Add a shared-memory message broker. by ehpor · Pull Request #272 · spacetelescope/catkit2

ehpor · 2024-12-17T01:07:04Z

This PR adds a MessageBroker class for distributing arbitrary-sized messages. Messages are put in shared memory, either CPU or GPU shared memory (the latter not implemented), and can be distributed between processes.

A Message contains:

A topic. We use a string, e.g. boston_dm/correction_howfs/value or boston_dm/correction_howfs/set, which follows a hierarchical naming convention.
Metadata. We chose to use a limited number of key-value pairs. The keys are stored statically for each message, the values are stored for each message.
Data. This is an array-like structure. The shape and dtype of this array does not have to be the same from message to message.

The MessageBroker allows us to listen to new messages at any level in the hierarchical structure. This means that we can listen to either a single message topic or many message topics at the same time. The MessageBroker also allows us to send messages with any topic.

Relies on #289.

ehpor · 2025-03-19T01:58:39Z

Performed the first benchmarks:

4.09438e+06 messages per second
244 ns per message

This is single-threaded. The biggest bottlenecks are:

~40%: the lookup of the topic header and event for a topic.
~25%: the Event structure providing notification of other processes (MacOS specific).
~15%: the allocators (combined freelist and pool allocator).
~15%: getting a timestamp.
~5%: misc processing.

(benchmark using samply)

For the topic header and event lookup for a topic: this can in principle be condensed into a single hash map/calculation. But I'll postpone this for later. This is enough for now. All low-hanging fruit has been taken care of.

ehpor · 2025-03-19T06:34:21Z

When walking home, I realized that the subtopics are very predictable. Doing a full lookup every subtopic might be avoidable with clever ordering. This is also work for the future, if required.

ehpor · 2025-03-20T01:10:46Z

I'm having some problems with the Python bindings. The polymorphism for Memory, SharedMemory and LocalMemory is not working as expected. I don't really understand why not. Pybind11 knows about the inheritance structure and should upcast the shared_ptr to a shared_ptr just fine.

ehpor · 2025-04-19T00:54:10Z

I've moved to a cheaper memory allocator, based on a thread-local pool fed by a buddy allocator. This is the final iteration of the allocator. I've also moved to a single Event structure for signaling.

New benchmarks:

5.54217e+06 messages per second
180 ns per message

Distributed as follows:

10% generating UUIDs
23% allocating and deallocating
21% getting topic header
7% getting a timestamp
4% looping through subtopics
4% signaling the event
3% updating the frame counter per topic
9% updating the framerates
19% misc other stuff

…rs seeded by true randomness.

This has been replaced by Event.

ivalaginja

I am slowly getting accustomed to this new concept but it's taking some time. I have a first round of questions that I was hoping to get answers to before being able to wrap this up on my side.

ivalaginja · 2025-05-06T12:46:29Z

+    data = b'hello world'
+    arr = np.frombuffer(data, dtype='uint8')
+
+    broker.publish_data(topic, data)


How is this different from publish_message()? Also, what is data? Is it a synonym for payload? Trying to relate things to the figures in #272 (comment).

@ehpor Please correct me if I am wrong. (and yes, I did dive a bit into the C++ code to see what's going on)

The difference is that publish_message already has a message object ready to be published to the broker, but that is not the case here. In this example, we only have topic and data, without any message object being created yet.

Without publishing a message, the broker has no access to any information.
So publish_data() needs to do several things:

Assign a trace_id, which is needed to trace our message.

Prepare a message, meaning it creates a message object (Header and Payload, see the message content fig. ) ready for publication.

Set the data information for the message, so it is complete and ready to be published.

Publish the message, making it available to the broker.

Hence, assert (retrieved_message.payload == arr).all() should pass. payload stores the data of the message.

@ivalaginja Data is not a synonym for payload. A payload is part of a message. Data is not part of a message. But a payload is data.

The reason for this function is similar to the Datastream.submit_data() vs. Datastream.request_new_frame() & Datastream.submit_frame(). The first is a convenience function for the latter two. Maybe my naming can be improved?

@alexisyslau Exactly. Unless you're hyperoptimizing, most people are not using the Datastream.request_new_frame() and Datastream.submit_frame() separately. Using those allows, in some cases, to remove a copy of the image. But most people will be using the Datastream.submit_data(), which combines the two, and the memory copy in between.

The same thing is true here. I'm expecting very little use of MessageBroker.prepare_message() & MessageBroker.publish_message() in favor of MessageBroker.publish_data() which combines the two and the memory copy in between. However, similar to the case of Datastreams, I want it to be available when people need it.

I see! Thanks to both of you! I think the naming is fine, especially when you trace it back to the data streams. Makes sense!

ivalaginja · 2025-05-06T12:50:32Z

+    broker.publish_data(topic, data * 2)
+    message_2 = broker.get_newest_message(topic)
+
+    assert str(message_2.trace_id) != str(message_1.trace_id)


What are the differences between trace_id, message_id and Partial frame ID?

A trace ID is set for each new information. It is typically assigned to the original data that came in, ie. the source of the "information". Eg. a camera service would create new trace IDs for each camera image. But a WFS reconstructor service would retain the same trace ID of the image that it computed the reconstruction for. Therefore, by using the trace ID, we can trace the flow of that camera information through the entire system. The technical term for this is "distributed tracing".

A payload ID is unique per data. In case partial messages are submitted (ie. pixel streaming), the payload for these messages will be the same, so they retain the same payload ID. A payload ID identifies a unique array.

