summaryrefslogtreecommitdiff
path: root/qpid/cpp/src
Commit message (Collapse)AuthorAgeFilesLines
* Fix QPID-4269Stephen D. Huston2012-09-111-1/+34
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1383229 13f79535-47bb-0310-9956-ffa450edef68
* QPID-4194 : re-enable queue events in CATCHUP state.Michael Goulish2012-08-141-1/+6
| | | | | | | | pavel moravec's fix -- without this, newbie broker with a replication queue will not replicate messages received during CATCHUP. git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1373004 13f79535-47bb-0310-9956-ffa450edef68
* QPID-4142 C++ broker connection counting confused using auth. Merge from ↵Charles E. Rolke2012-08-136-0/+63
| | | | | | trunk to 0.18. git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1372597 13f79535-47bb-0310-9956-ffa450edef68
* NO-JIRA: Trivial correction to log message.Alan Conway2012-08-071-1/+1
| | | | | | Log message logic was inverted, making the message confusing. git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1370394 13f79535-47bb-0310-9956-ffa450edef68
* NO-JIRA: Remove debug logginng statement left in by mistake.Alan Conway2012-08-071-1/+0
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1370313 13f79535-47bb-0310-9956-ffa450edef68
* QPID-4191: HA removing self address breaks if a VIP is used.Alan Conway2012-08-0711-67/+68
| | | | | | | | | | | | | | | | | | | Pre this patch the HA broker removed its own address from the set of cluster addresses to form the set of failover addresses. The goal was avoid useless self-connection attempts. However this was broken with a Virtual IP address where a single address is used for the entire cluster. The remove-self is not essential, self-connection attempts are prevented elsewhere. Backup brokers will be prevented from connecting to self by the same connection-observer as normal clients, and this patch addes self-connection checks ins This patch - removes the code to remove self-addresses - adds self-connection checks in ConnectionObserver - adds & reorders some log statements & comments for greater clarity. git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1370308 13f79535-47bb-0310-9956-ffa450edef68
* Revert "WIP: BrokerReplicator, QueueReplicator error handling."Alan Conway2012-08-014-12/+5
| | | | | | This reverts commit r1367543, which was committed by accident. git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1368145 13f79535-47bb-0310-9956-ffa450edef68
* QPID-4175: HA code rationalize loggingAlan Conway2012-07-316-18/+34
| | | | | | | | | | | Clean up and rationalize log messages and levels. notice: Major broker-level events: connecting, failing-over, primary active, backup ready. info: Major queue level events: subscriptions ready, replicators created etc. debug: Detailed replication events: accept/reject conections, details of queue replication protocol. trace: dumping raw QMF messages git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1367547 13f79535-47bb-0310-9956-ffa450edef68
* WIP: BrokerReplicator, QueueReplicator error handling.Alan Conway2012-07-314-5/+12
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1367543 13f79535-47bb-0310-9956-ffa450edef68
* NO-JIRA: Fix typo in makefiles, incorrect file name.Alan Conway2012-07-302-2/+2
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1367106 13f79535-47bb-0310-9956-ffa450edef68
* QPID-4107 HA does not replicate alternate-exchangeAlan Conway2012-07-266-55/+244
| | | | | | | | Set alternate exchange on replicated queues and exchanges. If the exchange is available, set it immediately. Otherwise remember what needs to be set so it can be set when the exchange becomes available. git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1366206 13f79535-47bb-0310-9956-ffa450edef68
* QPID-4159: HA missing messages in failover test.Alan Conway2012-07-267-34/+40
| | | | | | | | | | | | Fix test_failover_send_receive showing missing messages. With this fix, ran with -DDURATION=2 overnight with no failures. - Primary, RemoteBackup: Only report "ready" once per remote backup. - HaBroker: Put membership updates under mutex. - ReplicatingSubscription: Check for backup missing messages at the front. - ha_tests.py: Added assertion to test_priority_ring, verify primary queue as expected. git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1366180 13f79535-47bb-0310-9956-ffa450edef68
