Hi Folks,

A few issues were reported with OpenMAMA RC2 which have now been resolved (thanks to all involved in testing so far).

Please find RC3 now available complete with artefacts: https://github.com/finos/OpenMAMA/releases/tag/OpenMAMA-6.3.2-rc3

Let’s aim for a 2 week test window so please get testing.

This would close this RC window on 16th March, by which point if no major issues are reported, we will be getting ready to release 6.3.2 final.

Cheers,

Frank

Hi Folks,

A few issues were reported with OpenMAMA RC1 which have now been resolved (thanks to all involved in testing so far).

Please find RC2 now available complete with artefacts: https://github.com/OpenMAMA/OpenMAMA/releases/tag/OpenMAMA-6.3.2-rc2

Let’s aim for a 2 week test window so please get testing.

This would close this RC window on 6th February, by which point if no major issues are reported, we will be getting ready to release 6.3.2 final.

Cheers,

Frank

Hi Folks,

We plan on taking a cut for OpenMAMA 6.3.2 RC1 around the end of this week (16^th December cut-off). If anyone has anything they want to submit before then please let me know so I can delay the cut-off if necessary, otherwise we will continue as planned.

There are some key changes included in this release including:

• MAMA Resource Pool support added for C, C++, C# and Java
• Common headers moved to wombat subdirectory (see https://lists.openmama.org/g/Openmama-dev/topic/openmama_top_level_common/82750583 for more details on this and how to work around if this is an issue for your projects)
• C++11 now adopted as standard for project
• Lots of enhancements to CI including static analysis and security scanning
• Project is now compliant and certified with Open Source Security Foundation (OpenSSF) best practices
• Updated payload registered character to support unprintable characters
• Unit tests now available for MAMDA and C#
• Several other maintenance updates and bugfixes: https://github.com/finos/OpenMAMA/pulls?q=is%3Apr+milestone%3AOpenMAMA-6.3.2

This release also includes new changes to support:

• CentOS Stream 8
• CentOS Stream 9
• RHEL 8
• RHEL 9
• Ubuntu 22.04 LTS
• Visual Studio 2022

It also drops support for:

• Fedora (we now favour CentOS stream for getting ahead of upstream RH changes)
• CentOS 6 (EOL November 2020)
• CentOS 8 (EOL December 2021)
• Ubuntu 16.04 (EOL April 2021)

Cheers,

Frank

Hi Folks,

The following is to make testing the most recent changes easier particularly on windows… I have my own fork of vcpkg which points to the latest mama resource pool code.

Run this git command to grab a cut of my fork of vcpkg pointing to the latest openmama resource pool changes from my working branch

    git clone -b openmama-resource-pool-cut https://github.com/fquinner/vcpkg.git vcpkg-openmama

Then set up vcpkg:

    cd vcpkg-openmama

    .\bootstrap-vcpkg.bat

And install OpenMAMA into this environment:

    vcpkg install openmama:x64-windows

Then use Visual Studio / other IDE to open each tutorial from https://github.com/finos/OpenMAMA/tree/next/tutorials/cpp as a cmake project. In your cmake configuration, then set CMAKE_TOOLCHAIN_FILE=\path\to\vcpkg-openmama\scripts\buildsystems\vcpkg.cmake to let your development environment pick up openmama via Cmake’s find_package with no further modifications / environment variables.

Cheers,

Frank

Frank Quinn

Cascadium

T: +44 (0) 28 8678 8015

E: fquinn@...

W: https://cascadium.io

Hi Folks,

Bumping this up to the mailing list again for review.

Again, any feedback let me know either via email / issue or Gitter.

Cheers,

Frank

Frank Quinn, Cascadium | +44 (0) 28 8678 8015 | https://cascadium.io

Hi Folks,

FYI there’s a bunch of updated documentation on the website including a fully integrated / amalgamated version of the developer guides as well as doxygen documentation which is no longer stuffed into a little iframe.

It’s now live at:

https://openmama.finos.org/openmama_developer_guide.html

And:

https://openmama.finos.org/openmama_api_reference_docs.html

Note the developer guide is still to be subject to a content review – this is more of a port / fix of some of things which were “already there” rather than an overall content refresh.

More content will be available too in the coming weeks / months.

Enjoy!

Cheers,

Frank

Frank Quinn

Cascadium

T: +44 (0) 28 8678 8015

E: fquinn@...

