Debian Codegraph @

CHANGES

raw
1 HikariCP Changes
2
3 Changes in 2.7.7
4
5 * fixed issue whereby configuration through the HikariConfigMXBean could not be altered due
6 to the sealed configuration change introduced in 2.7.5.
7
8 Changes in 2.7.6
9
10 * issue 1064 fixed regression where HikariConfig.copyStateTo() propagated the "sealed" status
11 of the source configuration to the target configuration -- preventing further changes.
12
13 Changes in 2.7.5
14
15 * issue 1061/pull 1062 fixed compatibility issue with requery.io caused by HikariCP's
16 proxied Statement class returning a driver-native ResultSet instance from getGeneratedKeys()
17 instead of returning a HikariCP proxied ResultSet.
18
19 * pull 1058 enable quantiles on Prometheus metrics.
20
21 * pull 1055 fixed incorrect JavaDoc for HikariConfigMXBean.getMinimumIdle() method.
22
23 * issue 1045/pull 1047 added Automatic-Module-Name to jar manifest to ensure that the Java 8
24 library plays well with the Java 9 module system.
25
26 * introduced the concept of a "sealed" configuration. Once a pool is started, attempts to
27 alter its configuration outside of the HikariConfigMXBean will result in an IllegalStateException.
28
29 Changes in 2.7.4
30
31 * pull 1026 added support for SQL Server's specific isolation level (SNAPSHOT).
32
33 * issue 926/pull 1022 HikariJNDIFactory should not throw a NamingException or else
34 cascading to other object factories cannot occur.
35
36 Changes in 2.7.3
37
38 * issue 1003 added PostgreSQL SQL State 0A000 to list of unrecoverable states calling
39 for eviction.
40
41 * pull 1002 updated micrometer support due to API changes in their release candidate.
42
43 Changes in 2.7.2
44
45 * issue 983 fix logic that determines how many idle connections can be removed, without
46 violating the minimumIdle contract.
47
48 * pull 987 add thread name to leak detection messages.
49
50 * issue 982 fix classloader order, try the ThreadContext classloader before other
51 classloaders.
52
53 * pull 977 log better messages when connection is evicted.
54
55 * fallback to four digit random pool suffix when SecurityManager prevents writing
56 to system properties for the purpose of JVM-wide unique pool identifiers.
57
58 Changes in 2.7.1
59
60 * issue 968 Wrong label order in MicrometerMetricsTracker for the connection usage
61 metric.
62
63 * issue 967 incorrect bitwise operator value in ConcurrentBag.requite method
64 intended to cause parkNanos() to be called every 256 iterations. Thanks to @ztkmkoo
65 for finding this.
66
67 Changes in 2.4.13
68
69 * backport more efficient contention handling in ConcurrentBag.requite from 2.6.x
70 branch.
71
72 * issue 955 fix possible race condition when Statements are closed on different
73 threads from which they were created.
74
75 Changes in 2.7.0
76
77 * added support for micrometer metrics (currently Alpha-level support).
78
79 * issue 905 mask JDBC password in URLs
80
81 * issue 940 fix Prometheus metric collector for multiple data config
82
83 * issue 941 add support for setting a default schema
84
85 * issue 955 fix possible race condition when Statements are closed on different
86 threads from which they were created.
87
88 Changes in 2.6.3
89
90 * issue 878 load driver class from ThreadContext classloader if it is not found
91 via the regular classloader.
92
93 Changes in 2.6.2
94
95 * issue 890 add support for Prometheus metrics and multiple HikariCP pools.
96
97 * issue 880 fix race condition caused by sorting collection while the condition of
98 sort can change.
99
100 * issue 876 add support for using a Prometheus CollectorRegistry other than the
101 default one.
102
103 * issue 867 support network timeout even for Connection.isValid().
104
105 * issue 866 mark commit state dirty when Connection.getMetaData() is called.
106
107 Changes in 2.6.1
108
109 * issue 821 if a disconnection class exception is thrown during initial connection
110 setup, do not set the flag that indicates that checkDriverSupport() is complete.
111
112 * issue 835 fix increased CPU consumption under heavy load caused by excessive
113 spinning in the ConcurrentBag.requite() method.
