How to Create Separate Arize Projects for Multiple Django Apps in One Service?

·Mar 12, 2026 06:35 AM·

Hi Teaam, I have a single django service which internally has multiple apps. For each app I want to create a project in arize. How do we do this?

tracer_provider = register(
  project_name="conversational-agent", # Default is 'default'
  auto_instrument=True # Auto-instrument your app based on installed OI dependencies
)

Should the above code snippet be in the place where we are creating or compiling langgraph's workflow? or should it be acting like a middleware between api invocation and graph compilation?

10 comments

· Sorted by Oldest

Chandrahaas J.
·
·
🔒[private user] how does it automatically connect to the arize container? where is it getting these creds from?
Chandrahaas J.
·
·
🔒[private user] how does it automatically connect to the arize container? where is it getting these creds from?
Chandrahaas J.
·
·
🔒[private user] what if i locally hosted phoenix, how do I connect to that?
Chandrahaas J.
·
·
WARNING Overriding of current TracerProvider is not allowed 🔭 OpenTelemetry Tracing Details 🔭 | Phoenix Project: input-summarizer | Span Processor: SimpleSpanProcessor | Collector Endpoint: 0.0.0.0:4317 | Transport: gRPC | Transport Headers: {} | | Using a default SpanProcessor. `add_span_processor` will overwrite this default. | | ⚠️ WARNING: It is strongly advised to use a BatchSpanProcessor in production environments. | | `register` has set this TracerProvider as the global OpenTelemetry default. | To disable this behavior, call `register` with `set_global_tracer_provider=False`. None of PyTorch, TensorFlow >= 2.0, or Flax have been found. Models won't be available and only tokenizers, configuration and file/data utilities can be used. WARNING Overriding of current TracerProvider is not allowed WARNING Attempting to instrument while already instrumented 🔭 OpenTelemetry Tracing Details 🔭 | Phoenix Project: conversational-agent | Span Processor: SimpleSpanProcessor | Collector Endpoint: 0.0.0.0:4317 | Transport: gRPC | Transport Headers: {} | | Using a default SpanProcessor. `add_span_processor` will overwrite this default. | | ⚠️ WARNING: It is strongly advised to use a BatchSpanProcessor in production environments. | | `register` has set this TracerProvider as the global OpenTelemetry default. | To disable this behavior, call `register` with `set_global_tracer_provider=False`. 🔒[private user] this is what my logs look like when I added the integration like you suggested. Can you explain whats going on and inform me if I have to change anything?
Chandrahaas J.
·
·
🔒[private user], this is how everything looks like. I have two different workflows which get triggered. I have put tracer_provider = register( project_name="conversational-agent", # Default is 'default' auto_instrument=True # Auto-instrument your app based on installed OI dependencies ) and tracer_provider = register( project_name="input-summarizer", # Default is 'default' auto_instrument=True # Auto-instrument your app based on installed OI dependencies ) for tracing both. are these both considered a single process?
Chandrahaas J.
·
·
🔒[private user] what is a process you mentioned?
Chandrahaas J.
·
·
then from what i understand, You have to setup the register the collector only once and this should probably be done in settings.py of django. After registering the collector, how do we change the project for each app?
Chandrahaas J.
·
·
🔒[private user] then from what i understand, You have to setup the register the collector only once and this should probably be done in settings.py of django. After registering the collector, how do we change the project for each app?
Chandrahaas J.
·
·
🔒[private user] yes please, give me the code for what the setup must look like in settings.py and what should be done in all the other apps?
Chandrahaas J.
·
·
🔒[private user] will the same code from Arize documentation work for phoenix?

Chandrahaas J.
·
·
🔒[private user] how does it automatically connect to the arize container? where is it getting these creds from?
Chandrahaas J.
·
·
🔒[private user] how does it automatically connect to the arize container? where is it getting these creds from?
Chandrahaas J.
·
·
🔒[private user] what if i locally hosted phoenix, how do I connect to that?
Chandrahaas J.
·
·
WARNING Overriding of current TracerProvider is not allowed 🔭 OpenTelemetry Tracing Details 🔭 | Phoenix Project: input-summarizer | Span Processor: SimpleSpanProcessor | Collector Endpoint: 0.0.0.0:4317 | Transport: gRPC | Transport Headers: {} | | Using a default SpanProcessor. `add_span_processor` will overwrite this default. | | ⚠️ WARNING: It is strongly advised to use a BatchSpanProcessor in production environments. | | `register` has set this TracerProvider as the global OpenTelemetry default. | To disable this behavior, call `register` with `set_global_tracer_provider=False`. None of PyTorch, TensorFlow >= 2.0, or Flax have been found. Models won't be available and only tokenizers, configuration and file/data utilities can be used. WARNING Overriding of current TracerProvider is not allowed WARNING Attempting to instrument while already instrumented 🔭 OpenTelemetry Tracing Details 🔭 | Phoenix Project: conversational-agent | Span Processor: SimpleSpanProcessor | Collector Endpoint: 0.0.0.0:4317 | Transport: gRPC | Transport Headers: {} | | Using a default SpanProcessor. `add_span_processor` will overwrite this default. | | ⚠️ WARNING: It is strongly advised to use a BatchSpanProcessor in production environments. | | `register` has set this TracerProvider as the global OpenTelemetry default. | To disable this behavior, call `register` with `set_global_tracer_provider=False`. 🔒[private user] this is what my logs look like when I added the integration like you suggested. Can you explain whats going on and inform me if I have to change anything?
Chandrahaas J.
·
·
🔒[private user], this is how everything looks like. I have two different workflows which get triggered. I have put tracer_provider = register( project_name="conversational-agent", # Default is 'default' auto_instrument=True # Auto-instrument your app based on installed OI dependencies ) and tracer_provider = register( project_name="input-summarizer", # Default is 'default' auto_instrument=True # Auto-instrument your app based on installed OI dependencies ) for tracing both. are these both considered a single process?
Chandrahaas J.
·
·
🔒[private user] what is a process you mentioned?
Chandrahaas J.
·
·
then from what i understand, You have to setup the register the collector only once and this should probably be done in settings.py of django. After registering the collector, how do we change the project for each app?
Chandrahaas J.
·
·
🔒[private user] then from what i understand, You have to setup the register the collector only once and this should probably be done in settings.py of django. After registering the collector, how do we change the project for each app?
Chandrahaas J.
·
·
🔒[private user] yes please, give me the code for what the setup must look like in settings.py and what should be done in all the other apps?
Chandrahaas J.
·
·
🔒[private user] will the same code from Arize documentation work for phoenix?