git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1751566 13f79535-47bb-0310-9956-ffa450edef68

Fix previous commit that did not respect the plugin boundaries for
messaging::shutdown.

git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1737887 13f79535-47bb-0310-9956-ffa450edef68

The leak was caused by the sys::Poller. When two pollers run in the same process
and one is idles, PollerHandles accumulate on the idle poller threads and are
never released. The HA broker uses the qpid::messaging API to get initial status from
other brokers. The messaging API creates a separate Poller from the broker.

The fix is to shut down the qpid::messaging poller as soon as initial status
checks are complete so it does not interfere with the broker's Poller.

git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1737852 13f79535-47bb-0310-9956-ffa450edef68

git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1717670 13f79535-47bb-0310-9956-ffa450edef68

git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1701130 13f79535-47bb-0310-9956-ffa450edef68

git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1691457 13f79535-47bb-0310-9956-ffa450edef68

* 0-10 path no longer hans on open when connecting to a broker not
  supporting that version
* the 'protocol'  connection option now supports specifying multiple
  protocols to try in order (as a coma separated list)
* the protocol defaults, i.e. the value assumed if the 'protocol'
  connection option is not specified, can now be set via the client
  config file (e.g. protocol-defaults=amqp1.0,amqp0-10) or an
  environment variable (e.g QPID_PROTOCOL_DEFAULTS=amqp1.0,amqp0-10)
* if neither the connection option nor the defaults are specified
  all valid versions will be tried (currently amqp0-10, then amqp1.0
  but this may change in a future version)

git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1686078 13f79535-47bb-0310-9956-ffa450edef68

Implements the "transactional retire and settle immediately" option for
transactions as specified in AMQP 1.0 in the qpid::messaging C++ client.

NOTE: Transactions over AMQP 1.0 require proton 0.9 or greater.  With older
versions, attempting a transactions over AMQP 1.0 will raise a link-detached
exception "Node not found: tx-transaction"

1. Added descriptor list to Variant with support in Encoder and PnData.

Required to support transactions, need to be able to create described lists.
Variant changes are source and binary compatible.

A Variant now has a Variant::List of descripors which can be numeric or string.
Nested descriptors are implemented by putting multiple descriptors in the list.

Other minor changes:
- Variant refactor: don't delete impl on every assignment.
- Add Variant constructors that take a string encoding.
  (new constructors, not defaulted arguments, so the change is binary and source compatible.)
- Growable buffer support for Encoder.
- Printing described Variant prints descriptors in form @descriptor value

2. Added transaction support to AMQP 1.0 client code

Added messaging/amqp/Transaction.h,cpp: transaction logic
- communicate with coordinator, send declare/dischange messages.
- add tx state info to transfers and acknowledgements.
- Sync session after discharge.
- A transactional session automatically acks any message retrieved by fetch/get
  to bring them into the transaction. This is consistent the 0-10 client.

Minor fixes to existing client code:
- Fix use of pn_drain API in C++ client to work with C++ and Java brokers.
- Make amqp::Exception derive from qpid::Exception

3. Fixes to existing broker code:

- Incoming.cpp fix: start async completion before processing message.
- Delay accept of dischage message till commit is complete.
- newSession - handle failover during session creation.

4. Added tests

interop_tests.py: transaction tests that can run against an external broker, see comments.

ha_tests.py: Enable transaction tests over AMQP 1.0.

Minor test fixes:
- brokertest.py don't set default logging if QPID_LOG env vars set.
- brokertest.py Pass kwargs to broker() create function.
- qpid-receive: capacity should never be larger than message count.
- Accept user:pass as well as user/pass in Url.
- brokertest.py: Always do a ready() check on all brokers.

If proton < 0.9 is used, transaction tests will be skipped or will downgrade to
the amqp0-10 protocol with a printed warning.

git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1662743 13f79535-47bb-0310-9956-ffa450edef68

Problem was a missing Py_INCREF in cpp/include/qpid/swig_python_typemaps.i

git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1626887 13f79535-47bb-0310-9956-ffa450edef68

This is partly not-a-bug, there is a client error handling issue that has been
corrected.

qpid-txtest2 initializes a queue with messages at the start and drains the
queues at the end. These operations are *not transactional*. Therefore
duplicates are expected if there is a failover during initialization or
draining. When duplicates were observed, there was indeed a failover at one of
these times.

Making these operations transactional is not enough to pass, now we see the test
fail with "no messages to fetch". This is explained as follows:

If there is a failover during a transaction, TransactionAborted is raised. The
client assumes the transaction was rolled back and re-plays it. However, if the
failover occurs at a critical point *after* the client has sent commit
but *before* it has received a response, then the the client *does not know*
whether the transaction was committed or rolled-back on the new primary.

Re-playing in this case can duplicate the transaction. Each transaction moves
messages from one queue to another so as long as transactions are atomic the
total number of messages will not change. However, if transactions are
duplicated, a transactional session may try to move more messages than exist on
the queue, hence "no messages to fetch". For example if thread 1 moves N
messages from q1 to q2, and thread 2 tries to move N+M messages back, then
thread 2 will fail.

This problem has been corrected as follows: C++ and python clients now raise the
following exceptions:

- TransactionAborted: The transaction has definitely been rolled back due to a
  connection failure before commit or a broker error (e.g. a store error) during commit.
  It can safely be replayed.

- TransactionUnknown: The transaction outcome is unknown because the connection
  failed at the critical time. There's no simple automatic way to know what
  happened without examining the state of the broker queues.

Unfortunately With this fix qpid-txtest2 is no longer useful test for TX
failover because it regularly raises TransactionUnknown and there's not much we
can do with that.

A better test of TX atomicity with failover is to run a pair of
qpid-send/qpid-receive with fail-over and verify that the number of
enqueues/dequeues and message depth are a multiple of the transaction size. See
the JIRA for such a test. (Note these test also sometimes raise
TransactionUnknown but it doesn't matter since all we are checking is that
messages go on and off the queues in multiple of the TX size.)  )

Note: the original bug also reported seeing missing messages from
qpid-txtest2. I don't have a good explanation for that but since the
qpid-send/receive test shows that transactions are atomic I am going to let that
go for now.

git-svn-id: https://svn.apache.org/repos/asf/qpid/trunk@1621211 13f79535-47bb-0310-9956-ffa450edef68