114
115 * issue 817 updated behavior of new initializationFailTimeout, please see the
116 official documentation for details.
117
118 * issue 742 add direct MXBean accessor methods to HikariDataSource for users who do
119 not want run run JMX.
120
121 Changes in 2.6.0
122
123 * Redesign of the contention code path resulting in doubling contended throughput; now
124 contended pool access retains 98% of the uncontended throughput.
125
126 * issue 793 add new HikariConfig method, setScheduledExecutor(ScheduledExecutorService),
127 and deprecate method setScheduledExecutorService(ScheduledThreadPoolExecutor). It is
128 unfortunate that the deprecated method has the more accurate name, but its signature
129 cannot be changed without breaking binary compatibility.
130
131 * issue 770 add a new property initializationFailTimeout, and deprecate configuration
132 property initializationFailFast.
133
134 * issue 774 significantly improve spike load handling.
135
136 * issues 518/769 add new metric for tracking how long physical connection acquisition is
137 taking. DropWizard histogram name "ConnectionCreation", and Prometheus summary name
138 "hikaricp_connection_creation_millis".
139
140 * issue 741 cancel HouseKeeper task on pool shutdown. If the ScheduledExecutor being used
141 did not belong to HikariCP, this task would remain scheduled after shutdown, causing a
142 memory leak.
143
144 * issue 781 more technically accurate wording of pool startup and shutdown log messages.
145
146 Changes in 2.5.1
147
148 * issue 719 only reset lastConnectionFailure after a successful dataSource.getConnection()
149 call.
150
151 * issue 716 do not scan deeper than 10 nested SQLExceptions, it's typically a trap ...
152 a chain that never terminates.
153
154 * issue 714 fix possible issue with cross-thread visibility. Change pool entry state from
155 AtomicInteger w/lazySet() to a volatile int with use of AtomicIntegerFieldUpdater.
156
157 Changes in 2.5.0
158
159 * Release 2.5.0 marks the start of a Java 8 HikariCP artifact. The Java 7 artifact is
160 now called "HikariCP-java7".
161
162 * HikariCP 2.5.0 and HikariCP-java7 2.4.8 have identical functionality.
163
164 Changes in 2.4.12
165
166 * issue 878 search for driverClass in both HikariCP class classloader and Thread Context
167 ClassLoader
168
169 Changes in 2.4.11
170
171 * issue 793 add new HikariConfig method, setScheduledExecutor(ScheduledExecutorService),
172 and deprecate method setScheduledExecutorService(ScheduledThreadPoolExecutor). It is
173 unfortunate that the deprecated method has the more accurate name, but its signature
174 cannot be changed without breaking binary compatibility.
175
176 * issue 600 ignore Java 8 default methods when generating proxy classes for Java 7.
177
178 Changes in 2.4.10
179
180 * Redesign of the contention code path resulting in doubling contended throughput; now
181 contended pool access retains 98% of the uncontended throughput.
182
183 * issue 770 add a new property initializationFailTimeout, and deprecate configuration
184 property initializationFailFast.
185
186 * issue 774 significantly improve spike load handling.
187
188 * issue 741 cancel HouseKeeper task on pool shutdown. If the ScheduledExecutor being used
189 did not belong to HikariCP, this task would remain scheduled after shutdown, causing a
190 memory leak.
191
192 Changes in 2.4.9
193
194 * issue 719 only reset lastConnectionFailure after a successful dataSource.getConnection()
195 call.
196
197 * issue 716 do not scan deeper than 10 nested SQLExceptions, it's typically a trap ...
198 a chain that never terminates.
199
200 * issue 714 fix possible issue with cross-thread visibility. Change pool entry state from
201 AtomicInteger w/lazySet() to a volatile int with use of AtomicIntegerFieldUpdater.
202
203 Changes in 2.4.8
204
205 * Release 2.4.8 marks the start of a Java 7 HikariCP artifact, HikariCP-java7, representing
206 support for Java 7 entering maintenance mode.
207
208 * Added Connection.commit() call to the fail-fast initialization for databases that
209 automatically start a new Connection in a transaction and throw an exception on close
210 if it is not committed.
