothing or |
100 |
result in a runtime error). |
101 |
|
102 |
=back |
103 |
|
104 |
=cut |
105 |
|
106 |
package BDB; |
107 |
|
108 |
no warnings; |
109 |
use strict 'vars'; |
110 |
|
111 |
use base 'Exporter'; |
112 |
|
113 |
BEGIN { |
114 |
our $VERSION = '1.7'; |
115 |
|
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( |
128 |
poll_fileno poll_cb poll_wait flush |
129 |
min_parallel max_parallel max_idle |
130 |
nreqs nready npending nthreads |
131 |
max_poll_time max_poll_reqs |
132 |
); |
133 |
|
134 |
require XSLoader; |
135 |
XSLoader::load ("BDB", $VERSION); |
136 |
} |
137 |
|
138 |
=head2 WIN32 FILENAMES/DATABASE NAME MESS |
139 |
|
140 |
Perl on Win32 supports only ASCII filenames (the reason is that it abuses |
141 |
an internal flag to store wether a filename is Unicode or ANSI, but that |
142 |
flag is used for somethign else in the perl core, so there is no way to |
143 |
detect wether a filename is ANSI or Unicode-encoded). The BDB module |
144 |
tries to work around this issue by assuming that the filename is an ANSI |
145 |
filename and BDB was built for unicode support. |
146 |
|
147 |
=head2 BERKELEYDB FUNCTIONS |
148 |
|
149 |
All of these are functions. The create functions simply return a new |
150 |
object and never block. All the remaining functions take an optional |
151 |
callback as last argument. If it is missing, then the function will be |
152 |
executed synchronously. In both cases, C<$!> will reflect the return value |
153 |
of the function. |
154 |
|
155 |
BDB functions that cannot block (mostly functions that manipulate |
156 |
settings) are method calls on the relevant objects, so the rule of thumb |
157 |
is: if it's a method, it's not blocking, if it's a function, it takes a |
158 |
callback as last argument. |
159 |
|
160 |
In the following, C<$int> signifies an integer return value, |
161 |
C<bdb_filename> is a "filename" (octets on unix, madness on windows), |
162 |
C<U32> is an unsigned 32 bit integer, C<int> is some integer, C<NV> is a |
163 |
floating point value. |
164 |
|
165 |
Most C<SV *> types are generic perl scalars (for input and output of data |
166 |
values). |
167 |
|
168 |
The various C<DB_ENV> etc. arguments are handles return by |
169 |
C<db_env_create>, C<db_create>, C<txn_begin> and so on. If they have an |
170 |
appended C<_ornull> this means they are optional and you can pass C<undef> |
171 |
for them, resulting a NULL pointer on the C level. |
172 |
|
173 |
The C<SV *callback> is the optional callback function to call when the |
174 |
request is completed. This last callback argument is special: the callback |
175 |
is simply the last argument passed. If there are "optional" arguments |
176 |
before the callback they can be left out. The callback itself can be left |
177 |
out or specified as C<undef>, in which case the function will be executed |
178 |
synchronously. |
179 |
|
180 |
For example, C<db_env_txn_checkpoint> usually is called with all integer |
181 |
arguments zero. These can be left out, so all of these specify a call |
182 |
to C<< DB_ENV->txn_checkpoint >>, to be executed asynchronously with a |
183 |
callback to be called: |
184 |
|
185 |
db_env_txn_checkpoint $db_env, 0, 0, 0, sub { }; |
186 |
db_env_txn_checkpoint $db_env, 0, 0, sub { }; |
187 |
db_env_txn_checkpoint $db_env, sub { }; |
188 |
|
189 |
While these all specify a call to C<< DB_ENV->txn_checkpoint >> to be |
190 |
executed synchronously: |
191 |
|
192 |
db_env_txn_checkpoint $db_env, 0, 0, 0, undef; |
193 |
db_env_txn_checkpoint $db_env, 0, 0, 0; |
194 |
db_env_txn_checkpoint $db_env, 0; |
195 |
|
196 |
=head3 BDB functions |
197 |
|
198 |
Functions in the BDB namespace, exported by default: |