* QPID-4165: Fix unsused variable warnings when compiling with -DNDEBUGAlan Conway2012-07-243-5/+5
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1365251 13f79535-47bb-0310-9956-ffa450edef68
* NO-JIRA: Fix typos, update comments, update log messages.Alan Conway2012-07-247-15/+32
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1365046 13f79535-47bb-0310-9956-ffa450edef68
* QPID-4163: Deadlock between LinkRegistry and TimerTask locksAlan Conway2012-07-242-10/+12
| | | | | | | | Previously TimerTask::cancel was being called with locks held in this stack trace, causing a deadlock. Moved call to cancel outside of the lock. git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1365045 13f79535-47bb-0310-9956-ffa450edef68
* QPID-4159: HA Missing messages in failover test.Alan Conway2012-07-243-10/+27
| | | | | | | | | | | | QueueGuard was taking its snapshot of the initial queue range *before* it registered its QueueObserver. That means it was possible to have unguarded messages between the end of the snapshot and the first position protected by the guard. Fixed race condition in QueueRange constructor: Must call getPosition() *after* getFront() since both may be advancing and we want to end up with a valid range front <= back+1. git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1365044 13f79535-47bb-0310-9956-ffa450edef68
* QPID-4158: HA transition to CATCHUP status too early; merged from 1363520Gordon Sim2012-07-201-16/+4
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1363852 13f79535-47bb-0310-9956-ffa450edef68
* QPID-4156: HA close window for clients to connect before HA broker is ↵Gordon Sim2012-07-203-10/+31
| | | | | | initialized; merged r1363498 git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1363839 13f79535-47bb-0310-9956-ffa450edef68
* NO-JIRA: Removed use of boost::bind which was causing build failure on windows.Alan Conway2012-07-191-2/+2
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1363489 13f79535-47bb-0310-9956-ffa450edef68
* QPID-4148: HA Not setting initial queues for new RemoteBackupsAlan Conway2012-07-193-14/+21
| | | | | | | | Fix bug introduced by r1362584: "QPID-4144 HA broker deadlocks on broker::QueueRegistry lock and ha::Primary lock" Stopped setting initial queues on new (i.e. not expected) RemoteBackups. git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1363488 13f79535-47bb-0310-9956-ffa450edef68
* QPID-4145: HA Minor fixes to recoveryAlan Conway2012-07-197-27/+44
| | | | | | | | | - Demote timed-out backups from ready to catch-up. - Don't cancel connected backups on timeout, only disconnected ones. - Don't allow promotion of a catch-up broker. - Minor logging improvement. git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1363487 13f79535-47bb-0310-9956-ffa450edef68
* QPID-4144 HA broker deadlocks on broker::QueueRegistry lock and ha::Primary lockAlan Conway2012-07-195-52/+44
| | | | | | | | | | | | Running tests repeatedly, the broker deadlocked with the attached stack trace. The problem call sequences are: 1. QueueRegistry::destroy takes QueuerRegistry lock > ConfigurationObserver::queueDestroy > ha::Primary::queueDestroy takes Primary lock. 2. ConnectionObserver::opened cals Primary::opened lock> RemoteBackup>getQueues().eachQueue This patch breaks the deadlock at both ends: QueueRegistry no longer holds the lock across the observer call and Primary does not hold the lock across eachQueue. git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1363486 13f79535-47bb-0310-9956-ffa450edef68
* QPID-4138: HA sporadic failure in ha_testsAlan Conway2012-07-194-1/+8
| | | | | | | | | This was another test issue: trace logging had been left on by mistake, which was slowing down the tests to the point they were hitting test time-outs. Also in this patch: Include symbolic satus in membership update log message. git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1363485 13f79535-47bb-0310-9956-ffa450edef68
* QPID-4136: HA sporadic failures in ha_testsAlan Conway2012-07-172-2/+6
| | | | | | | Caused by a test bug, default timeout was too short. Fixed the test to set a high timeout and increased default timeout. git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1362654 13f79535-47bb-0310-9956-ffa450edef68