211
212 * feature 694: report if a previously reported leaked connection is returned to the pool
213
214 * issue 689: log a warning if default transaction isolation level cannot be detected.
215 This can occur with pseudo-database drivers such as the one for JSonar
216
217 * issue 674: fix regression caused by pull request #450 (overzealous optimisation)
218
219 Changes in 2.4.7
220
221 * Miscellaneous stability improvements.
222
223 * Removed Oracle SQL state 61000, added specific error code (2399) to evict connections
224 when it is encountered.
225
226 * issue 664: do not recycle PoolEntry objects that have closed their held connection.
227
228 * issue 641, 643: reflection used method String.toUpperCase() without a Locale, which causes
229 problems in some locales such as Turkish.
230
231 * pull 632: added support for Prometheus metrics tracker.
232
233 * issue 650: detect Amazon Redshift connection refused error codes.
234
235 Changes in 2.4.6
236
237 * Added Oracle SQL error code 61000 (exceeded maximum connect time) to evict connections
238 when it is encountered.
239
240 * issue 621: fix NPE in shutdown if invoked during initialization.
241
242 * issue 606, 610: housekeeper thread was running before all class members were
243 initialized, causing an unrecoverable exception and disabling idle connection
244 retirement (maximumLifetime still applied).
245
246 Changes in 2.4.5
247
248 * issue 596: fix bug that occurs when minimumIdle is set to 0, but maximumPoolSize is
249 not explicitly set.
250
251 * issue 594: fix incompatibility with various libraries caused by storing a non-String
252 object in System properties.
253
254 * issue 593: improve logging when network timeout is not supported by the driver
255
256 * issue 591: improve thread-safety of Statement proxies
257
258 Changes in 2.4.4
259
260 * Generate unique sequential pool names, even across container classloaders to avoid
261 JMX name collisions when registering pools.
262
263 * Improve pool stability when running on computers using power-saving or sleep modes
264 where wake-up previously caused pool to grow to maximum size.
265
266 * Improve pool stability under severe thread-starvation conditions. Previous code
267 could interpret prolonged starvation (exceeding one minute) as non-contiguous clock
268 advancement, and reacted by soft-evicting connections (unnecessarily).
269
270 * Added connection timeout rate to Dropwizard metrics (ConnectionTimeoutRate).
271
272 * issue 563: Do not start the house-keeping thread until after pool initialisation
273 has succeeded.
274
275 * issue 559: Ensure the pool refill after house-keeping does not enqueue more add
276 connection requests if there are already minimumIdle requests pending.
277
278 * issue 555: include support for Java 8 interface 'default' methods during proxy
279 generation.
280
281 * issue 547: decreased allowable minimum connectionTimeout and validationTimeout to
282 250ms.
283
284 * issue 495: implemented iterator() method on custom FastList to support Tomcat
285 memory leak detection.
286
287 Changes in 2.3.13
288
289 * issue 512: reduce the number of calls made to Connection.getAutoCommit().
290
291 Changes in 2.4.3
292
293 * Improve pool shutdown behavior. Stop active connection acquisition once the
294 shutdown sequence has initiated.
295
296 * Improved detection and reporting of ambiguous pool configuration, when both the
297 connection URL and DataSource class names are specified.
298
299 Changes in 2.4.2
300
301 * Improve accuracy of timeouts for getConnection() calls by accounting for possibly
302 long delay aliveness tests.
303
304 * Improve adherence to minimumIdle goal by closing idle connections starting from
305 longest idle time to shortest. Additionally, stop when minimumIdle is reached even
306 if connections exceeding idleTimeout remain (but are still within maxLifetime).
307
308 * Introduce larger variance into maxLifetime to avoid mass connection closing and
309 subsequent new connection creation load on the database. Connections now have a
310 maximum lifetime between 97.5-100% of configured maxLifetime. In the case of the
311 default 30 minute lifetime, this generates actual lifetimes with a maximum deviation
312 of 45 seconds. Currently, no attempt is made to further avoid clustering that may
313 occur due to randomness.