199 |
|
200 |
$env = db_env_create (U32 env_flags = 0) |
201 |
flags: RPCCLIENT |
202 |
|
203 |
db_env_open (DB_ENV *env, bdb_filename db_home, U32 open_flags, int mode, SV *callback = &PL_sv_undef) |
204 |
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 |
205 |
db_env_close (DB_ENV *env, U32 flags = 0, SV *callback = &PL_sv_undef) |
206 |
db_env_txn_checkpoint (DB_ENV *env, U32 kbyte = 0, U32 min = 0, U32 flags = 0, SV *callback = &PL_sv_undef) |
207 |
flags: FORCE |
208 |
db_env_lock_detect (DB_ENV *env, U32 flags = 0, U32 atype = DB_LOCK_DEFAULT, SV *dummy = 0, SV *callback = &PL_sv_undef) |
209 |
atype: LOCK_DEFAULT LOCK_EXPIRE LOCK_MAXLOCKS LOCK_MAXWRITE LOCK_MINLOCKS LOCK_MINWRITE LOCK_OLDEST LOCK_RANDOM LOCK_YOUNGEST |
210 |
db_env_memp_sync (DB_ENV *env, SV *dummy = 0, SV *callback = &PL_sv_undef) |
211 |
db_env_memp_trickle (DB_ENV *env, int percent, SV *dummy = 0, SV *callback = &PL_sv_undef) |
212 |
db_env_dbremove (DB_ENV *env, DB_TXN_ornull *txnid, bdb_filename file, bdb_filename database, U32 flags = 0, SV *callback = &PL_sv_undef) |
213 |
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) |
214 |
|
215 |
$db = db_create (DB_ENV *env = 0, U32 flags = 0) |
216 |
flags: XA_CREATE |
217 |
|
218 |
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) |
219 |
flags: AUTO_COMMIT CREATE EXCL MULTIVERSION NOMMAP RDONLY READ_UNCOMMITTED THREAD TRUNCATE |
220 |
db_close (DB *db, U32 flags = 0, SV *callback = &PL_sv_undef) |
221 |
flags: DB_NOSYNC |
222 |
db_upgrade (DB *db, bdb_filename file, U32 flags = 0, SV *callback = &PL_sv_undef) |
223 |
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) |
224 |
flags: FREELIST_ONLY FREE_SPACE |
225 |
db_sync (DB *db, U32 flags = 0, SV *callback = &PL_sv_undef) |
226 |
db_key_range (DB *db, DB_TXN_ornull *txn, SV *key, SV *key_range, U32 flags = 0, SV *callback = &PL_sv_undef) |
227 |
db_put (DB *db, DB_TXN_ornull *txn, SV *key, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef) |
228 |
flags: APPEND NODUPDATA NOOVERWRITE |
229 |
db_get (DB *db, DB_TXN_ornull *txn, SV *key, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef) |
230 |
flags: CONSUME CONSUME_WAIT GET_BOTH SET_RECNO MULTIPLE READ_COMMITTED READ_UNCOMMITTED RMW |
231 |
db_pget (DB *db, DB_TXN_ornull *txn, SV *key, SV *pkey, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef) |
232 |
flags: CONSUME CONSUME_WAIT GET_BOTH SET_RECNO MULTIPLE READ_COMMITTED READ_UNCOMMITTED RMW |
233 |
db_del (DB *db, DB_TXN_ornull *txn, SV *key, U32 flags = 0, SV *callback = &PL_sv_undef) |
234 |
db_txn_commit (DB_TXN *txn, U32 flags = 0, SV *callback = &PL_sv_undef) |
235 |
flags: TXN_NOSYNC TXN_SYNC |
236 |
db_txn_abort (DB_TXN *txn, SV *callback = &PL_sv_undef) |
237 |
|
238 |
db_c_close (DBC *dbc, SV *callback = &PL_sv_undef) |
239 |
db_c_count (DBC *dbc, SV *count, U32 flags = 0, SV *callback = &PL_sv_undef) |
240 |
db_c_put (DBC *dbc, SV *key, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef) |
241 |
flags: AFTER BEFORE CURRENT KEYFIRST KEYLAST NODUPDATA |
242 |
db_c_get (DBC *dbc, SV *key, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef) |
243 |
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 |
244 |
db_c_pget (DBC *dbc, SV *key, SV *pkey, SV *data, U32 flags = 0, SV *callback = &PL_sv_undef) |
245 |
db_c_del (DBC *dbc, U32 flags = 0, SV *callback = &PL_sv_undef) |
246 |
|
247 |
db_sequence_open (DB_SEQUENCE *seq, DB_TXN_ornull *txnid, SV *key, U32 flags = 0, SV *callback = &PL_sv_undef) |
248 |
flags: CREATE EXCL |
249 |
db_sequence_close (DB_SEQUENCE *seq, U32 flags = 0, SV *callback = &PL_sv_undef) |
250 |
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) |
251 |