* QPID-4133: (Jason Dillaman) Non-responsive peer in federated link can result ↵Alan Conway2012-07-171-1/+3
| | | | | | | | in entire cluster shutdown Fix is to not start link processing until the connection is open. git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1362653 13f79535-47bb-0310-9956-ffa450edef68
* NO-JIRA: Added missing EXTERN statements causing build breakage.Alan Conway2012-07-171-2/+3
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1362636 13f79535-47bb-0310-9956-ffa450edef68
* QPID-4036 QPID-4192 cluster connection leak on failed auth. merge from trunk.Charles E. Rolke2012-07-171-0/+2
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1362628 13f79535-47bb-0310-9956-ffa450edef68
* QPID-4126: HA primary times out expected backups.Alan Conway2012-07-117-33/+112
| | | | | | | | | | | | After a failure, the newly-promoted primary broker guards messages for each of the backups that were connected at the time of the failure. It waits for them to reconnect to the new primary before becoming active. This patch introduces a time-out so that if an expected backup fails to fail-over within the time limit, the primary will cancel the guards for that backup and carry on. git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1360348 13f79535-47bb-0310-9956-ffa450edef68
* QPID-4128: HA should not use amq.failover for replication linksAlan Conway2012-07-112-2/+3
| | | | | | | | | | | HA replication links should not use the amq.failover exchange to get updates reconnect targets. amq.failover provides the client failover list, HA manages a separate failover list for brokers. Replication links should be using the broker list, and not allow it to be overwritten by amq.failover updates. Review requested for 0.18, 2 line fix. git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1360228 13f79535-47bb-0310-9956-ffa450edef68
* QPID-4118: HA does not work with authentication and authorization.Alan Conway2012-07-116-31/+122
| | | | | | | | | - Updated test framework to use credentials - Updated BrokerReplicator to use HA identity to create configuration - Updated documentation with a HA security section. - Updated qpid-ha to take --sasl-mechanism git-svn-id: https://svn.apache.org/repos/asf/qpid/branches/0.18@1360227 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3757: clean up demux when subscription manager is destroyedGordon Sim2012-07-092-1/+12
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1359099 13f79535-47bb-0310-9956-ffa450edef68
* QPID-4117: honour alternate-exchange option specified within x-declare for a ↵Gordon Sim2012-07-062-2/+24
| | | | | | link git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1358321 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3883: Using application headers in messages causes a very large slowdownAndrew Stitcher2012-07-065-17/+64
| | | | | | | Add subject to outgoing messsage before encoding it to save a round trip decode-encode. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1358275 13f79535-47bb-0310-9956-ffa450edef68
* QPID-3883: Using application headers in messages causes a very large slowdownAndrew Stitcher2012-07-061-81/+224
| | | | | | Encode Variants directly without translating to FieldTables git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1358274 13f79535-47bb-0310-9956-ffa450edef68
* NO-JIRA: Changed VariantImpl to use a union more sensibly and avoid ugly and ↵Andrew Stitcher2012-07-061-39/+33
| | | | | | | | potentially buggy reinterpret_casts. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1358273 13f79535-47bb-0310-9956-ffa450edef68
* NO-JIRA: Fix previous TimerTask fix to workaround stupid windows.h symbol ↵Andrew Stitcher2012-07-062-3/+3
| | | | | | corruption git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1358265 13f79535-47bb-0310-9956-ffa450edef68
* QPID-4085: HA review and fix lock scopes.Alan Conway2012-07-053-45/+53
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1357852 13f79535-47bb-0310-9956-ffa450edef68
* NO-JIRA: Remove dead code.Alan Conway2012-07-056-63/+0
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1357851 13f79535-47bb-0310-9956-ffa450edef68