A message/frame ID is sequential numbering per topic. The trace ID and payload ID are not. It allows us to figure out whether we missed frames, and to request the previous/next message for a particular message.

A partial frame ID is a numbering within a message. It increments sequentially, starting from zero. It gets incremented each time a partial message is submitted. I didn't test the partial message generation in this PR. It probably needs more work.

ivalaginja · 2025-05-06T12:57:35Z

+    waiting.wait(1)
+    assert waiting.is_set()


If this times out without a signal, shouldn't waiting.is_set() return False?

Yes, this is checking that the wait didn't time out. I could also do:

assert waiting.wait(1)

I'll add comments.

Added more comments to the test.

Oh I had inverted the purpose of the test in my head. Makes sense! Thanks!

alexisyslau · 2025-05-06T15:12:20Z

+	ArrayInfo &array_info = message.m_Header->payload_info.array_info;
+
+	array_info.data_type = 'u';
+	array_info.byte_order = '=';


Most of the CPUs and GPUs should be little-endian so using the native order should be ok normally. But is it better to force it to be little-endian <? This will make sure we always interpret the array in the same way.

Since the data type here is a byte (u), which is one byte long, the byte order doesn't matter. But, I'm not sure why I chose = compared to @, which are native byte order and native alignment, and native byte order and no alignment respectively. Again, no difference in this case, but native everything should be the default in other parts of the code, e.g. when submitting an array.

ivalaginja

I left a couple more questions about some details, but over all things make a lot more sense now. I also still need to run these tests.

ivalaginja · 2025-05-07T07:20:33Z

+def test_message_subscription(broker):
+    topic = "test_message_subscription"
+
+    subscription_newest = broker.subscribe(topic, mode=MessageSubscriptionMode.NewestOnly)


New question: why do we need to subscribe to the broker, and then work through the subscription? Intuitively, I would have asked the broker directly for a next message, next message ID etc. Is it only to be able to define the subscription mode or is there more behind it?

One big problem currently with datastreams is that the next frame ID is linked to your opened datastream. Furthermore, service proxies cache the datastream object, since closing and reopening is an expensive operation and so we are minimizing that. This means that if you do stream.get_next_frame() on the same stream, given to you on the same service, the threads will compete for the next frame. Ie. the threads are bound to the same subscription.

By splitting the subscription up from the message broker / datastream, you avoid this problem. The next frame id to read is stored in the subscription, where it's supposed to be, rather than in the message broker / datastream. This was an oversight when developing catkit2 datastreams, and I'm kinda surprised that we haven't run into bugs of this kind so far.

ivalaginja · 2025-05-07T07:20:36Z

+    assert message_2.payload[0] == 10
+    assert message_3.payload[0] == 11
+
+    assert subscription_newest.try_get_next_message() is None


So this means once I get a message through a subscription, that message does not exist anymore? Would make sense, just double-checking - and trying to find the correct vocabulary for this ^^

Yep, a single subscription can only access a certain message once. The try_get_next_frame() function is there for datastreams as well, but hidden behind a timeout=0, which reads badly.

So, there is get_next_message() which gets the next message in your subscription, according to your MessageSubscriptionMode, which will wait for that message if it's not there yet or return immediately if it is. And we have try_get_next_message() which doesn't wait and will return None if it's not available yet.

ivalaginja · 2025-05-07T07:21:17Z

+    assert subscription_sequential.try_get_next_message() is None
+
+    subscription_newest2 = broker.subscribe(topic, starting_frame_id=1, mode=MessageSubscriptionMode.NewestOnly)
+    assert subscription_newest2.next_message_id == 2


Why is this 2 if you give it a starting ID of 2? Does it not give you the same one? The sequential subscription below shows the behavior I expect, so why does the "newest" subscription mode automatically up the message id?

MessageSubscriptionMode.Sequential will try to give you messages without dropped frames, even if there are newer messages available. MessageSubscriptionMode.NewestOnly will instead drop frames to give you the most recent update. In this case, subscription_newest2 is trying to read frame 1, but sees that frame 2 is already there, so it will skip frame 1 and give you frame 2. subscription_sequential2 will not skip frames, and so will give you frame 1 first, then frame 2 afterwards (untested here, because that was already tested a few asserts above).

(Potentially stupid) follow-up question: How come there is frame 2 already there?

We submitted three frames to the message broker: frame 0, 1 and 2.

Or I'm misunderstanding the question.

Submitted where? I only see them submitted at the very top, and we access them all later on, so they shouldn't be there anymore, no? Or does a new subscription have access to all the messages?

Messages are not consumed. You can access historical data as well. The last N messages are kept for each topic (currently N = 32), similar to how the last N frames are kept for datastreams.

Sorry, I should have said that earlier. This is different than Apacha Kafka or Aeron, if you're familiar with those. In those frameworks, messages are only delivered once, and only accessible if published after you created your subscription. In our message broker, messages are delivered once and then can be accessed later, kinda like a time series database. I'm not fully convinced about the latter capability, since I don't think we ever used it on HiCAT. And it complicates the networking interface that I'm building now.