flags: TXN_NOSYNC |
252 |
db_sequence_remove (DB_SEQUENCE *seq, DB_TXN_ornull *txnid = 0, U32 flags = 0, SV *callback = &PL_sv_undef) |
253 |
flags: TXN_NOSYNC |
254 |
|
255 |
=head4 db_txn_finish (DB_TXN *txn, U32 flags = 0, SV *callback = &PL_sv_undef) |
256 |
|
257 |
This is not actually a Berkeley DB function but a BDB module |
258 |
extension. The background for this exytension is: It is very annoying to |
259 |
have to check every single BDB function for error returns and provide a |
260 |
codepath out of your transaction. While the BDB module still makes this |
261 |
possible, it contains the following extensions: |
262 |
|
263 |
When a transaction-protected function returns any operating system |
264 |
error (errno > 0), BDB will set the C<TXN_DEADLOCK> flag on the |
265 |
transaction. This flag is also set by Berkeley DB functions themselves |
266 |
when an operation fails with LOCK_DEADLOCK, and it causes all further |
267 |
operations on that transaction (including C<db_txn_commit>) to fail. |
268 |
|
269 |
The C<db_txn_finish> request will look at this flag, and, if it is set, |
270 |
will automatically call C<db_txn_abort> (setting errno to C<LOCK_DEADLOCK> |
271 |
if it isn't set to something else yet). If it isn't set, it will call |
272 |
C<db_txn_commit> and return the error normally. |
273 |
|
274 |
How to use this? Easy: just write your transaction normally: |
275 |
|
276 |
my $txn = $db_env->txn_begin; |
277 |
db_get $db, $txn, "key", my $data; |
278 |
db_put $db, $txn, "key", $data + 1 unless $! == BDB::NOTFOUND; |
279 |
db_txn_finish $txn; |
280 |
die "transaction failed" if $!; |
281 |
|
282 |
That is, handle only the expected errors. If something unexpected happens |
283 |
(EIO, LOCK_NOTGRANTED or a deadlock in either db_get or db_put), then the remaining |
284 |
requests (db_put in this case) will simply be skipped (they will fail with |
285 |
LOCK_DEADLOCK) and the transaction will be aborted. |
286 |
|
287 |
You can use the C<< $txn->failed >> method to check wether a transaction |
288 |
has failed in this way and abort further processing (excluding |
289 |
C<db_txn_finish>). |
290 |
|
291 |
=head3 DB_ENV/database environment methods |
292 |
|
293 |
Methods available on DB_ENV/$env handles: |
294 |
|
295 |
DESTROY (DB_ENV_ornull *env) |
296 |
CODE: |
297 |
if (env) |
298 |
env->close (env, 0); |
299 |
|
300 |
$int = $env->set_data_dir (const char *dir) |
301 |
$int = $env->set_tmp_dir (const char *dir) |
302 |
$int = $env->set_lg_dir (const char *dir) |
303 |
$int = $env->set_shm_key (long shm_key) |
304 |
$int = $env->set_cachesize (U32 gbytes, U32 bytes, int ncache = 0) |
305 |
$int = $env->set_flags (U32 flags, int onoff = 1) |
306 |
$int = $env->log_set_config (U32 flags, int onoff = 1) [v4.7] |
307 |
$int = $env->set_intermediate_dir_mode (const char *modestring) [v4.7] |
308 |
$env->set_errfile (FILE *errfile = 0) |
309 |
$env->set_msgfile (FILE *msgfile = 0) |
310 |
$int = $env->set_verbose (U32 which, int onoff = 1) |
311 |
$int = $env->set_encrypt (const char *password, U32 flags = 0) |
312 |
$int = $env->set_timeout (NV timeout_seconds, U32 flags = SET_TXN_TIMEOUT) |
313 |
$int = $env->set_mp_max_openfd (int maxopenfd); |
314 |
$int = $env->set_mp_max_write (int maxwrite, int maxwrite_sleep); |
315 |
$int = $env->set_mp_mmapsize (int mmapsize_mb) |
316 |
$int = $env->set_lk_detect (U32 detect = DB_LOCK_DEFAULT) |
317 |
$int = $env->set_lk_max_lockers (U32 max) |
318 |
$int = $env->set_lk_max_locks (U32 max) |
319 |
$int = $env->set_lk_max_objects (U32 max) |
320 |
$int = $env->set_lg_bsize (U32 max) |
321 |
$int = $env->set_lg_max (U32 max) |
322 |
$int = $env->mutex_set_increment (U32 increment) |