314
315 * Ongoing com.zaxxer.hikari.metrics refactors. This is not considered public API until
316 such time as we announce it. Caveat lector.
317
318 * Performance improvements in the getConnection()/close() hot path.
319
320 * issue 452: fixed race condition when creating an rapidly ramping connections in the
321 pool.
322
323 * issue 415: removed use of java.beans classes to allow use of HikariCP with the
324 Zulu JRE compact3 profile.
325
326 * issue 406: execute validation query during connection setup to make sure it is
327 valid SQL.
328
329 Changes in 2.3.12
330
331 * Fixed issue with proxy generation whereby the generated classes contain the
332 major version number for Java 8, which makes them incompatible with the Java 7
333 runtime.
334
335 Changes in 2.4.1
336
337 * issue 380: housekeeper was not being scheduled in the case of a user specified
338 ScheduledExecutorService instance.
339
340 * issue 340: rollback change that elides setting the readonly property if the user
341 never explicitly configured it. See discussion in the Github issue tracker. Also
342 fixes binary ABI breakage between 2.3.9 and 2.4.0.
343
344 * issue 379: stop closing idle connections, to keep minimumIdle connections in pool
345
346 * issue 375: fixed InvalidPathException in HikariConfig
347
348 * issue 362: fixed NullPointerException in closing connection (closing statements)
349
350 * issue 357: allow altering the username & password through JMX at runtime
351
352 * issue 349: handle integer Transaction isolation level
353
354 * Throw SQLTransientConnectionException instead of SQLTimeoutException
355
356 * for validating connection, if network time out is set, do not set query timeout too
357
358 * ResultSet.getStatement() should return StatementProxy
359
360 Changes in 2.4.0
361
362 * Consolidated distribution into single JVM target (Java 7/8). Java 6 support has
363 entered maintenance mode, bug fixes will continue on the 2.3.x branch.
364
365 * Removed runtime dependency on Javassist by pre-generating proxy classes at build-time.
366
367 * Significantly reduced overhead, and increased reliability, of ConcurrentBag.
368
369 * Reduced garbage generation by 2-3x.
370
371 * Add connection soft-eviction and replacement if backward system clock motion or
372 significant forward jumps (greater than 1 minute) are detected.
373
374 * Pool configuration properties and DataSource methods previously marked as
375 @Deprecated have been removed.
376
377 * Deprecated HikariDataSource.shutdown() in favor of close().
378
379 * Improve shutdown performance.
380
381 * Allow user specified ScheduledThreadPoolExecutor for housekeeping timer. Useful
382 in applications with dozens or hundreds of pools in the same JVM.
383
384 * Reduce overhead and accuracy of Dropwizard gauges.
385
386 Changes in 2.3.7
387
388 * Try harder at resolving the driver by various means when both driverClassName and jdbcUrl
389 have been specified.
390
391 * Allow a specifically set DataSource instance to override other settings such as jdbcUrl,
392 dataSourceClassName, or driverClassName.
393
394 * Fixed issue where, in the case of a driver-based configuration (jdbcUrl), we were not
395 initialising the network timeout Executor.
396
397 * Fixed race condition uncovered during load-testing in which the connections in the pool
398 can spike to the maximum pool size when many connections reach their maxLifetime at the
399 same time.
400
401 Changes in 2.3.6
402
403 * Allow explicit definition of driverClassName to override DriverManager.getDriver(url)
404 located driver.
405
406 * Fixed a rare issue where a Connection that is held out of the pool, and never used by
407 the holding thread, upon returning to the pool might be given to another thread without
408 an aliveness test.
409
410 Changes in 2.3.5
411
412 * Fixed regression caused by enhancement #279 that imposed a runtime dependency on
413 Dropwizard metrics.
414
415 Changes in 2.3.4
416
417 * Fixed class cast exception when setting the HealthCheckRegistry via JNDI lookup.
418
419 * Allow Dropwizard MetricRegistry/HealthCheckRegistry to be set after pool startup --
420 one time only.
421
422 * Make logger in BaseHikariPool non-static and use getClass() to log messages as the
423 implementation class rather than as BaseHikariPool.
424
425 * Removed deprecation from connectionInitSql, it will be allowed.