Ah gotcha!! I don't know those in particular, but that's the type of systems I was in touch with previously. Makes sense, thanks a lot for the effort you put into all these explanations!

ivalaginja · 2025-05-07T07:30:23Z

@ehpor I am unsure what I am doing wrong. I am trying to reinstall catkit2 so that I can test the tests, but the installation fails with the below error message. I am slowly acquiring PTSD from Cmake errors 😬

(catkit2) maintenances-MacBook-Pro-6:catkit2 ilaginja$ pip install -e .
Obtaining file:///Users/ilaginja/repos/catkit2
  Installing build dependencies ... done
  Checking if build backend supports build_editable ... done
  Getting requirements to build editable ... done
  Installing backend dependencies ... done
  Preparing editable metadata (pyproject.toml) ... done
Building wheels for collected packages: catkit2
  Building editable for catkit2 (pyproject.toml) ... error
  error: subprocess-exited-with-error
  
  × Building editable for catkit2 (pyproject.toml) did not run successfully.
  │ exit code: 1
  ╰─> [41 lines of output]
      2025-05-07 09:28:20,644 - scikit_build_core - INFO - RUN: /Users/ilaginja/opt/miniconda3/envs/catkit2/bin/cmake -E capabilities
      2025-05-07 09:28:20,674 - scikit_build_core - INFO - CMake version: 3.29.3
      *** scikit-build-core 0.11.2 using CMake 3.29.3 (editable)
      2025-05-07 09:28:20,724 - scikit_build_core - INFO - Build directory: /private/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmptmubhn3l/build
      *** Configuring CMake...
      2025-05-07 09:28:20,765 - scikit_build_core - INFO - RUN: ninja --version
      2025-05-07 09:28:21,413 - scikit_build_core - INFO - Ninja version: 1.11.1
      2025-05-07 09:28:21,415 - scikit_build_core - INFO - RUN: /Users/ilaginja/opt/miniconda3/envs/catkit2/bin/cmake -S. -B/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmptmubhn3l/build -DCMAKE_BUILD_TYPE:STRING=Release -C/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmptmubhn3l/build/CMakeInit.txt -DCMAKE_INSTALL_PREFIX=/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmptmubhn3l/wheel/platlib -DCMAKE_MAKE_PROGRAM=ninja
      loading initial cache file /var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmptmubhn3l/build/CMakeInit.txt
      -- The C compiler identification is AppleClang 13.0.0.13000027
      -- The CXX compiler identification is AppleClang 13.0.0.13000027
      -- Detecting C compiler ABI info
      -- Detecting C compiler ABI info - done
      -- Check for working C compiler: /Library/Developer/CommandLineTools/usr/bin/clang - skipped
      -- Detecting C compile features
      -- Detecting C compile features - done
      -- Detecting CXX compiler ABI info
      -- Detecting CXX compiler ABI info - done
      -- Check for working CXX compiler: /Library/Developer/CommandLineTools/usr/bin/clang++ - skipped
      -- Detecting CXX compile features
      -- Detecting CXX compile features - done
      CMake Error at catkit_core/CMakeLists.txt:41 (find_package):
        By not providing "Findcppzmq.cmake" in CMAKE_MODULE_PATH this project has
        asked CMake to find a package configuration file provided by "cppzmq", but
        CMake did not find one.
      
        Could not find a package configuration file provided by "cppzmq" with any
        of the following names:
      
          cppzmqConfig.cmake
          cppzmq-config.cmake
      
        Add the installation prefix of "cppzmq" to CMAKE_PREFIX_PATH or set
        "cppzmq_DIR" to a directory containing one of the above files.  If "cppzmq"
        provides a separate development package or SDK, be sure it has been
        installed.
      
      
      -- Configuring incomplete, errors occurred!
      
      *** CMake configuration failed
      [end of output]
  
  note: This error originates from a subprocess, and is likely not a problem with pip.
  ERROR: Failed building editable for catkit2
Failed to build catkit2
ERROR: Could not build wheels for catkit2, which is required to install pyproject.toml-based projects
(catkit2) maintenances-MacBook-Pro-6:catkit2 ilaginja$

ehpor · 2025-05-07T07:35:23Z

For me, this has always been a problem of not activating the environment, or, worse, conda telling me the environment is activated while it is not. After deactivating the environment, then reactivating it again, it works for me. Another check is to see if cppzmq is installed conda list cppzmq. It should be.

FYI: I've been becoming so fed up with this, that there is #312 which removes cppzmq as dependency (not the root cause btw), and #276 (for doing a correct build of the library).

ivalaginja · 2025-05-07T07:38:33Z

... it actually seems like I don't have cppmq installed? No output when I run conda list cppzmq, or conda list |grep cpp (tried the latter just in case I mistyped).

ivalaginja · 2025-05-07T07:41:08Z

I'll just install it manually. God knows on what outdated env file I created this environment.

ivalaginja · 2025-05-07T07:43:18Z

Uuughh. Time for a completely new environment from scratch?

(catkit2) maintenances-MacBook-Pro-6:catkit2 ilaginja$ pip install -e .                                        
Obtaining file:///Users/ilaginja/repos/catkit2                                                                 
  Installing build dependencies ... done                                                                       
  Checking if build backend supports build_editable ... done                                                   
  Getting requirements to build editable ... done                                                              
  Installing backend dependencies ... done                                                                     
  Preparing editable metadata (pyproject.toml) ... done                                                        
Building wheels for collected packages: catkit2
  Building editable for catkit2 (pyproject.toml) ... error
  error: subprocess-exited-with-error
  