323 |
$int = $env->mutex_set_tas_spins (U32 tas_spins) |
324 |
$int = $env->mutex_set_max (U32 max) |
325 |
$int = $env->mutex_set_align (U32 align) |
326 |
|
327 |
$txn = $env->txn_begin (DB_TXN_ornull *parent = 0, U32 flags = 0) |
328 |
flags: READ_COMMITTED READ_UNCOMMITTED TXN_NOSYNC TXN_NOWAIT TXN_SNAPSHOT TXN_SYNC TXN_WAIT TXN_WRITE_NOSYNC |
329 |
|
330 |
=head4 Example: |
331 |
|
332 |
use AnyEvent; |
333 |
use BDB; |
334 |
|
335 |
our $FH; open $FH, "<&=" . BDB::poll_fileno; |
336 |
our $WATCHER = AnyEvent->io (fh => $FH, poll => 'r', cb => \&BDB::poll_cb); |
337 |
|
338 |
BDB::min_parallel 8; |
339 |
|
340 |
my $env = db_env_create; |
341 |
|
342 |
mkdir "bdtest", 0700; |
343 |
db_env_open |
344 |
$env, |
345 |
"bdtest", |
346 |
BDB::INIT_LOCK | BDB::INIT_LOG | BDB::INIT_MPOOL | BDB::INIT_TXN | BDB::RECOVER | BDB::USE_ENVIRON | BDB::CREATE, |
347 |
0600; |
348 |
|
349 |
$env->set_flags (BDB::AUTO_COMMIT | BDB::TXN_NOSYNC, 1); |
350 |
|
351 |
|
352 |
=head3 DB/database methods |
353 |
|
354 |
Methods available on DB/$db handles: |
355 |
|
356 |
DESTROY (DB_ornull *db) |
357 |
CODE: |
358 |
if (db) |
359 |
{ |
360 |
SV *env = (SV *)db->app_private; |
361 |
db->close (db, 0); |
362 |
SvREFCNT_dec (env); |
363 |
} |
364 |
|
365 |
$int = $db->set_cachesize (U32 gbytes, U32 bytes, int ncache = 0) |
366 |
$int = $db->set_flags (U32 flags) |
367 |
flags: CHKSUM ENCRYPT TXN_NOT_DURABLE |
368 |
Btree: DUP DUPSORT RECNUM REVSPLITOFF |
369 |
Hash: DUP DUPSORT |
370 |
Queue: INORDER |
371 |
Recno: RENUMBER SNAPSHOT |
372 |
|
373 |
$int = $db->set_encrypt (const char *password, U32 flags) |
374 |
$int = $db->set_lorder (int lorder) |
375 |
$int = $db->set_bt_minkey (U32 minkey) |
376 |
$int = $db->set_re_delim (int delim) |
377 |
$int = $db->set_re_pad (int re_pad) |
378 |
$int = $db->set_re_source (char *source) |
379 |
$int = $db->set_re_len (U32 re_len) |
380 |
$int = $db->set_h_ffactor (U32 h_ffactor) |
381 |
$int = $db->set_h_nelem (U32 h_nelem) |
382 |
$int = $db->set_q_extentsize (U32 extentsize) |
383 |
|
384 |
$dbc = $db->cursor (DB_TXN_ornull *txn = 0, U32 flags = 0) |
385 |
flags: READ_COMMITTED READ_UNCOMMITTED WRITECURSOR TXN_SNAPSHOT |
386 |
$seq = $db->sequence (U32 flags = 0) |
387 |
|
388 |
=head4 Example: |
389 |
|
390 |
my $db = db_create $env; |
391 |
db_open $db, undef, "table", undef, BDB::BTREE, BDB::AUTO_COMMIT | BDB::CREATE | BDB::READ_UNCOMMITTED, 0600; |
392 |
|
393 |
for (1..1000) { |
394 |
db_put $db, undef, "key $_", "data $_"; |
395 |
|
396 |
db_key_range $db, undef, "key $_", my $keyrange; |
397 |
my ($lt, $eq, $gt) = @$keyrange; |
398 |
} |
399 |
|
400 |
db_del $db, undef, "key $_" for 1..1000; |
401 |
|
402 |
db_sync $db; |
403 |
|
404 |
|
405 |
=head3 DB_TXN/transaction methods |
406 |
|
407 |
Methods available on DB_TXN/$txn handles: |
408 |
|
409 |
DESTROY (DB_TXN_ornull *txn) |
410 |
CODE: |
411 |
if (txn) |
412 |
txn->abort (txn); |
413 |
|
414 |
$int = $txn->set_timeout (NV timeout_seconds, U32 flags = SET_TXN_TIMEOUT) |
415 |
flags: SET_LOCK_TIMEOUT SET_TXN_TIMEOUT |
416 |
|
417 |
$bool = $txn->failed |
418 |
# see db_txn_finish documentation, above |
419 |
|
420 |
|
421 |
=head3 DBC/cursor methods |
422 |
|
423 |
Methods available on DBC/$dbc handles: |
424 |
|
425 |
DESTROY (DBC_ornull *dbc) |
426 |
CODE: |
427 |
if (dbc) |
428 |
dbc->c_close (dbc); |
429 |
|
430 |
$int = $cursor->set_priority ($priority = PRIORITY_*) |
431 |
|
432 |
=head4 Example: |
433 |
|
434 |
my $c = $db->cursor; |
435 |
|
436 |
for (;;) { |
437 |
db_c_get $c, my $key, my $data, BDB::NEXT; |
438 |
warn "<$!,$key,$data>"; |
439 |
last if $!; |
440 |
} |
441 |
|
442 |
db_c_close $c; |
443 |
|
444 |
|
445 |