426
427 * Made suspect/resume lock non-static (should be be shared across pools).
428
429 * Improved unwrap() behavior in the Hibernate HikariConnectionProvider.
430
431 * Improved leak detection log
432
433 Changes in 2.3.3
434
435 * Fixed bad interaction with PostgeSQL JDBC driver whereby a SQLException thrown by
436 PostgreSQL where the getNextException() call returns the original exception and causes
437 an infinite loop in HikariCP (and eventual stack overflow).
438
439 * Throw a typed Exception rather than a simple RuntimeException when pool initialization
440 fails.
441
442 * Allow Dropwizard Metrics and HealthChecks to be configured by a JNDI lookup.
443
444 Changes in 2.3.2
445
446 * Add support for Dropwizard HealthChecks through the introduction of two initial health
447 checks: ConnectivityCheck and Connection99Percent. See the Github project wiki for
448 documentation.
449
450 * Allow a lower maxLifetime setting of 30 seconds (compared to previous 120 second limit)
451
452 * Improve the message displayed when a connection leak is detected.
453
454 * Fixed a bug where Connection.setNetworkTimeout() was called on an already closed connection
455 resulting in a warning log from the AS400 JDBC driver.
456
457 Changes in 2.3.1
458
459 * Work around a bug in the MySQL Connector/J implementation of Connection.setNetworkTimeout()
460 that results in non-deterministic asynchronous application of the timeout, resulting in an
461 NPE from the MySQL driver when setNetworkTimeout() is followed immediately by close().
462
463 * Introduced a separate validationTimeout property, distict from connectionTimeout, to allow
464 greater control for some deployments that desire a long (or infinite) connectionTimeout
465 but expect the aliveness check to succeed for fail within a different (shorter) amount of
466 time.
467
468 Changes in 2.3.0
469
470 * Support pool suspend/resume to support certain failover scenarios.
471
472 * Fix theoretical race in JDBC 4.0 detection support.
473
474 * Improve shutdown() semantics to avoid exceptions as connections are forcefully
475 aborted.
476
477 * Unregister Codahale metrics at shutdown, if metrics are enabled.
478
479 * Major internal project layout restructuring to allow shared use of common code
480 between the Java 6/7 and Java 8 versions.
481
482 * Fixed bug where two pools in the same VM (and ClassLoading domain), using drivers
483 with differing JDBC support levels, would fail unless both pools were using
484 connectionTestQuery.
485
486 * Improved timeliness of maxLifetime evictions, while increasing performance of
487 getConnection() slightly as a side-effect.
488
489 * Fixed bug in HikariDataSource unwrap() semantics.
490
491 * Allow a lower leakDetectionThreshold of 2 seconds.
492
493 * Fixed bug when using the HikariJNDIFactory that required the presence of
494 Codahale metrics.
495
496 * Support initializationFailFast even when minimumIdle = 0
497
498 * Log internal pool inconsistencies rather than throwing exceptions that might
499 disrupt internal executors.
500
501 * Guard against poor or unreliable System.nanoTime() implementations.
502
503 Changes in 2.2.5
504
505 * Fixes for Java 6 compatibility.
506
507 * Implement full transaction state tracking. This allows HikariCP to bypass
508 the automatic rollback when connections are returned to the pool if the
509 transaction state is "clean".
510
511 * Rename MBean closeIdleConnections() to softEvictConnections() and implement
512 "evict on return" semantics.
513
514 * Fixed bug in code that sets HikariConfig values from a Properties instance
515 that prevented defaults from being read properly.
516
517 * Fixed an obscure bug in connection creation with a driver that throws an
518 exception when setTransactionIsolation() is called with the value returned
519 by getTransactionIsolation(). We now bypass setTransactionIsolation() if
520 the user has not configured an isolation level (using the default).
521
522 * Fix a bug where DataSource.loginTimeout() was always being set to 1 second.
523
524 * Fix bug where some drivers return 0 from Connection.getNetworkTimeout(),
525 and yet throw SQLFeatureNotSupportedException when setNetworkTimeout() is
526 called. This broke they way that HikariCP had implemented JDBC 4.1 support
527 detection.