  × Building editable for catkit2 (pyproject.toml) did not run successfully.
  │ exit code: 1
  ╰─> [37 lines of output]
      2025-05-07 09:42:40,744 - scikit_build_core - INFO - RUN: /Users/ilaginja/opt/miniconda3/envs/catkit2/bin/cmake -E capabilities
      2025-05-07 09:42:40,772 - scikit_build_core - INFO - CMake version: 3.29.4
      *** scikit-build-core 0.11.2 using CMake 3.29.4 (editable)
      2025-05-07 09:42:40,820 - scikit_build_core - INFO - Build directory: /private/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmpbgzqkwlc/build
      *** Configuring CMake...
      2025-05-07 09:42:40,855 - scikit_build_core - INFO - RUN: ninja --version
      2025-05-07 09:42:41,118 - scikit_build_core - INFO - Ninja version: 1.11.1
      2025-05-07 09:42:41,120 - scikit_build_core - INFO - RUN: /Users/ilaginja/opt/miniconda3/envs/catkit2/bin/cmake -S. -B/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmpbgzqkwlc/build -DCMAKE_BUILD_TYPE:STRING=Release -C/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmpbgzqkwlc/build/CMakeInit.txt -DCMAKE_INSTALL_PREFIX=/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmpbgzqkwlc/wheel/platlib -DCMAKE_MAKE_PROGRAM=ninja
      loading initial cache file /var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmpbgzqkwlc/build/CMakeInit.txt
      -- The C compiler identification is AppleClang 13.0.0.13000027
      -- The CXX compiler identification is AppleClang 13.0.0.13000027
      -- Detecting C compiler ABI info
      -- Detecting C compiler ABI info - done
      -- Check for working C compiler: /Library/Developer/CommandLineTools/usr/bin/clang - skipped
      -- Detecting C compile features
      -- Detecting C compile features - done
      -- Detecting CXX compiler ABI info
      -- Detecting CXX compiler ABI info - done
      -- Check for working CXX compiler: /Library/Developer/CommandLineTools/usr/bin/clang++ - skipped
      -- Detecting CXX compile features
      -- Detecting CXX compile features - done
      CMake Error at catkit_core/CMakeLists.txt:45 (find_package):
        Could not find a package configuration file provided by "Eigen3" with any
        of the following names:
      
          Eigen3Config.cmake
          eigen3-config.cmake
      
        Add the installation prefix of "Eigen3" to CMAKE_PREFIX_PATH or set
        "Eigen3_DIR" to a directory containing one of the above files.  If "Eigen3"
        provides a separate development package or SDK, be sure it has been
        installed.
      
      
      -- Configuring incomplete, errors occurred!
      
      *** CMake configuration failed
      [end of output]
  
  note: This error originates from a subprocess, and is likely not a problem with pip.
  ERROR: Failed building editable for catkit2
Failed to build catkit2
ERROR: Could not build wheels for catkit2, which is required to install pyproject.toml-based projects
(catkit2) maintenances-MacBook-Pro-6:catkit2 ilaginja$

ivalaginja · 2025-05-07T07:50:25Z

@ehpor I really don't know what I am doing wrong. I ran conda env update --file environment.yml on this branch, then pip install -e . and I still can't install things.

(catkit2) maintenances-MacBook-Pro-6:catkit2 ilaginja$ pip install -e .
Obtaining file:///Users/ilaginja/repos/catkit2
  Installing build dependencies ... done
  Checking if build backend supports build_editable ... done
  Getting requirements to build editable ... done
  Installing backend dependencies ... done
  Preparing editable metadata (pyproject.toml) ... done
Building wheels for collected packages: catkit2
  Building editable for catkit2 (pyproject.toml) ... error
  error: subprocess-exited-with-error
  
  × Building editable for catkit2 (pyproject.toml) did not run successfully.
  │ exit code: 1
  ╰─> [94 lines of output]
      2025-05-07 09:49:08,924 - scikit_build_core - INFO - RUN: /Users/ilaginja/opt/miniconda3/envs/catkit2/bin/cmake -E capabilities
      2025-05-07 09:49:08,975 - scikit_build_core - INFO - CMake version: 3.27.4
      *** scikit-build-core 0.10.7 using CMake 3.27.4 (editable)
      2025-05-07 09:49:09,068 - scikit_build_core - INFO - Build directory: /private/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build
      *** Configuring CMake...
      2025-05-07 09:49:09,130 - scikit_build_core - INFO - RUN: ninja --version
      2025-05-07 09:49:09,380 - scikit_build_core - INFO - Ninja version: 1.11.1
      2025-05-07 09:49:09,384 - scikit_build_core - INFO - RUN: /Users/ilaginja/opt/miniconda3/envs/catkit2/bin/cmake -S. -B/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build -DCMAKE_BUILD_TYPE:STRING=Release -C/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/CMakeInit.txt -DCMAKE_INSTALL_PREFIX=/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/wheel/platlib -DCMAKE_MAKE_PROGRAM=ninja
      loading initial cache file /var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/CMakeInit.txt
      -- The C compiler identification is AppleClang 13.0.0.13000027
      -- The CXX compiler identification is AppleClang 13.0.0.13000027
      -- Detecting C compiler ABI info
      -- Detecting C compiler ABI info - done
      -- Check for working C compiler: /Library/Developer/CommandLineTools/usr/bin/gcc - skipped
      -- Detecting C compile features
      -- Detecting C compile features - done
      -- Detecting CXX compiler ABI info
      -- Detecting CXX compiler ABI info - done
      -- Check for working CXX compiler: /Library/Developer/CommandLineTools/usr/bin/g++ - skipped
      -- Detecting CXX compile features
      -- Detecting CXX compile features - done
      -- Found nlohmann_json: /Users/ilaginja/opt/miniconda3/envs/catkit2/lib/cmake/nlohmann_json/nlohmann_jsonConfig.cmake (found version "3.9.1")
      -- Found Protobuf: /Users/ilaginja/opt/miniconda3/envs/catkit2/lib/libprotobuf.dylib (found version "3.20.3")
      CMake Warning (dev) at /Users/ilaginja/opt/miniconda3/envs/catkit2/lib/python3.7/site-packages/pybind11/share/cmake/pybind11/FindPythonLibsNew.cmake:101 (message):
        Policy CMP0148 is not set: The FindPythonInterp and FindPythonLibs modules
        are removed.  Run "cmake --help-policy CMP0148" for policy details.  Use
        the cmake_policy command to set the policy and suppress this warning, or
        preferably upgrade to using FindPython, either by calling it explicitly
        before pybind11, or by setting PYBIND11_FINDPYTHON ON before pybind11.
      Call Stack (most recent call first):
        /Users/ilaginja/opt/miniconda3/envs/catkit2/lib/python3.7/site-packages/pybind11/share/cmake/pybind11/pybind11Tools.cmake:50 (find_package)
        /Users/ilaginja/opt/miniconda3/envs/catkit2/lib/python3.7/site-packages/pybind11/share/cmake/pybind11/pybind11Common.cmake:228 (include)
        /Users/ilaginja/opt/miniconda3/envs/catkit2/lib/python3.7/site-packages/pybind11/share/cmake/pybind11/pybind11Config.cmake:250 (include)
        catkit2/CMakeLists.txt:10 (find_package)
      This warning is for project developers.  Use -Wno-dev to suppress it.
      