W: https://cascadium.io

Hi Folks,

Now that CentOS has been EOL for 16 months, could I ask for feedback again on the below please? 1, 2 or 3?

Cheers,

Frank

Thanks your consideration folks, this RFC is now considered approved and work will begin in the coming weeks.

Cheers,

Frank

Hi Folks,

I have just opened a new RFC to cover the OpenMAMA Resource Pool originally discussed on the mailing list to try and make the OpenMAMA API more accessible for new users without sacrificing performance or functionality:

https://openmama.finos.org/openmama_rfc_mama_resource_pool.html

As per our RFC process, we will leave this open for a period of 2 weeks, after which point if there are no major concerns, the RFC design will be considered approved, and work can begin on the recommended option in the document.

For details on how to provide constructive RFC feedback, please refer to https://openmama.github.io/openmama_rfc_process.html#how-to-provide-rfc-feedback

I’ll be on Gitter for any quick informal questions anyone may have throughout as well.

Cheers,

Frank

Frank Quinn

Cascadium

T: +44 (0) 28 8678 8015

E: fquinn@...

W: https://cascadium.io

Hi Mike,

Subscription activations may be created upfront before mama_start if not using the throttle yes.

The behaviour you have identified is similar to behaviour outlined in the OpenMAMA Developer’s guide in the MAMA Subscription Callbacks section where it is possible to receive an onMsg before the onCreate has even been fired. It’s the same sort of principle in both cases where the thread performing the setup differs from the thread performing the dispatch. Note I expect the behaviour you have observed will only happen if you’re (quite rightly) not using the default queue for your subscriptions.

If you want to prevent this behaviour, adding your own MAMA Queue event to initialized all unthrottled subscriptions when the default queue unblocks could be an option.

Cheers,

Frank

Hi Folks,

It has been a while since the last release but we’re pleased to announce the release of OpenMAMA OMNM 1.0.0. This release includes:

• Complete interface compatibility with OpenMAMA
• Complete unit test coverage for all official OpenMAMA unit tests
• Platform support for all OpenMAMA supported platforms including Windows and OSX
• A set of methods to support extending omnm for use in other payload bridges with reduced code required:
• omnmmsgPayloadImpl_[gs]etExtenderClosure to allow the extender
  to attach their own internal object to the omnm payload
• omnmmsgPayloadImpl_updateVectorMsgPayload to allow the extender
  to update vectors of messages without needing an intermediate
  mamaMsg for each submsg.

Release artifacts can be found here:

https://github.com/cascadium/OpenMAMA-omnm/releases

Or by running:

yum / apt upgrade openmama-omnm

From the OpenMAMA thirdparty repository hosted by our friends at cloudsmith:

https://cloudsmith.io/~openmama/repos/openmama-thirdparty/packages/

The first implementation based on this base bridge will be released soon too and adds support for msgpack.

Cheers,

Frank

Frank Quinn

Cascadium

T: +44 (0) 28 8678 8015

E: fquinn@...

W: http://cascadium.io

Hi OpenMAMA Dev,

I wanted to clarify the expected behavior surrounding creation of subscriptions during application startup – should subscription activations always be executed by the thread dispatching events off the throttle queue?

Taking mamalistenc as an example, subscriptions can be created before we start dispatching on the default event queue. If the throttle rate is non-zero, the subscriptions’ activations are added to the throttle queue after the set interval has expired, but will not be executed until we start dispatching on the queue when we call mama_start.

The other behavior we see is if the throttle is disabled (the rate is set to <=0). The subscriptions’ activations are executed immediately on the calling thread – ignoring the queue - before the mama_start call due to the below code snippet. This results in the subscription being created on the transport and messages being received from the middleware before the mama_start call. Is this expected/allowed behavior?

Throttle.c:

if (self->mRate <= 0.0)

{

    /* throttle is not in use, execute immediately */

    mama_log (MAMA_LOG_LEVEL_FINEST, "wombatThrottle_dispatch (): "

              "no throttle; triggering action.");

    actionCb (closure1, closure2);

}

Thanks in advance,

Mike

Hi Folks,

With more and more packaging options being made available to OpenMAMA, it was only a matter of time before the fact that our build system includes top level header files was to become an issue.

This was raised by our friends at Microsoft / vcpkg when they discovered that it conflicted with one of their headers.

I have suggested a workaround https://github.com/Cheney-W/vcpkg/pull/1 which was ultimately merged upstream, but it involves moving these trailing headers into the “wombat” subdirectory. So that’s:

• destroyhandle.h
• platform.h
• list.h
• lookup2.h
• property.h
• timers.h
• wlock.h

Which will all shortly be moved into the wombat subdirectory through our build system when installing.