528
529 Changes in 2.2.4
530
531 * Generate proxy classes into the same protection domain as the HikariCP
532 loaded classes. This solves issues with signed jars.
533
534 * Improve accuracy of pool statistics available to JMX and logged at debug
535 level (at a slight performance cost).
536
537 * Fixed issue where after a database down condition, and when minimumIdle is
538 set to 0, when the database connectivity is restored the connections could
539 ramp up to the maximum pool size. Eventually, idleTimeout and maxLifetime
540 would restore normal pool conditions, but it was still undesirable behavior.
541
542 * Improved connection timeout handling by using Connection.setNetworkTimeout()
543 if available (JDBC 4.1).
544
545 * driverClassName is no longer a required property when jdbcUrl is specified.
546 Omitting this property only works for compliant drivers.
547
548 * Add auto-detection of support for Statement.setQueryTimeout() used in the
549 alive check. Fixes failures with test queries on the PostgreSQL driver
550 when not using JDBC4 isValid() alive checks.
551
552 * The pool now defaults to fail-fast initialization. If you need to start
553 your application without/before the database, you will need to explicitly
554 set initializationFailFast to false.
555
556 * Dropwizard/Codahale metrics are now supported via the setMetricRegistry()
557 method in HikariConfig and in HikariDataSource.
558
559 * Fixed issue with pool initialization of MySQL after default value of
560 initializationFailFast property was changed to false.
561
562 * Further shadow runtime dependency on Codahale metrics from reflection
563 performed by Spring and other IoC containers.
564
565 * Fix issue where network timeout was not properly restored to its default
566 value after modifying it for the duration of the addConnection() method.
567
568 Changes in 2.1.0
569
570 * Significant internal refactor supporting creation of new proxy instances
571 (throwaway) around Connections for each call to getConnection(). This
572 can avoid issues where a thread continues to try to use a connection
573 after it is closed [returned to the pool].
574
575 * Allow HikariConfig(String propertyFileName) to load properties file from
576 classloader as a stream, with fall-back to the file-system.
577
578 * Allow loading of properties file specified by -Dhikaricp.configurationFile
579 system property when using the default HikariConfig() or HikariDataSource()
580 constructors.
581
582 * Fixed accounting issue with totalConnections when aborting connections
583 during shutdown, causing a warning message to be logged.
584
585 * Fixed regression in Java 8 codeline that would prevent minimumIdle from
586 being set before maxPoolSize.
587
588 * Fixed regression with Tomcat carping about ThreadLocal variables held after
589 web application restart
590
591 * Change to make HikariConfig.getTransactionIsolation()/setTransactionIsolation()
592 follow proper bean semantics.
593
594 * Fixed issue where connections created in the pool would skip the alive check
595 the first time they were used.
596
597 Changes in 2.0.1
598
599 * Split project into Java 6/7 and Java 8 components.
600
601 * Fixed issue in JNDI object factory which would not allow JNDI-defined
602 DataSource properties to pass-thru to the pool.
603
604 * Fixed issue where under certain conditions getConnection() could
605 timeout prematurely.
606
607 * Fixed issue where user-defined pool name would be overridden by the
608 automatically generated name.
609
610 * Fixed NPE when one of either username and password is defined, and the
611 other is null.
612
613 * Fixed issue tracking the statements when there are mixed statement
614 types (Statement, PreparedStatement, etc.) open on the connection and
615 the number of unclosed statements exceeds 32.
616
617 * Fixed issue where housekeeping threads would add idle connections
618 even when minimumIdle was 0.
619
620 * Fixed issue where Wrapper.isWrapperFor() and Wrapper.unwrap() calls
621 did not recurse as per specification.
622
623 * HikariDataSource now implements the Closable interface.
624
625 * Integrated change to allow specifying a ThreadGroup for thread
626 creation is certain restricted environments.
627
628 Changes in 1.4.0
629
630 *) Fix bug that did not allow minIdle property to be set.
631
632 Changes in 1.3.9
633
634 *) Added pool name to housekeeping thread name to make thread dumps
635 more meaningful in containers with multiple pools.
636
637 *) Improved shutdown semantics; make a concerted effort to close
638 idle connections and abort or close active connections.