      -- Found PythonInterp: /Users/ilaginja/opt/miniconda3/envs/catkit2/bin/python3.7 (found suitable version "3.7.16", minimum required is "3.7")
      -- Found PythonLibs: /Users/ilaginja/opt/miniconda3/envs/catkit2/lib/libpython3.7m.dylib
      -- Performing Test HAS_FLTO_THIN
      -- Performing Test HAS_FLTO_THIN - Success
      -- Found pybind11: /Users/ilaginja/opt/miniconda3/envs/catkit2/lib/python3.7/site-packages/pybind11/include (found version "2.13.6")
      -- Found nlohmann_json: /Users/ilaginja/opt/miniconda3/envs/catkit2/lib/cmake/nlohmann_json/nlohmann_jsonConfig.cmake (found suitable version "3.9.1", minimum required is "3.2.0")
      -- Configuring done (4.4s)
      -- Generating done (0.1s)
      -- Build files have been written to: /var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build
      *** Building project with Ninja...
      2025-05-07 09:49:13,954 - scikit_build_core - INFO - RUN: /Users/ilaginja/opt/miniconda3/envs/catkit2/bin/cmake --build /var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build -v
      Change Dir: '/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build'
      
      Run Build Command(s): ninja -v
      [1/61] cd /var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit2 && /Users/ilaginja/opt/miniconda3/envs/catkit2/bin/protoc --proto_path=/Users/ilaginja/repos/catkit2/catkit2/../proto/ --python_out=/Users/ilaginja/repos/catkit2/catkit2/./proto/ /Users/ilaginja/repos/catkit2/catkit2/../proto/core.proto /Users/ilaginja/repos/catkit2/catkit2/../proto/logging.proto /Users/ilaginja/repos/catkit2/catkit2/../proto/service.proto /Users/ilaginja/repos/catkit2/catkit2/../proto/testbed.proto /Users/ilaginja/repos/catkit2/catkit2/../proto/tracing.proto
      [2/61] cd /var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core && /Users/ilaginja/opt/miniconda3/envs/catkit2/bin/protoc --cpp_out /var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen/ -I /Users/ilaginja/repos/catkit2/catkit_core -I /Users/ilaginja/repos/catkit2/proto /Users/ilaginja/repos/catkit2/proto/core.proto
      [3/61] cd /var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core && /Users/ilaginja/opt/miniconda3/envs/catkit2/bin/protoc --cpp_out /var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen/ -I /Users/ilaginja/repos/catkit2/catkit_core -I /Users/ilaginja/repos/catkit2/proto /Users/ilaginja/repos/catkit2/proto/service.proto
      [4/61] cd /var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core && /Users/ilaginja/opt/miniconda3/envs/catkit2/bin/protoc --cpp_out /var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen/ -I /Users/ilaginja/repos/catkit2/catkit_core -I /Users/ilaginja/repos/catkit2/proto /Users/ilaginja/repos/catkit2/proto/tracing.proto
      [5/61] cd /var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core && /Users/ilaginja/opt/miniconda3/envs/catkit2/bin/protoc --cpp_out /var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen/ -I /Users/ilaginja/repos/catkit2/catkit_core -I /Users/ilaginja/repos/catkit2/proto /Users/ilaginja/repos/catkit2/proto/logging.proto
      [6/61] cd /var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core && /Users/ilaginja/opt/miniconda3/envs/catkit2/bin/protoc --cpp_out /var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen/ -I /Users/ilaginja/repos/catkit2/catkit_core -I /Users/ilaginja/repos/catkit2/proto /Users/ilaginja/repos/catkit2/proto/testbed.proto
      [7/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/LocalMemory.cpp.o -MF catkit_core/CMakeFiles/catkit_core.dir/LocalMemory.cpp.o.d -o catkit_core/CMakeFiles/catkit_core.dir/LocalMemory.cpp.o -c /Users/ilaginja/repos/catkit2/catkit_core/LocalMemory.cpp
      [8/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/HostName.cpp.o -MF catkit_core/CMakeFiles/catkit_core.dir/HostName.cpp.o.d -o catkit_core/CMakeFiles/catkit_core.dir/HostName.cpp.o -c /Users/ilaginja/repos/catkit2/catkit_core/HostName.cpp