If you make direct use of these headers, to fix this in your code if you use these headers directly, you may simply add $MAMA_ROOT/include/wombat to your include paths (or better still, update your #include statements to use <wombat/>). The former is something you should be able to do now before the change lands so you are forwards compatible with it.

This is planned to be included in the next upcoming OpenMAMA release (currently 6.3.2).

If anyone desperately wants a configuration option to allow the legacy header location, that can also be considered, though since we’re trying to get included in as many upstream packing targets as possible, it will not be the default for our packaging going forward.

Cheers,

Frank

Frank Quinn

Cascadium

T: +44 (0) 28 8678 8015

E: fquinn@...

W: http://cascadium.io

Hi Folks,

We are currently considering options for trying to make the OpenMAMA API easier to work with and more accessible for new users.

When I look at the OpenMAMA API's strengths and weaknesses, the setup and management of the subscriptions themselves is one of the more difficult things to manage especially with a new user. These are things which we seasoned users would take for granted but I think new users would struggle with.

That is why I'm suggesting a new "MamaResourcePool" api.

The idea is to manage OpenMAMA resources via OpenMAMA itself rather than depending on the user to manage complex state, setup, shutdown etc. I'm taking the approach of starting with what I (pretending to be a new OpenMAMA user) would like to see as opposed to things which would be technically easier to do etc.

So taking for example a new subscription, this is the entire code I'm proposing (in C here but should translate easily to other languages) should be required:

// Set up the event handlers for OpenMAMA

mamaMsgCallbacks callbacks;

memset(&callbacks, 0, sizeof(callbacks));

callbacks.onMsg = subscriptionOnMsg;

callbacks.onCreate = subscriptionOnCreate;

callbacks.onError = subscriptionOnError;

// Use a new "resource pool"

mamaResourcePool resourcePool;

const char* poolName = NULL;

const void* closure = NULL;

// Create the resource pool, allocating queues, defaults etc from configuration

mamaResourcePool_create(&resourcePool, poolName);

// Actually... subscribe to something via something convenient like a URI

mamaResourcePool_createSubscriptionFromUri(&subscription, "bridge://transport/source/topic", &callbacks, closure);

// Block

mama_start();

And then you would start receiving events. No transport creation, source creation, queue creation, encouraging users use default queue etc.

I’d also suggest that the callbacks onCreate and onError should be optional, with the default implementation simply logging those events.

I foresee similar with lambdas etc. The idea is that mamaResourcePool will:

·         When it is created:

o    Create a number of event queues for subscriptions (which could be configurable in properties based on the poolName)

o    Load all payload and middleware bridges found on the PATH / LD_LIBRARY_PATH etc

o    Call mama_open() (if not already open)

o    Create and manage internal stores of all managed subscriptions, transports, sources etc

·         Then when a resource is created inside:

o    Find / create transport bridge

o    Find / create transport

o    Find / create source

o    Find / create subscription (e.g. multiple callbacks for same subscription)

o    Find / create anything which is required to acquire the requested resource

·         Be convenient and usable, but still flexible (it would return existing OpenMama objects so they can be manipulated)

·         Take the heavy lifting and thread safety pain away from the application developer

·         Defer default "is this basic sub" or "is this market data sub" to the bridge, configuration or as an API call in the resource pool itself

My guess is that many of you may already have an API that does this in each of the provided languages… but newcomers won't have that. This would vastly reduce the complexity and error checking required for a new user, as well as remove the expectation around deciding “which subscription API to use” from the application developer.

If there is any feedback to this approach, or in particular if anyone has recently worked with a new user coming online with OpenMAMA please let us know or pop into Gitter to discuss!

Cheers,

Frank

Frank Quinn

Cascadium

T: +44 (0) 28 8678 8015

E: fquinn@...

W: http://cascadium.io

That’s right Mac users – OpenMAMA has been approved and is now included in homebrew’s set of packages.

So to install OpenMAMA on a Mac, simply

brew update

brew install openmama

…and yes this will also work on the new Apple M1 machines.

This also has been verified as working with linuxbrew (if anyone uses that).

Any issues please raise an issue on the OpenMAMA github page!

Cheers,

Frank

Frank Quinn

Cascadium

T: +44 (0) 28 8678 8015

E: fquinn@...

W: http://cascadium.io