639
640 *) Performance enhancements.
641
642 Changes in 1.3.8
643
644 *) Fixed incorrect logic when using JDBC4 isValid() test for alive
645 status of connection.
646
647 Changes in 1.3.7
648
649 *) Added JNDI object factory (com.zaxxer.hikari.HikariJNDIFactory)
650 for Tomcat and other containers that prefer JNDI-registered DataSource
651 factories.
652
653 *) Fix NPE that can occur when connections cannot be created and
654 callers to getConnection() timeout.
655
656 *) Various bug fixes and minor enhancements.
657
658 Changes in 1.3.6
659
660 *) Include connection failure cause in calls to getConnection() that
661 timeout (due to connection failure). Removed chatty logging.
662
663 *) Java8 Compatibility fixes.
664
665 *) Include pool name in logging messages. Thanks for the contribution
666 @jaredstehler.
667
668 Changes in 1.3.5
669
670 *) Fixed a regression in the Javassist code generation.
671
672 *) Various bug fixes and minor enhancements.
673
674 Changes in 1.3.4
675
676 *) Added new property isolateInternalQueries used to control whether
677 internal pool queries such as connection alive tests are isolated
678 in their own transaction.
679
680 *) Added properties for DriverManager (driverClassName) and JDBC URL-based
681 (jdbcUrl) configuration. 1999 called and wants its JDBC driver back.
682
683 *) Added new username and password properties to allow default authentication
684 for connections.
685
686 *) Added support for the getConnection(username, password) method signature
687 to HikariDataSource.
688
689 *) Added new property readOnly to control the default read-only status
690 of connections in the pool.
691
692 *) Deprecated acquireIncrement property.
693
694 *) Deprecated acquireRetries property.
695
696 *) Deprecated acquireRetryDelay property.
697
698 *) Deprecated minimumPoolSize property.
699
700 *) Added new property minimumIdle used to control the minimum number of
701 idle connections the pool should try to maintain on a running basis.
702
703 *) Added evictConnection(Connection) method to HikariDataSource to allow
704 special cases where users wish to forcibly eject a connection from
705 the pool. To use used cautiously, read the JavaDoc before using.
706
707 *) Various bug fixes and minor enhancements.
708
709 Changes in 1.3.3
710
711 *) Removed shared state contention that was causing excessive CPU cache-line
712 flushing. Nearly 4x improvement in Connection acquisition/release
713 performance.
714
715 *) Fixed issue with Tomcat carping about ThreadLocal variables held after
716 web application restart.
717
718 *) Fixed issue where the maximum configured connections could be overrun
719 during large burst requests for connections.
720
721 Changes in 1.3.2
722
723 *) Java 6 compatibility.
724
725 *) HikariDataSource now extends HikariConfig, so pool properties can be
726 set directly on a HikariDataSource without need to create a HikariConfig.
727 The cost of doing so is a small runtime cost due to the fact that an
728 "is initialized" check must be performed on every invocation of
729 getConnection() due to lazy initialization of the pool.
730
731 *) Added Sybase-specific disconnect error codes to SQLException snooping.
732
733 *) Added HikariConfig.setCatalog() method to set DataSource catalog.
734
735 *) Add DataSource.close() method that is synonymous with shutdown().
736
737 *) More performance improvements (never ending).
738
739 *) Moved benchmarks to https://github.com/brettwooldridge/HikariCP-benchmark
740
741 Changes in 1.3.0
742
743 *) Pool is now backed by a custom lock-less ConcurrentBag that
744 provides superior performance to LinkedBlockingQueue and
745 LinkedTransferQueue for usage patterns of connection pools.
746
747 *) Fixed bugs reported against the 1.2.9 release.
748
749 *) Added more detailed logging for broken connections and failures
750 during new connection creation.
751
752 Changes in 1.2.9
753
754 *) Added a fail-fast option for pool initialization. If enabled, a
755 RuntimeException will be thrown if there are errors during pool
756 initialization.