      [9/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/BuddyAllocator.cpp.o -MF catkit_core/CMakeFiles/catkit_core.dir/BuddyAllocator.cpp.o.d -o catkit_core/CMakeFiles/catkit_core.dir/BuddyAllocator.cpp.o -c /Users/ilaginja/repos/catkit2/catkit_core/BuddyAllocator.cpp
      [10/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/Log.cpp.o -MF catkit_core/CMakeFiles/catkit_core.dir/Log.cpp.o.d -o catkit_core/CMakeFiles/catkit_core.dir/Log.cpp.o -c /Users/ilaginja/repos/catkit2/catkit_core/Log.cpp
      [11/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/Event.cpp.o -MF catkit_core/CMakeFiles/catkit_core.dir/Event.cpp.o.d -o catkit_core/CMakeFiles/catkit_core.dir/Event.cpp.o -c /Users/ilaginja/repos/catkit2/catkit_core/Event.cpp
      [12/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/HybridPoolAllocator.cpp.o -MF catkit_core/CMakeFiles/catkit_core.dir/HybridPoolAllocator.cpp.o.d -o catkit_core/CMakeFiles/catkit_core.dir/HybridPoolAllocator.cpp.o -c /Users/ilaginja/repos/catkit2/catkit_core/HybridPoolAllocator.cpp
      [13/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/HashMap.cpp.o -MF catkit_core/CMakeFiles/catkit_core.dir/HashMap.cpp.o.d -o catkit_core/CMakeFiles/catkit_core.dir/HashMap.cpp.o -c /Users/ilaginja/repos/catkit2/catkit_core/HashMap.cpp
      [14/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/LogConsole.cpp.o -MF catkit_core/CMakeFiles/catkit_core.dir/LogConsole.cpp.o.d -o catkit_core/CMakeFiles/catkit_core.dir/LogConsole.cpp.o -c /Users/ilaginja/repos/catkit2/catkit_core/LogConsole.cpp
      [15/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/PoolAllocator.cpp.o -MF catkit_core/CMakeFiles/catkit_core.dir/PoolAllocator.cpp.o.d -o catkit_core/CMakeFiles/catkit_core.dir/PoolAllocator.cpp.o -c /Users/ilaginja/repos/catkit2/catkit_core/PoolAllocator.cpp
      [16/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/LogFile.cpp.o -MF catkit_core/CMakeFiles/catkit_core.dir/LogFile.cpp.o.d -o catkit_core/CMakeFiles/catkit_core.dir/LogFile.cpp.o -c /Users/ilaginja/repos/catkit2/catkit_core/LogFile.cpp
      [17/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/Client.cpp.o -MF catkit_core/CMakeFiles/catkit_core.dir/Client.cpp.o.d -o catkit_core/CMakeFiles/catkit_core.dir/Client.cpp.o -c /Users/ilaginja/repos/catkit2/catkit_core/Client.cpp
      [18/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/Command.cpp.o -MF catkit_core/CMakeFiles/catkit_core.dir/Command.cpp.o.d -o catkit_core/CMakeFiles/catkit_core.dir/Command.cpp.o -c /Users/ilaginja/repos/catkit2/catkit_core/Command.cpp
      [19/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/ServiceState.cpp.o -MF catkit_core/CMakeFiles/catkit_core.dir/ServiceState.cpp.o.d -o catkit_core/CMakeFiles/catkit_core.dir/ServiceState.cpp.o -c /Users/ilaginja/repos/catkit2/catkit_core/ServiceState.cpp
      [20/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/SharedMemory.cpp.o -MF catkit_core/CMakeFiles/catkit_core.dir/SharedMemory.cpp.o.d -o catkit_core/CMakeFiles/catkit_core.dir/SharedMemory.cpp.o -c /Users/ilaginja/repos/catkit2/catkit_core/SharedMemory.cpp
      [21/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/DataStream.cpp.o -MF catkit_core/CMakeFiles/catkit_core.dir/DataStream.cpp.o.d -o catkit_core/CMakeFiles/catkit_core.dir/DataStream.cpp.o -c /Users/ilaginja/repos/catkit2/catkit_core/DataStream.cpp
      [22/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/Server.cpp.o -MF catkit_core/CMakeFiles/catkit_core.dir/Server.cpp.o.d -o catkit_core/CMakeFiles/catkit_core.dir/Server.cpp.o -c /Users/ilaginja/repos/catkit2/catkit_core/Server.cpp
      [23/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/Property.cpp.o -MF catkit_core/CMakeFiles/catkit_core.dir/Property.cpp.o.d -o catkit_core/CMakeFiles/catkit_core.dir/Property.cpp.o -c /Users/ilaginja/repos/catkit2/catkit_core/Property.cpp
      [24/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/Timing.cpp.o -MF catkit_core/CMakeFiles/catkit_core.dir/Timing.cpp.o.d -o catkit_core/CMakeFiles/catkit_core.dir/Timing.cpp.o -c /Users/ilaginja/repos/catkit2/catkit_core/Timing.cpp
