Back in 8e40b2d4f4f242271d3dfcda4f9b96d3f94cee1b thread safety was added
to TypedProperties by synchronizing all relevant methods. This was
simple and effective but suffers from performance issues in read-heavy
use-cases.
This commit improves the performance by using a read/write locking
mechanism so reads can execute concurrently.
No new tests were added since the original commit added tests to verify
thread safety.
Adds management and metrics to the outgoing AMQP senders created by a
remote peer that has message federation configured from the local broker
including a view for the remote policies that tracks number of sent
messages for all producers. Also adds a top level federation management
view that shows counts for incoming and outgoing messages for all
federation consumer and producers.
Add a more declarative state handling API for initialization, start,
stop and shutdown to federation resources and the broker connection API
to make management resource create and cleanup more reliable and robust
to avoid and leaked management objects when reconnection attempts stop
or broker connections are manually stopped or are stopped by
configuration updates or configuration remove.
This commit enforces a consistent style for parentheses, namely that
there is no padding. Futhermore, it updates all the code that violates
this styling so that the code is styled consistently across the entire
code-base.
This commit enforces a consistent style for commas, namely that a space
should follow a comma. Futhermore, it update all the code that violates
this styling up to date so that the code is styled consistently across
the entire code-base.
Add basic management views for AMQP broker connections and implement control
types for AMQP federation features along with the broker connection management
views. Some initial work also to provide support for other broker connection
features to add management control and also plan for future views of incoming
broker connections and management of AMQP federation resources.
177e6820b541e0a71b952eebf503a4d2235910c5
e34fd09ce325a551bf2eaf579854e2e107435502
e065d25b6fb4cdfa7f5b72afd52d6f4771b50634
af9bd7b84aad32e4fe30f2c8909e51cf7300b475
f4bdacbc4cb46b308f80391940890c74bf111ecd
74951d9221b29336c6575de9390a8557114e10eb
fe0ca4d84fc587b701e4b83ae8e5f2240c222842
b08b91a32faa89cdff4403b532e60d96084f2d4a
85857ab8eb1b1eca2498306239740e0e39d2acd4
e02ec69021f24220c0b91b8324382c9630415276
bf9f6d213bf6b3bf4fd7e15430ada8c9954e69ce
2d806807b8d9006cf642392ad5da9878394cebae
52a4157bbb898ee8fd8d7f13fffca71a9c368630
53bb3ea1833cddda6c2ded70d3149286c8207570
d5f152c7254c378707b12ad4a0b798fdb8cb7b10
b3529dcea428fa697aacbceacc6641e47cfb74ba
246bf083914c7acbb05a7fe0904c471331242c39
8a04ee07de8a7c863daf37a07a1176131151caa0
6cb68f2ce923ea3c2400c21967b2003175c8a0f4
9873fccf744c0cb0a25dd905fab67ea52ef7aa7d
See PR for individual commits which were squashed to form this change. This closes#5320
The broker uses unique IDs for logging statements. As logging changes
over time we need a way to "retire" these IDs (e.g. when certain
logging statements are no longer needed) to ensure they are not used
This commit does the following:
- Removes all the logging methods which are no longer used and
"retires" the corresponding ID (i.e. adds them to the `retiredIDs`
list for that LogBundle).
- Updates the validation for IDs that have been retired or are in
active use including a new suggestion about a valid ID to use when an
invalid ID is found.
- Fixes all the regular expressions in all the various uses of
`@LogBundle` to ensure there are no overlaps to prevent duplicates
between bundles.
Changes from myself and Robbie Gemmell (see PR). This closes#5303.
Throughout the years, the standard mechanism for storing passwords has evolved.
In the beginning, passwords were stored in plaintext. Developers are now
encouraged to leverage adaptive one-way functions to store a password. Using a
two-way function by default for storing passwords without a warning could lead
users to a false sense of security.
PriorityLinkedList has multiple sub-lists, before this commit PriorityLinkedList::setNodeStore would set the same node store between all the lists.
When a removeWithID was called for an item on list[0] the remove from list[4] would always succeed first. This operation would work correctly most of the time except
when tail and head is being used. Many NullPointerExceptions would be seen while iterating on the list for remove operations, and the navigation would be completely broken.
A test was added to PriorityLinkedListTest to make sure the correct lists were used however I was not able to reproduce the NPE condition in that test.
AccumulatedInPageSoakTest reproduced the exact condition for the NPE when significant load is used.
We've traditionally used org.apache.activemq.artemis.utils.Base64 for
Base64 encoding/decoding. This implementation is based on public domain
code from http://iharder.net/base64.
In Java 8 java.util.Base64 was introduced. I assumed we hadn't switched
to this implementation for performance reasons so I created a simple
JMH-based test to compare the two implementations and it appears to me
that java.util.Base64 is significantly faster than our current
implementation. Using the JDK's class will simplify our code and
improve performance. Also, it should be 100% backwards compatible
since Base64 encoding/decoding is standardized.
The test in this commit was distilled down from a much more complex
integration test that rarely reproduced the problem. It is short and
sweet and reproduces the problem every time.
The problem exists in the iterator's `remove()` method where it uses
`index` instead of `i` when calculating a new highest priority.
Updates artemis-log-annotation-processor to use artemis-project so that
artemis-pom can reference artemis-log-annotation-processor without cycle.
Split out its tests to their own module to faciltate, also exercising the
profile mechanism to enable the processor usage with trigger file.
Simplify disabling processing in the module using maven.compiler.proc prop
available since maven-compiler-plugin 3.13.0
Uses a dummy non-processor path at root to 'disable' processsing on JDK < 23,
accounting for Maven 3 not being able to unset maven.compiler.proc from a
parent, and JDKs < 21 requiring newest builds to support -proc:full value
needed otherwise to reenable processing once explicitly disabled.
This commit uses lambdas or method references wherever possible. There
are still a handful of places that appear like they could be changed but
couldn't mainly because they use "this" and the meaning of "this"
changes when using a lambda.
