Commit Graph

2248 Commits (fb95531f07a0ed4deb85a8d5286414b304f93652)
 

Author SHA1 Message Date
Nitin 1b1b63fb3e renamed ProxyLeakTask as it has nothing to do with 'proxy' 9 years ago
Ron Klein 1f5b6c72ec Update README.md
typo
9 years ago
Nitin a6486c6417 fix: set last access even for readonly and set isCommitStateDirty wisely 9 years ago
Nitin 4065283619 better variation 9 years ago
Nitin 9d34443031 bug fix: do not call next exception if delegate == closed 9 years ago
Nitin 1a9494596f test failure fix 3 9 years ago
Nitin 5a41fee89b test failure fix 2 9 years ago
Nitin 5238d2d4c4 fix test failure 9 years ago
Nitin bd3e570192 added, corrected validation and test 9 years ago
Nitin 3095b6dc9e avoid re-setting twice in validation 9 years ago
Nitin 7b1301eeec removed unused import 9 years ago
Nitin 47ac982cc3 create leak task first and reverted long to int change 9 years ago
Nitin 2e5dd85133 replaced volatile long with int and import static TimeUnit.* 9 years ago
Nitin 0f75311732 using UNINITIALIZED 9 years ago
Nitin 2b6da07d26 cleanup 9 years ago
Nitin 82f7c15ace changes as suggested 9 years ago
Nitin 838d929067 using 'Failed' like every where else, instead of 'Unable'. 9 years ago
Nitin 6436870bea separate validationTimeout and connectionTimeout and validation, init
networkTimeout only once.
9 years ago
Brett Wooldridge 6e06620976 I didn't like this change, but missed it in the review. Readability over performance in this case. Not to mention that the JVM is likely to hoist the instanceof check and cast out of the loop because it is final invariant. 9 years ago
Brett Wooldridge 1f244af32f Issue #547 Cleanup and additional logging. 9 years ago
Brett Wooldridge fb09f10298 Log a warning (once) when it is not possible to honor a validation timeout of less than 1 second. 9 years ago
Brett Wooldridge 0e3677afa3 Use same terminology as other validation log messages, ie. "is less than" or "is more than". 9 years ago
Brett Wooldridge 6b535d542d Merge branch 'dev' of github.com:nitincchauhan/HikariCP into dev 9 years ago
Nitin e46034f56d bugfix: calling sValid() with 0 is infinite. made minimum 1sec 9 years ago
Brett Wooldridge 62fdb2c6c3 Merge pull request #567 from nitincchauhan/dev
avoid 'more' exceptions after 'un-recoverable' exception thrown before for the connection
9 years ago
Nitin e7da3f5dc7 reverted as suggested 9 years ago
Brett Wooldridge d75c69a51b [maven-release-plugin] prepare for next development iteration 9 years ago
Brett Wooldridge d2a348aee8 [maven-release-plugin] prepare release HikariCP-2.4.4-RC1 9 years ago
Brett Wooldridge d82663b1ca Prepare RC1 9 years ago
Brett Wooldridge af2546b837 Prepare RC1 9 years ago
Nitin 0d0eec16ea moved check to call(), start thread in other loop 9 years ago
Brett Wooldridge 501f2ab1d8 Updated for 2.3.13 9 years ago
Nitin 14ceda3b0c for re-throwing exception, setting log level to error 9 years ago
Nitin 4c2486e4a0 removed static log, like parent PoolBase. checking instanceof out of
loop
9 years ago
Nitin a007e730be race test: exit loop on exception 9 years ago
Nitin d3bc0a155a it is 1 thread so have queue size that mostly do not require to queue
jobs on caller
9 years ago
Nitin 3610590d64 poolName was missing in few logs 9 years ago
Nitin ba09a68fe1 better thread names and log in validateNumerics() with poolName 9 years ago
Nitin f46fc1e9e5 setting allowCoreThreadTimeOut first 9 years ago
Nitin 8bece2de76 consistent name for threads 9 years ago
Nitin c2523133d4 over-riding method instead of passing true as default from other places 9 years ago
Nitin 2bc3d98901 unique 'closing' reason for each situation 9 years ago
Nitin 76d490d2de removed explicit call to fillInStackTrace() 9 years ago
Nitin ce03fc68c9 removed word 'pool' from thread name 9 years ago
Nitin e156b0d719 removed misleading 'extra' comma 9 years ago
Nitin 9936e95afe moved log 'closed' at the very end 9 years ago
Nitin f23f358ad4 avoid 'more' exceptions after connection is detected throwing
un-recoverable exception
9 years ago
Brett Wooldridge ad17fb09d1 Update CHANGES 9 years ago
Brett Wooldridge e350d1f15f Issue #566 Remove separate Counter metric for timeouts, count is implicitly captured and available from the new Meter metric for timeout rates. 9 years ago
Brett Wooldridge 650ed4c2e2 Update CHANGES 9 years ago