=head3 DB_SEQUENCE/sequence methods |
446 |
|
447 |
Methods available on DB_SEQUENCE/$seq handles: |
448 |
|
449 |
DESTROY (DB_SEQUENCE_ornull *seq) |
450 |
CODE: |
451 |
if (seq) |
452 |
seq->close (seq, 0); |
453 |
|
454 |
$int = $seq->initial_value (db_seq_t value) |
455 |
$int = $seq->set_cachesize (U32 size) |
456 |
$int = $seq->set_flags (U32 flags) |
457 |
flags: SEQ_DEC SEQ_INC SEQ_WRAP |
458 |
$int = $seq->set_range (db_seq_t min, db_seq_t max) |
459 |
|
460 |
=head4 Example: |
461 |
|
462 |
my $seq = $db->sequence; |
463 |
|
464 |
db_sequence_open $seq, undef, "seq", BDB::CREATE; |
465 |
db_sequence_get $seq, undef, 1, my $value; |
466 |
|
467 |
|
468 |
=head2 SUPPORT FUNCTIONS |
469 |
|
470 |
=head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION |
471 |
|
472 |
=over 4 |
473 |
|
474 |
=item $msg = BDB::strerror [$errno] |
475 |
|
476 |
Returns the string corresponding to the given errno value. If no argument |
477 |
is given, use C<$!>. |
478 |
|
479 |
Note that the BDB module also patches the C<$!> variable directly, so you |
480 |
should be able to get a bdb error string by simply stringifying C<$!>. |
481 |
|
482 |
=item $fileno = BDB::poll_fileno |
483 |
|
484 |
Return the I<request result pipe file descriptor>. This filehandle must be |
485 |
polled for reading by some mechanism outside this module (e.g. Event or |
486 |
select, see below or the SYNOPSIS). If the pipe becomes readable you have |
487 |
to call C<poll_cb> to check the results. |
488 |
|
489 |
See C<poll_cb> for an example. |
490 |
|
491 |
=item BDB::poll_cb |
492 |
|
493 |
Process some outstanding events on the result pipe. You have to call this |
494 |
regularly. Returns the number of events processed. Returns immediately |
495 |
when no events are outstanding. The amount of events processed depends on |
496 |
the settings of C<BDB::max_poll_req> and C<BDB::max_poll_time>. |
497 |
|
498 |
If not all requests were processed for whatever reason, the filehandle |
499 |
will still be ready when C<poll_cb> returns. |
500 |
|
501 |
Example: Install an Event watcher that automatically calls |
502 |
BDB::poll_cb with high priority: |
503 |
|
504 |
Event->io (fd => BDB::poll_fileno, |
505 |
poll => 'r', async => 1, |
506 |
cb => \&BDB::poll_cb); |
507 |
|
508 |
=item BDB::max_poll_reqs $nreqs |
509 |
|
510 |
=item BDB::max_poll_time $seconds |
511 |
|
512 |
These set the maximum number of requests (default C<0>, meaning infinity) |
513 |
that are being processed by C<BDB::poll_cb> in one call, respectively |
514 |
the maximum amount of time (default C<0>, meaning infinity) spent in |
515 |
C<BDB::poll_cb> to process requests (more correctly the mininum amount |
516 |
of time C<poll_cb> is allowed to use). |
517 |
|
518 |
Setting C<max_poll_time> to a non-zero value creates an overhead of one |
519 |
syscall per request processed, which is not normally a problem unless your |
520 |
callbacks are really really fast or your OS is really really slow (I am |
521 |
not mentioning Solaris here). Using C<max_poll_reqs> incurs no overhead. |
522 |
|
523 |
Setting these is useful if you want to ensure some level of |
524 |
interactiveness when perl is not fast enough to process all requests in |
525 |
time. |
526 |
|
527 |
For interactive programs, values such as C<0.01> to C<0.1> should be fine. |
528 |
|
529 |
Example: Install an EV watcher that automatically calls |
530 |
BDB::poll_cb with low priority, to ensure that other parts of the |
531 |
program get the CPU sometimes even under high load. |
532 |
|
533 |
# try not to spend much more than 0.1s in poll_cb |
534 |
BDB::max_poll_time 0.1; |
535 |
|
536 |
my $bdb_poll = EV::io BDB::poll_fileno, EV::READ, \&BDB::poll_cb); |
