summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
...
* QPID-3158 - Defect in the CRAM-MD5-HEX mechanism - CRAMMD5HexInitialiser ↵Robert Gemmell2011-03-315-11/+244
| | | | | | | | fails to pad bytes in range 0A-0F with leading zero. Add testcase to test CRAM-MD5-HEX mechanism. Guard against nulls in SASL SaslServerFactory.getMechanismNames implementations to avoid dependency on mechanism registration order. Applied patch from Keith Wall <keith.wall@gmail.com> git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1087249 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3129: cluster_tests.LongTests.test_failover hangsAlan Conway2011-03-302-11/+28
| | | | | | | | | | | | | | | | | Fix is a race condition in posix/Socket.cpp Socket::connect. When connecting to a port on the same host which no longer has a process associated with it the OS occasionally chooses the remote port (which is unoccupied) as the port to bind the local end of the socket, resulting in a "circular" connection. This seems like something the OS should prevent but I have confirmed that the sporadic hangs in cluster_tests.LongTests.test_failover on RHEL5 are caused by such a circular connection. The fix is to detect circular connections and raise an error. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1087052 13f79535-47bb-0310-9956-ffa450edef68
* qpid-3171Michael Goulish2011-03-301-1/+1
| | | | | | | | | | | | | | | | | | | | The registration of the codec happens on a different thread from the use of the codec. It is possible for the registration to occur after the first attempted use. In my testing, this happened 3% of the time -- 165 times out of 5000 tests -- when using RDMA transport, and 0 times out of 5000 when using TCP. Which is why we didn't notice it earlier. We have a function that tells when we are ready to encode -- CyrusSecurityLayer::canEncode. But it does not check the validity of the codec pointer before using it, so it cores in this situation. I believe simply checking that pointer is probably the best solution. Introducing that check caused the crash not to show up in 10,000 trials. There were also no hangs. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1087047 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3167: add a unit test of SimpleAMQQueue#processQueue to check delivery ↵Robert Gemmell2011-03-3023-77/+264
| | | | | | when subscriptions with unique selectors are in use git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1087000 13f79535-47bb-0310-9956-ffa450edef68
* NO-JIRA: add missing ASF licenceRobert Gemmell2011-03-291-0/+20
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1086777 13f79535-47bb-0310-9956-ffa450edef68
* NO-JIRA: add missing ASF licenceRobert Gemmell2011-03-291-0/+20
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1086769 13f79535-47bb-0310-9956-ffa450edef68
* NO-JIRA - Fixed case error in the automake file.Ted Ross2011-03-291-1/+1
| | | | | | | This caused a missing dependency in the qpidtypes library. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1086663 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3166: updates based on review feedback from Keith WallRobert Gemmell2011-03-281-25/+24
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1086169 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3162: move unregistration of ServerConnections from the ↵Robert Gemmell2011-03-271-3/+11
| | | | | | ConnectionRegistry into the setState() method to ensure it occurs in all cases. Also move registration into setState(). git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1086041 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3165: ensure all subscriptions are checked before making the decision ↵Robert Gemmell2011-03-271-22/+69
| | | | | | on whether to stop delivering. Use a boolean instead of doing a 0/1 toggle and update variables to generally clarify logic. Use an int instead of a long for the iteration decrementing git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1086040 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3166: add system test using multiple batch transacted producers with ↵Robert Gemmell2011-03-272-2/+250
| | | | | | multiple consumers using unique selectors. Exposes issue detailed in QPID-3165. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1086039 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3156: Use MapAccessor to retrieve Boolean propertiesAndrew Donald Kennedy2011-03-241-15/+3
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1085187 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3169: add usage message if command is unknown.Kenneth Anthony Giusti2011-03-241-0/+3
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1085147 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3168: fix queue flow control --help textKenneth Anthony Giusti2011-03-242-204/+159
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1085065 13f79535-47bb-0310-9956-ffa450edef68
* NO-JIRA - ACL Modules "reload" method now returns OK status if there was no ↵Ted Ross2011-03-231-1/+4
| | | | | | error. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1084768 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3156Rajith Muditha Attapattu2011-03-231-1/+8
| | | | | | | | The address parser treats 'true' as a string while it treats 'True' as a boolean. Therefore I modified the AddressHelper class to expect both forms. The MapAccessor class used for the rest of code already handles this case. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1084462 13f79535-47bb-0310-9956-ffa450edef68
* QPID-2930Rajith Muditha Attapattu2011-03-232-0/+4
