intergral / deep-python-client Goto Github PK

Describe the bug
When the application with deep is shutdown there is sometimes an error thrown.

To Reproduce
Reproduction is not easy. Essentially if deep shuts down while set_trace is running there can be an exception thrown.

Expected behavior
No exception on shutdown.

Desktop (please complete the following information):

• OS: linux
• Version: main
• Python Version cpython 3.12

Additional context
Probably just need to add atexit handler.

Is your feature request related to a problem? Please describe.
Add a span event for snapshots. As it is possible to have more than 1 snapshot within the scope of a span we should use events rather than attributes. This provides a scope to set the snapshot id, context id and tracepoint ids as attributes within the events. The events can then also set the name as the snapshot location, and provide scoped time.

Describe the solution you'd like
Call add_event when a snapshot is collected.

Describe the bug
Creating a new tracepoint logger for deep requires the order to be set to -1 to override the default provider.

To Reproduce
Steps to reproduce the behavior:

1. Create a new plugin to act as tracepoint logger
2. register the plugin
3. it will not be used until the order is set to -1

Expected behavior
Shouldn't need to specify the order on custom plugins to override plugins that do not allow for multiple.

Desktop (please complete the following information):

• Version [e.g. 22] main@ac1ec83e
• Python Version [e.g. cpython 3.12] cpython 3.10

Additional context
Add any other context about the problem here.

Describe the bug
The snapshot duration is missing.

To Reproduce
Steps to reproduce the behavior:

1. Send snapsshot
2. see duration is missing

Expected behavior
Duration should be set to the time taken to process the sanpshot.

Screenshots

Desktop (please complete the following information):

• Version [e.g. 22] 1.1.0
• Python Version [e.g. cpython 3.12] cpython

Additional context
Add any other context about the problem here.

Describe the bug
The error below is given when using the api to create tracepoints (_deep.register_tracepoint("simple_test.py", 41, metrics=[MetricDefinition("uuid_length", "histogram")])

2024-02-05 18:19:15,907 - deep - ERROR - Error registering metric histogram deep_uuid_length
Traceback (most recent call last):
  File "/home/bdonnell/repo/github/intergral/test-app-python/venv/lib/python3.10/site-packages/deep/api/plugin/metric/otel_metrics.py", line 107, in histogram
    histogram = self.__check_cache(name, 'histogram',
  File "/home/bdonnell/repo/github/intergral/test-app-python/venv/lib/python3.10/site-packages/deep/api/plugin/metric/otel_metrics.py", line 47, in __check_cache
    default = from_default()
  File "/home/bdonnell/repo/github/intergral/test-app-python/venv/lib/python3.10/site-packages/deep/api/plugin/metric/otel_metrics.py", line 108, in <lambda>
    lambda: get_meter('deep').create_histogram(f'{namespace}_{name}', unit,
  File "/home/bdonnell/repo/github/intergral/test-app-python/venv/lib/python3.10/site-packages/opentelemetry/sdk/metrics/_internal/__init__.py", line 193, in create_histogram
    ) = self._is_instrument_registered(name, _Histogram, unit, description)
  File "/home/bdonnell/repo/github/intergral/test-app-python/venv/lib/python3.10/site-packages/opentelemetry/metrics/_internal/__init__.py", line 225, in _is_instrument_registered
    instrument_id = ",".join(
TypeError: sequence item 2: expected str instance, NoneType found

To Reproduce
Steps to reproduce the behavior:

1. Use the api deep.register_tracepoint to create a metric with just a name a type.
2. See error

Expected behavior
A metric should be created with default values.

Desktop (please complete the following information):

• Python Version [e.g. cpython 3.12] cpython 3.10.12

Additional context
Add any other context about the problem here.

Describe the bug
The tracepoint attribute is missing from the snapshots. This affects the search in Deep.

To Reproduce
Steps to reproduce the behavior:

1. Create a snapshot
2. Inspect the attributes
3. See error

Expected behavior
The attribute tracepoint should be attached to the snapshots to connect all snapshots to the tracepoint that triggered them.

Desktop (please complete the following information):

• Version [e.g. 22] 1.1.0

Additional context
Add any other context about the problem here.

Is your feature request related to a problem? Please describe.
When customising the deep agent it is a bit awkward registering plugins.

Describe the solution you'd like
Some way to make it simpler such as:

deep.start({}, plugins=[MyDeepPlugin])

Describe alternatives you've considered
A clear and concise description of any alternative solutions or features you've considered.

Additional context
Add any other context or screenshots about the feature request here.