537 |
|
538 |
=item BDB::poll_wait |
539 |
|
540 |
If there are any outstanding requests and none of them in the result |
541 |
phase, wait till the result filehandle becomes ready for reading (simply |
542 |
does a C<select> on the filehandle. This is useful if you want to |
543 |
synchronously wait for some requests to finish). |
544 |
|
545 |
See C<nreqs> for an example. |
546 |
|
547 |
=item BDB::poll |
548 |
|
549 |
Waits until some requests have been handled. |
550 |
|
551 |
Returns the number of requests processed, but is otherwise strictly |
552 |
equivalent to: |
553 |
|
554 |
BDB::poll_wait, BDB::poll_cb |
555 |
|
556 |
=item BDB::flush |
557 |
|
558 |
Wait till all outstanding BDB requests have been handled. |
559 |
|
560 |
Strictly equivalent to: |
561 |
|
562 |
BDB::poll_wait, BDB::poll_cb |
563 |
while BDB::nreqs; |
564 |
|
565 |
=back |
566 |
|
567 |
=head3 VERSION CHECKING |
568 |
|
569 |
BerkeleyDB comes in various versions, many of them have minor |
570 |
incompatibilities. This means that traditional "at least version x.x" |
571 |
checks are often not sufficient. |
572 |
|
573 |
Example: set the log_autoremove option in a way compatible with <v.47 and |
574 |
v4.7. Note the use of & on the constants to avoid triggering a compiletime |
575 |
bug when the symbol isn't available. |
576 |
|
577 |
$DB_ENV->set_flags (&BDB::LOG_AUTOREMOVE ) if BDB::VERSION v0, v4.7; |
578 |
$DB_ENV->log_set_config (&BDB::LOG_AUTO_REMOVE) if BDB::VERSION v4.7; |
579 |
|
580 |
=over 4 |
581 |
|
582 |
=item BDB::VERSION |
583 |
|
584 |
The C<BDB::VERSION> function, when called without arguments, returns the |
585 |
Berkeley DB version as a v-string (usually with 3 components). You should |
586 |
use C<lt> and C<ge> operators exclusively to make comparisons. |
587 |
|
588 |
Example: check for at least version 4.7. |
589 |
|
590 |
BDB::VERSION ge v4.7 or die; |
591 |
|
592 |
=item BDB::VERSION min-version |
593 |
|
594 |
Returns true if the BDB version is at least the given version (specified |
595 |
as a v-string), false otherwise. |
596 |
|
597 |
Example: check for at least version 4.5. |
598 |
|
599 |
BDB::VERSION v4.7 or die; |
600 |
|
601 |
=item BDB::VERSION min-version, max-version |
602 |
|
603 |
Returns true of the BDB version is at least version C<min-version> (specify C<undef> or C<v0> for any minimum version) |
604 |
and less then C<max-version>. |
605 |
|
606 |
Example: check wether version is strictly less then v4.7. |
607 |
|
608 |
BDB::VERSION v0, v4.7 |
609 |
or die "version 4.7 is not yet supported"; |
610 |
|
611 |
=back |
612 |
|
613 |
=cut |
614 |
|
615 |
sub VERSION { |
616 |
if (@_ > 0) { |
617 |
return undef if VERSION_v lt $_[0]; |
618 |
if (@_ > 1) { |
619 |
return undef if VERSION_v ge $_[1]; |
620 |
} |
621 |
} |
622 |
|
623 |
VERSION_v |
624 |
} |
625 |
|
626 |
=head3 CONTROLLING THE NUMBER OF THREADS |
627 |
|
628 |
=over 4 |
629 |
|
630 |
=item BDB::min_parallel $nthreads |
631 |
|
632 |
Set the minimum number of BDB threads to C<$nthreads>. The current |
633 |
default is C<8>, which means eight asynchronous operations can execute |
634 |
concurrently at any one time (the number of outstanding requests, |
635 |
however, is unlimited). |
636 |
|
637 |
BDB starts threads only on demand, when an BDB request is queued and |
638 |
no free thread exists. Please note that queueing up a hundred requests can |
639 |
create demand for a hundred threads, even if it turns out that everything |
640 |
is in the cache and could have been processed faster by a single thread. |
641 |
|
642 |
It is recommended to keep the number of threads relatively low, as some |
643 |
Linux kernel versions will scale negatively with the number of threads |