      [25/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/Shareable.cpp.o -MF catkit_core/CMakeFiles/catkit_core.dir/Shareable.cpp.o.d -o catkit_core/CMakeFiles/catkit_core.dir/Shareable.cpp.o -c /Users/ilaginja/repos/catkit2/catkit_core/Shareable.cpp
      [26/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/MessageBroker.cpp.o -MF catkit_core/CMakeFiles/catkit_core.dir/MessageBroker.cpp.o.d -o catkit_core/CMakeFiles/catkit_core.dir/MessageBroker.cpp.o -c /Users/ilaginja/repos/catkit2/catkit_core/MessageBroker.cpp
      [27/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/proto/core.pb.cc.o -MF catkit_core/CMakeFiles/catkit_core.dir/proto/core.pb.cc.o.d -o catkit_core/CMakeFiles/catkit_core.dir/proto/core.pb.cc.o -c /var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/proto/core.pb.cc
      FAILED: catkit_core/CMakeFiles/catkit_core.dir/proto/core.pb.cc.o
      /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/proto/core.pb.cc.o -MF catkit_core/CMakeFiles/catkit_core.dir/proto/core.pb.cc.o.d -o catkit_core/CMakeFiles/catkit_core.dir/proto/core.pb.cc.o -c /var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/proto/core.pb.cc
      clang: error: no such file or directory: '/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/proto/core.pb.cc'
      clang: error: no input files
      [28/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/Util.cpp.o -MF catkit_core/CMakeFiles/catkit_core.dir/Util.cpp.o.d -o catkit_core/CMakeFiles/catkit_core.dir/Util.cpp.o -c /Users/ilaginja/repos/catkit2/catkit_core/Util.cpp
      [29/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/LogForwarder.cpp.o -MF catkit_core/CMakeFiles/catkit_core.dir/LogForwarder.cpp.o.d -o catkit_core/CMakeFiles/catkit_core.dir/LogForwarder.cpp.o -c /Users/ilaginja/repos/catkit2/catkit_core/LogForwarder.cpp
      [30/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/Tensor.cpp.o -MF catkit_core/CMakeFiles/catkit_core.dir/Tensor.cpp.o.d -o catkit_core/CMakeFiles/catkit_core.dir/Tensor.cpp.o -c /Users/ilaginja/repos/catkit2/catkit_core/Tensor.cpp
      [31/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/Uuid.cpp.o -MF catkit_core/CMakeFiles/catkit_core.dir/Uuid.cpp.o.d -o catkit_core/CMakeFiles/catkit_core.dir/Uuid.cpp.o -c /Users/ilaginja/repos/catkit2/catkit_core/Uuid.cpp
      [32/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/Tracing.cpp.o -MF catkit_core/CMakeFiles/catkit_core.dir/Tracing.cpp.o.d -o catkit_core/CMakeFiles/catkit_core.dir/Tracing.cpp.o -c /Users/ilaginja/repos/catkit2/catkit_core/Tracing.cpp
      [33/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/ServiceProxy.cpp.o -MF catkit_core/CMakeFiles/catkit_core.dir/ServiceProxy.cpp.o.d -o catkit_core/CMakeFiles/catkit_core.dir/ServiceProxy.cpp.o -c /Users/ilaginja/repos/catkit2/catkit_core/ServiceProxy.cpp
      [34/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/Types.cpp.o -MF catkit_core/CMakeFiles/catkit_core.dir/Types.cpp.o.d -o catkit_core/CMakeFiles/catkit_core.dir/Types.cpp.o -c /Users/ilaginja/repos/catkit2/catkit_core/Types.cpp
      [35/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/Service.cpp.o -MF catkit_core/CMakeFiles/catkit_core.dir/Service.cpp.o.d -o catkit_core/CMakeFiles/catkit_core.dir/Service.cpp.o -c /Users/ilaginja/repos/catkit2/catkit_core/Service.cpp
      [36/61] /Library/Developer/CommandLineTools/usr/bin/g++ -DPROTOBUF_USE_DLLS -I/var/folders/28/w2xb9gbd7_g3wxh48m73qv840000gp/T/tmp0xk3ekoc/build/catkit_core/gen -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include -isystem /Users/ilaginja/opt/miniconda3/envs/catkit2/include/eigen3 -O3 -DNDEBUG -std=gnu++17 -arch x86_64 -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX12.1.sdk -fPIC -g -MD -MT catkit_core/CMakeFiles/catkit_core.dir/TestbedProxy.cpp.o -MF catkit_core/CMakeFiles/catkit_core.dir/TestbedProxy.cpp.o.d -o catkit_core/CMakeFiles/catkit_core.dir/TestbedProxy.cpp.o -c /Users/ilaginja/repos/catkit2/catkit_core/TestbedProxy.cpp
      ninja: build stopped: subcommand failed.
      