| | | | | | | Excluding the testGetPropertyNames test case until we make a decision on whether to apply the same fix for 0-8 codepath. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1084420 13f79535-47bb-0310-9956-ffa450edef68
* QPID-2930Rajith Muditha Attapattu2011-03-221-1/+1
| | | | | | | | Removed the explicit use of addressing in the test case to enable it to be run under any profile. This seems better than excluding it from the default and 0.8 profiles as the test doesn't really have anything to do with addressing. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1084414 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3159 - Applied patch from Angus SalkeldTed Ross2011-03-221-2/+2
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1084359 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3160: Remove file/directory defaults from pre-generated man page.Alan Conway2011-03-222-69/+71
| | | | | | | | | | These paths will be different depending on build configuration so they can't be included in a static man page. They will still be shown by qpidd --help. Also updated the instructions in Makefile.am to include this edit step. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1084339 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3160: Check in pre-generated man pageAlan Conway2011-03-224-6/+313
| | | | | | | | | | | | | Previously the man page was generated automatically from the output of qpidd --help. To make a distribution, you had to build the entire source tree just to generate the man page. This automation was of limited value as the core broker options do not change often. This commit includes a pre-generated man page. At each release we will manually re-generate and check in a new man page. Instructions are provided in docs/man/Makefile.am git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1084282 13f79535-47bb-0310-9956-ffa450edef68
* NO-JIRA - Fixed two problems in the qmf-interop tests:Ted Ross2011-03-221-1/+2
| | | | | | | | | | 1) Added a 2-second delay on console startup. This closes the window on a race condition that occurs when the console and agent start at exactly the same time. Note that this race has been designed out of the QMFv2 protocol. 2) Fixed a reference to an incorrect variable name in the assert(cond) message. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1084053 13f79535-47bb-0310-9956-ffa450edef68
* QPID-2930Rajith Muditha Attapattu2011-03-221-3/+13
| | | | | | | | | I had checked in an incomplete test case with rev 1082719 I had made certain changes to the test case for testing another potential fix, but forgot to revert it back to the original before committing. Re applied the changes to adequately test the issue. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1084048 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3147: Added missing qpid/DisableExceptionLogging.h to source list in ↵Alan Conway2011-03-211-0/+1
| | | | | | Makefile.am. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1083961 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3147: Misconfigured tracing/logging can lead to hung threads in logging ↵Alan Conway2011-03-213-5/+62
| | | | | | | | | | | stack The hang was caused by re-entrant attempts to initialize the Logger singleton when an exception was thrown during logger configuration. The fix is to disable exception logging temporarily while the logger is constructed. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1083884 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3155: Install module libraries during install-exec phaseAlan Conway2011-03-186-22/+22
| | | | | | | | The C++ automake system has 2 install phases, exec and data. Module libraries are installed incorrectly in the data phase. Fix the makefiles to install them in the exec phase. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1083077 13f79535-47bb-0310-9956-ffa450edef68
* NO-JIRA: Fix qpidd late/overran warnings.Alan Conway2011-03-182-9/+13
| | | | | | | Warnings for late-and-overrun tasks were not being correctly reported. Lateness for overrun tasks below the late threshold were not being reported. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1082954 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3153Michael Goulish2011-03-184-4/+8
| | | | | | | added check for ais_exec running -- or don't run these clustered tests git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1082812 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3152Michael Goulish2011-03-181-6/+7
| | | | | | | The non-clustered versions of the sasl_fed_ex tests should probably not attempt to load cluster.so. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1082804 13f79535-47bb-0310-9956-ffa450edef68
* Added a test case for the issue highlighted by QPID-2930Rajith Muditha Attapattu2011-03-171-0/+28
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1082719 13f79535-47bb-0310-9956-ffa450edef68
* QPID-2959Rajith Muditha Attapattu2011-03-173-6/+16
| | | | | | | | | Added a temp hack to get the testReplyToWithNamelessExchange passing with the java.0.10 test profile. In doing so I realized that "qpid.declare_exchanges" has no affect on the producer side. Therefore I added a check in BasicMessageProducer_0_10.java to check this flag before issuing an exchange_declare. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1082715 13f79535-47bb-0310-9956-ffa450edef68
* Added testReplyToWithNamelessExchange as a test case for QPID-2959Rajith Muditha Attapattu2011-03-171-33/+32
| | | | | | | | | | | | Added testReplyToWithCustomExchange as a test case for QPID-3011 Removed testAddressBasedReplyTo as the above test cases cover it adequately. Currently the testReplyToWithNamelessExchange fails with the java.0.10 test profile. This is due to the default being BURL and the replyTo address is evaluated as a BURL and as part of the logic an exchange declare is being set. The C++ broker does seem to ignore the exchange declare of a known type (the nameless exchange in this case), while the java broker throws an exception. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1082709 13f79535-47bb-0310-9956-ffa450edef68