644 |
(higher parallelity => MUCH higher latency). With current Linux 2.6 |
645 |
versions, 4-32 threads should be fine. |
646 |
|
647 |
Under most circumstances you don't need to call this function, as the |
648 |
module selects a default that is suitable for low to moderate load. |
649 |
|
650 |
=item BDB::max_parallel $nthreads |
651 |
|
652 |
Sets the maximum number of BDB threads to C<$nthreads>. If more than the |
653 |
specified number of threads are currently running, this function kills |
654 |
them. This function blocks until the limit is reached. |
655 |
|
656 |
While C<$nthreads> are zero, aio requests get queued but not executed |
657 |
until the number of threads has been increased again. |
658 |
|
659 |
This module automatically runs C<max_parallel 0> at program end, to ensure |
660 |
that all threads are killed and that there are no outstanding requests. |
661 |
|
662 |
Under normal circumstances you don't need to call this function. |
663 |
|
664 |
=item BDB::max_idle $nthreads |
665 |
|
666 |
Limit the number of threads (default: 4) that are allowed to idle (i.e., |
667 |
threads that did not get a request to process within 10 seconds). That |
668 |
means if a thread becomes idle while C<$nthreads> other threads are also |
669 |
idle, it will free its resources and exit. |
670 |
|
671 |
This is useful when you allow a large number of threads (e.g. 100 or 1000) |
672 |
to allow for extremely high load situations, but want to free resources |
673 |
under normal circumstances (1000 threads can easily consume 30MB of RAM). |
674 |
|
675 |
The default is probably ok in most situations, especially if thread |
676 |
creation is fast. If thread creation is very slow on your system you might |
677 |
want to use larger values. |
678 |
|
679 |
=item $oldmaxreqs = BDB::max_outstanding $maxreqs |
680 |
|
681 |
This is a very bad function to use in interactive programs because it |
682 |
blocks, and a bad way to reduce concurrency because it is inexact: Better |
683 |
use an C<aio_group> together with a feed callback. |
684 |
|
685 |
Sets the maximum number of outstanding requests to C<$nreqs>. If you |
686 |
to queue up more than this number of requests, the next call to the |
687 |
C<poll_cb> (and C<poll_some> and other functions calling C<poll_cb>) |
688 |
function will block until the limit is no longer exceeded. |
689 |
|
690 |
The default value is very large, so there is no practical limit on the |
691 |
number of outstanding requests. |
692 |
|
693 |
You can still queue as many requests as you want. Therefore, |
694 |
C<max_oustsanding> is mainly useful in simple scripts (with low values) or |
695 |
as a stop gap to shield against fatal memory overflow (with large values). |
696 |
|
697 |
=item BDB::set_sync_prepare $cb |
698 |
|
699 |
Sets a callback that is called whenever a request is created without an |
700 |
explicit callback. It has to return two code references. The first is used |
701 |
as the request callback (it should save the return status), and the second |
702 |
is called to wait until the first callback has been called (it must set |
703 |
C<$!> to the return status). |
704 |
|
705 |
This mechanism can be used to include BDB into other event mechanisms, |
706 |
such as L<AnyEvent::BDB> or L<Coro::BDB>. |
707 |
|
708 |
The default implementation works like this: |
709 |
|
710 |
sub { |
711 |
my $status; |
712 |
( |
713 |
sub { $status = $! }, |
714 |
sub { BDB::poll while !defined $status; $! = $status }, |
715 |
) |
716 |
} |
717 |
|
718 |
It simply blocks the process till the request has finished and then sets |
719 |
C<$!> to the return value. This means that if you don't use a callback, |