* QPID-4085: HA failover test: wait for all receivers to be connected.Alan Conway2012-07-051-12/+13
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1357850 13f79535-47bb-0310-9956-ffa450edef68
* QPID-4085: HA add missing lock in HaBroker::getKnownBrokers()Alan Conway2012-07-051-0/+1
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1357849 13f79535-47bb-0310-9956-ffa450edef68
* QPID-4085: HA failover test: client failure: "connection not yet open"Alan Conway2012-07-051-1/+1
| | | | | | | | Sporadic failure of ha_tests.LongTests.test_failover_send_receive with client error. "connection not yet open". Fixed by replacing Exception with TransportFailure so reconnect logic will apply. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1357848 13f79535-47bb-0310-9956-ffa450edef68
* QPID-4085: HA message-loss race condition, handling replication event after ↵Alan Conway2012-07-054-8/+18
| | | | | | | | | | | response. If the backup broker receives a declare event for a queue after receiving a queue response for the same queue, it removes the queue and replaces it with the new one from the reponse. Previously it did not remove the corresponding bridge so things fail when we attempt to create it. Corrected to remove the bridge also. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1357846 13f79535-47bb-0310-9956-ffa450edef68
* NO-JIRA: Fix for potential Timer deadlock issue:Andrew Stitcher2012-07-052-11/+51
| | | | | | | | | | | | | | | | | - Previously we used a mutex to prevent cancelling a TimerTask whilst it was still executing from within its callback in another thread. This violates the principle that you shouldn't hold locks when calling the arbitrary code in a callback, and so is subject to potential deadlock problems. - This fix only works if no timer callback calls TimerTask::cancel(); this is true with the current code. And there is no good reason to call cancel() from within a callback, as cancel is the default behviour in any case - you have to specifically reschedule a recurring timer. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1357827 13f79535-47bb-0310-9956-ffa450edef68
* NO-JIRA: Trivial fix for crash-on-exit bug.Alan Conway2012-07-043-3/+2
| | | | | | | | | Primary::~Primary() no longer resets HaBroker::observer. It is unnecessary since HaBroker unregisters the observer before Primary is destroyed, and it cause sporadic crash on exit because the HaBroker::observer can be reset before ~Primary is called. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1357370 13f79535-47bb-0310-9956-ffa450edef68
* NO-JIRA: Don't run the IPv6 test unless the machine has global IPv6 addressesAndrew Stitcher2012-07-031-0/+13
| | | | | | | - The qpid code deliberately won't use IPv6 unless it has IPv6 configured on the machine, if it's not fully there don't try to test it. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1356946 13f79535-47bb-0310-9956-ffa450edef68
* NO-JIRA: Removed dead code.Alan Conway2012-07-031-8/+0
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1356936 13f79535-47bb-0310-9956-ffa450edef68
* NO-JIRA: Typo in previous fixAndrew Stitcher2012-07-021-1/+1
| | | | | | (r1352646) git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1356280 13f79535-47bb-0310-9956-ffa450edef68
* NO-JIRA: Fixed erroneous use of operator+Andrew Stitcher2012-07-011-1/+4
| | | | | | | operator+(const char*, int) was used where operator+(std::string&, std::string&) was meant git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1355847 13f79535-47bb-0310-9956-ffa450edef68
* QPID-4021: Badly behaved clients can still clog up the brokerAndrew Stitcher2012-06-282-2/+12
| | | | | | | | | | Ameliorate the problem by only turning timeout off after receiving 3 frames from the sender. This avoids an unauthenticed client causing a DoS by just hanging before completing authentication in most cases. This is far from a good fix, but should mostly avoid the issue until it can be fixed in a neat way. git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1355142 13f79535-47bb-0310-9956-ffa450edef68
* NO-JIRA: Add missing EXTERN declarations needed by HA module.Alan Conway2012-06-281-6/+10
| | | | git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1355138 13f79535-47bb-0310-9956-ffa450edef68
View Lorry Controller Status