      
      *** CMake build failed
      [end of output]
  
  note: This error originates from a subprocess, and is likely not a problem with pip.
  ERROR: Failed building editable for catkit2
Failed to build catkit2
ERROR: Could not build wheels for catkit2, which is required to install pyproject.toml-based projects

ehpor · 2025-05-07T07:57:12Z

@ivalaginja That looks like the same error that @brynickson had during the hackathon. I don't remember how she fixed that or what the problem was. I do remember if felt like magic.

ehpor · 2025-05-07T07:58:21Z

And yes, protobuf is also something that I intend to rip out after the message broker transition. We will not need it anymore after that.

ivalaginja · 2025-05-07T07:59:31Z

@ivalaginja That looks like the same error that @brynickson had during the hackathon. I don't remember how she fixed that or what the problem was. I do remember if felt like magic.

@raphaelpclt @brynickson any pointers on how this was fixed?

ivalaginja · 2025-05-07T14:28:24Z

@ivalaginja That looks like the same error that @brynickson had during the hackathon. I don't remember how she fixed that or what the problem was. I do remember if felt like magic.

@raphaelpclt @brynickson any pointers on how this was fixed?

@ehpor I am now told by @raphaelpclt that you were actually working through this with @brynickson, who is currently out of office ^^

raphaelpclt · 2025-05-07T14:36:50Z

@ivalaginja That looks like the same error that @brynickson had during the hackathon. I don't remember how she fixed that or what the problem was. I do remember if felt like magic.

@raphaelpclt @brynickson any pointers on how this was fixed?

@ehpor I am now told by @raphaelpclt that you were actually working through this with @brynickson, who is currently out of office ^^

From what I remember it was a compiling tools issue - a package needing to be updated. Maybe cmake?

ivalaginja · 2025-05-07T14:39:19Z

Ha. @raphaelpclt and I found a solution: upgrade cmake manually to version 3.31.2

Installation worked now. Tests tbd.

ivalaginja

The python test code makes sense to me, and the tests pass on my MacOs 12.2. Note that I run a conda environment with Python 3.9.12 (THD2 default) and I had to upgrade cmake to 3.31.2.

If you could still answer my last question that'd be great, but that's just for my own understanding anyway.

I guess we better hold off of merging until we get more reviews in?

ehpor · 2025-05-07T16:09:09Z

Yep, forgot to show my testing:

Tested on Windows 11, AMD 5900X + benchmarks, Python 3.7
Tested on Ubuntu 20.04, AMD Threadripper 5995WX + benchmarks, Python 3.8 & 3.11
Tested on MacOS Sonoma 14.7, Macbook M2 Pro + benchmarks, Python 3.8, 3.11

There are two PRs open (#310 & #311), which reference this one, that make improvements / bugfixes. And a third one on the way working on network communications between message brokers (introducing a ShardedMessageBroker class).

ivalaginja · 2025-05-07T16:19:54Z

@ehpor btw, does this PR require rewriting some of the catkit2 docs? Or does this touch all our unwritten docs only? ^^

ehpor · 2025-05-07T16:32:49Z

Only touches the unwritten part of the docs 😭

ehpor · 2025-05-07T19:07:55Z

Merging at the end of today @RemiSoummer, unless you have objections.

RemiSoummer · 2025-05-07T21:36:23Z

@ehpor I haven't had a chance to even start looking at it, so go ahead if you need to merge it.

ehpor added the enhancement New feature or request label Dec 17, 2024

ehpor self-assigned this Dec 17, 2024

ehpor changed the title ~~Feature/message broker~~ Add a shared-memory message broker. Dec 17, 2024

ehpor force-pushed the feature/message_broker branch 4 times, most recently from cb233fb to 45704c2 Compare December 23, 2024 23:31

ehpor force-pushed the feature/message_broker branch from 714535a to 43f7f7b Compare January 24, 2025 00:18

ehpor force-pushed the feature/message_broker branch 7 times, most recently from 9ffe87f to f8c80fd Compare February 25, 2025 22:28

ehpor force-pushed the feature/message_broker branch 3 times, most recently from a9ccf76 to 7ece328 Compare March 18, 2025 19:03

ehpor added 8 commits April 23, 2025 08:08

Add UUID generator based on two MT19937 pseudo-random number generato…

1e043a2

…rs seeded by true randomness.

Add interface for Cuda shared memory.

428600d

Add rough layout for message broker.

6ed1920

Add benchmark for UUID generator.

797e995

Use a typename for UUIDs.

7b38451

Flesh out interface more.

b4ef884

Only use a single PublishMessage() method.

c47585f

Topic header copy constructor and copy assignment operator.

124d8b4

Remove unused Synchronization class.

19630b3

This has been replaced by Event.

This was referenced Apr 29, 2025

Be strict about shared memory lifetime #308

Merged

Add a benchmark for latency for events. #310

Merged

Add an event based on futex #311

Merged

ivalaginja requested a review from alexisyslau May 5, 2025 15:00

ivalaginja reviewed May 6, 2025

View reviewed changes

alexisyslau reviewed May 6, 2025

View reviewed changes

ehpor added 2 commits May 7, 2025 00:00

Add comments for event testing.

450b8a6

Use i + 10 to make frame IDs different from payload.

f99de86

ivalaginja reviewed May 7, 2025

View reviewed changes

ivalaginja approved these changes May 7, 2025

View reviewed changes

ehpor merged commit 8d94635 into develop May 7, 2025
6 checks passed

ehpor deleted the feature/message_broker branch May 7, 2025 22:07

Conversation

ehpor commented Dec 17, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

ehpor commented Mar 19, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

ehpor commented Mar 19, 2025

Uh oh!

ehpor commented Mar 20, 2025

Uh oh!

ehpor commented Apr 19, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

ivalaginja left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

alexisyslau May 6, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

alexisyslau May 6, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

ivalaginja left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

ivalaginja commented May 7, 2025

Uh oh!

ehpor commented May 7, 2025

Uh oh!

ivalaginja commented May 7, 2025

Uh oh!

ivalaginja commented May 7, 2025

ehpor commented Dec 17, 2024 •

edited

Loading

ehpor commented Mar 19, 2025 •

edited

Loading

ehpor commented Apr 19, 2025 •

edited

Loading

alexisyslau May 6, 2025 •

edited

Loading

alexisyslau May 6, 2025 •

edited

Loading

ehpor commented May 7, 2025 •

edited

Loading