757
758 *) Made the registration of the HikariCP MBeans optional. They now
759 default to not being registered. Registering them causes a minor
760 performance hit due to additional pool data collection in
761 HikariDataSource.getConnection().
762
763 *) Added the SQLException message to the log entry when an exception
764 occurs during physical connection acquisition.
765
766 *) Implemented an orderly shutdown of the pool via the shutdown() method
767 on HikariDataSource.
768
769 *) Listened to "Adele - Live At The Royal Albert Hall" on endless loop.
770
771 Changes in 1.2.8
772
773 *) Fixed a critical bug introduced in 1.2.7 occurring when the number of
774 concurrently open statements exceeds sixteen.
775
776 Changes in 1.2.7
777
778 *) Finally achieved performance parity between the generated delegates
779 and the former method of instrumenting driver classes directly.
780
781 *) Improved generated delegate code. Removed unnecessary casts, moved
782 to a static proxy factory rather than a singleton (performance win).
783
784 *) Improved performance of FastStatementList (primary source of speed-up
785 to reach parity with former instrumentation code).
786
787 *) Removed aliveness check on connection creation.
788
789 *) Track connection isolation level and only reset if the state has
790 become "dirty". Avoids unnecessary round trip to the DB during the
791 aliveness check.
792
793 *) Added interface IConnectionCustomizer and related HikariConfig
794 property 'connectionCustomizerClassName' to allow users to specify
795 a connection customization implementation.
796
797 Changes in 1.2.6
798
799 *) Fixed regression that caused IndexOutOfBounds when multiple unclosed
800 Statements existed at the time of Connection.close().
801
802 *) Fixed incorrect pom.xml dependency on Hibernate.
803
804 Changes in 1.2.5
805
806 *) Instrumentation mode (agent) removed due to narrowing gap between
807 delegation mode and instrumentation (and to simplify the code base).
808
809 *) Added setDataSource() to HikariConfig to allow a DataSource instance
810 to be explicitly wrapped by the pool. Only available when creating
811 HikariConfig programmatically or constructing HikariConfig from a
812 java.util.Properties instance.
813
814 *) Fixed Hibernate threading issue (certain usage patterns) introduced
815 in 1.2.2.
816
817 *) Fixed issue observed with PostgreSQL whereby the query that tests the
818 connection for "aliveness" also starts a transaction (when auto-commit
819 is false), thereby causing a later failure when we tried to set the
820 transaction isolation level.
821
822 *) Fixed issue where idleTimeout could not be set to 0, thereby disabling
823 it. Incorrect value validation caused 0 to be rejected as a valid value.
824
825 Changes in 1.2.4
826
827 *) Fix another Hibernate-related issue whereby an NPE is encountered when
828 a thread that was not the thread that obtained a Connection tries to
829 interact with that Connection.
830
831 Changes in 1.2.3
832
833 *) Fix internal (but suppressed) exception during rollback of connections
834 returned to the pool with auto-commit turned off.
835
836 *) Fix a reflection issue that causes Hibernate failures due to the
837 CallableStatement interface being incorrectly injected into statement
838 proxies that are PreparedStatement or Statement instances.
839
840 Changes in 1.2.2
841
842 *) Perform a rollback() on connections returned to the pool with
843 auto commit disabled.
844
845 *) Add a constructor for HikariConfig that accepts a Properties
846 object.
847
848 *) Speed improvements for delegate mode.
849
850 *) Fix a bug where connection timeouts could not be disabled.
851
852 *) Permit setting the DataSource logWriter either on the HikariDataSource
853 or via addDataSourceProperty() on the HikariConfig.
854
855 *) Add transactionIsolation property to allow setting the default
856 transaction isolation level for connections.
857
858 Changes in 1.2.1
859
860 *) Clear SQL warnings before returning a connection to the user.
861
862 *) Added asynchronous connection backfill strategy that triggers
863 when the pool becomes empty as a result of dispatching a
864 connection.
865
866 *) Changed default acquireIncrement to 1, set minimum timeout of
867 100ms for acquiring a connection.
868
869 Changes in 1.1.9
870
871 *) Added connectionInitSql property to allow setting connection
872 properties when a new connection is created.
873
874 *) Added setDataSourceProperties() setter to HikariConfig to
875 allow easier configuration though Spring.