* only assert if type is specifiedRafael H. Schloming2011-03-172-2/+5
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1082697 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3150Michael Goulish2011-03-171-1/+1
| | | | | | | | This is a test-only fix. Tell the client explicitly to use DIGEST-MD5, rather than possibly defaulting to a mech that the test has not prepared for. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1082685 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3116: rubygen with ruby 1.9Alan Conway2011-03-171-1/+2
| | | | | | Fix name mangling for reserved "type" symbol to work on ruby 1.8 and 1.9. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1082670 13f79535-47bb-0310-9956-ffa450edef68
* NO-JIRA: Put a timeout on calls to Thread.join to avoid tests hanging ↵Alan Conway2011-03-171-3/+7
| | | | | | indefinitely. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1082668 13f79535-47bb-0310-9956-ffa450edef68
* added support for address assertionRafael H. Schloming2011-03-172-1/+14
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1082478 13f79535-47bb-0310-9956-ffa450edef68
* added tcp_nodelay connection optionRafael H. Schloming2011-03-174-13/+20
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1082470 13f79535-47bb-0310-9956-ffa450edef68
* changed python clients durable default to FalseRafael H. Schloming2011-03-161-1/+1
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1082290 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3142: enable tests to run against a copy of an existing work directoryRobert Gemmell2011-03-153-2/+34
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1081858 13f79535-47bb-0310-9956-ffa450edef68
* NO-JIRA: Refactor RefCountedBuffer to be simpler and more flexible.Alan Conway2011-03-156-81/+98
| | | | | | | | | | RefCountedBuffer::pointer is replaced by a separate class BufferRef. BufferRef is independent of RefCountedBuffer, it can be used to manage other reference-counted buffer classes e.g. to implement buffer pools. BufferRef also provides beginning and end of the buffer, not just beginning. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1081631 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3144: Remove "loaded module" info log message.Alan Conway2011-03-141-2/+1
| | | | | | | | | | | | | These messages, printed at plugin load time, interfere with qpidd --check. It is supposed to print the PID of a running qpidd daemon to stdout, for use in scripts and the like. However with log-to-stdout enabled these messages obscure the PID. They are in any case of little value, individual plugins can log their start-up messages on initialization time which does not cause a problem git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1081548 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3109Rajith Muditha Attapattu2011-03-141-0/+16
| | | | | | | Added a test case to cover a producer created with a null destination (where a valid destination is supplied each time send is invoked). git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1081467 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3143Rajith Muditha Attapattu2011-03-142-1/+15
| | | | | | | | | | | Removed the line which sets the create option to 'always'. Please note that the default create option is set to 'never'. Therefore if just a queue name is specified the client will throw an exception if that queue is not present in the broker. However this check will only be performed when a producer or consumer is created using that destination. Also modified the existing test case to verify the default and the explicit create case. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1081460 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3138 some cosmetic cleanup & correction for IVECarl C. Trieloff2011-03-142-24/+34
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1081450 13f79535-47bb-0310-9956-ffa450edef68
* NO-JIRA: Minor improvements to test scripts qpid-cluster-benchmark and ↵Alan Conway2011-03-144-40/+76
| | | | | | | | | qpid-cpp-benchmark - qpid-cluster-benchmark: added command line options. - qpid-cpp-benchmark: clean up error handling, fixed a race condition. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1081396 13f79535-47bb-0310-9956-ffa450edef68
* NO_JIRA: Removed faulty test, test_dr_no_message was failing sporadically.Alan Conway2011-03-121-19/+0
| | | | | | | | test_dr_no_message was using unreliable sleeps to try to determine expiry of messages. There's no obvious way to implement the test correctly. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1080945 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3129: cluster_tests.LongTests.test_failover hangsAlan Conway2011-03-112-128/+83
| | | | | | | - simplified brokertest.py using subprocess.Popen file redirection instead of threads. - fixed the hang in test_failover git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1080786 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3137: ensure old api examples are installed into correct locationGordon Sim2011-03-118-7/+9
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1080679 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3136: Add option to disable defaults for queue threshold alertsGordon Sim2011-03-115-12/+15
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk/qpid@1080589 13f79535-47bb-0310-9956-ffa450edef68
View Lorry Controller Status