720 |
BDB will simply fall back to synchronous operations. |
721 |
|
722 |
=back |
723 |
|
724 |
=head3 STATISTICAL INFORMATION |
725 |
|
726 |
=over 4 |
727 |
|
728 |
=item BDB::nreqs |
729 |
|
730 |
Returns the number of requests currently in the ready, execute or pending |
731 |
states (i.e. for which their callback has not been invoked yet). |
732 |
|
733 |
Example: wait till there are no outstanding requests anymore: |
734 |
|
735 |
BDB::poll_wait, BDB::poll_cb |
736 |
while BDB::nreqs; |
737 |
|
738 |
=item BDB::nready |
739 |
|
740 |
Returns the number of requests currently in the ready state (not yet |
741 |
executed). |
742 |
|
743 |
=item BDB::npending |
744 |
|
745 |
Returns the number of requests currently in the pending state (executed, |
746 |
but not yet processed by poll_cb). |
747 |
|
748 |
=back |
749 |
|
750 |
=cut |
751 |
|
752 |
set_sync_prepare { |
753 |
my $status; |
754 |
( |
755 |
sub { |
756 |
$status = $!; |
757 |
}, |
758 |
sub { |
759 |
BDB::poll while !defined $status; |
760 |
$! = $status; |
761 |
}, |
762 |
) |
763 |
}; |
764 |
|
765 |
min_parallel 8; |
766 |
|
767 |
END { flush } |
768 |
|
769 |
1; |
770 |
|
771 |
=head2 FORK BEHAVIOUR |
772 |
|
773 |
This module should do "the right thing" when the process using it forks: |
774 |
|
775 |
Before the fork, BDB enters a quiescent state where no requests |
776 |
can be added in other threads and no results will be processed. After |
777 |
the fork the parent simply leaves the quiescent state and continues |
778 |
request/result processing, while the child frees the request/result queue |
779 |
(so that the requests started before the fork will only be handled in the |
780 |
parent). Threads will be started on demand until the limit set in the |
781 |
parent process has been reached again. |
782 |
|
783 |
In short: the parent will, after a short pause, continue as if fork had |
784 |
not been called, while the child will act as if BDB has not been used |
785 |
yet. |
786 |
|
787 |
Win32 note: there is no fork on win32, and perls emulation of it is too |
788 |
broken to be supported, so do not use BDB in a windows pseudo-fork, better |
789 |
yet, switch to a more capable platform. |
790 |
|
791 |
=head2 MEMORY USAGE |
792 |
|
793 |
Per-request usage: |
794 |
|
795 |
Each aio request uses - depending on your architecture - around 100-200 |
796 |
bytes of memory. In addition, stat requests need a stat buffer (possibly |
797 |
a few hundred bytes), readdir requires a result buffer and so on. Perl |
798 |
scalars and other data passed into aio requests will also be locked and |
799 |
will consume memory till the request has entered the done state. |
800 |
|
801 |
This is not awfully much, so queuing lots of requests is not usually a |
802 |
problem. |
803 |
|
804 |
Per-thread usage: |
805 |
|
806 |
In the execution phase, some aio requests require more memory for |
807 |
temporary buffers, and each thread requires a stack and other data |
808 |
structures (usually around 16k-128k, depending on the OS). |
809 |
|
810 |
=head1 KNOWN BUGS |
811 |
|
812 |
Known bugs will be fixed in the next release, except: |
813 |
|
814 |
If you use a transaction in any request, and the request returns |
815 |
with an operating system error or DB_LOCK_NOTGRANTED, the internal |
816 |
TXN_DEADLOCK flag will be set on the transaction. See C<db_txn_finish>, |
817 |
above. |
818 |
|
819 |
=head1 SEE ALSO |
820 |
|
821 |
L<AnyEvent::BDB> (event loop integration), L<Coro::BDB> (more natural |
822 |
syntax), L<IO::AIO> (nice to have). |
823 |
|
824 |
=head1 AUTHOR |
825 |
|
826 |
Marc Lehmann <schmorp@schmorp.de> |
827 |
http://home.schmorp.de/ |
828 |
|
829 |
=cut |
830 |
|