feat: add core dimensional models and service inventory report

- Add dim_otel_resources for resource attribute deduplication
- Add fct_otel_* views joining staging with resource dimension
- Add rpt_otel_service_inventory for service discovery
- Update staging models with resource_id foreign key
- Refresh test parquet files with consistent resource data

Author: smithclay

README.md

@@ -2,7 +2,7 @@
 
 A dbt package for transforming OpenTelemetry (OTLP) data into queryable, analytics-ready schemas.
 
-[![CI](https://github.com/smithclay/dbt-opentelemetry/actions/workflows/ci.yml/badge.svg)](https://github.com/smithclay/dbt-opentelemetry/actions/workflows/ci.yml)
+[![CI](https://github.com/smithclay/dbt-opentelemetry/actions/workflows/ci.yml/badge.svg)](https://github.com/smithclay/dbt_opentelemetry/actions/workflows/ci.yml)
 [![dbt Hub](https://img.shields.io/badge/dbt%20Hub-dbt__opentelemetry-orange)](https://hub.getdbt.com/)
 [![License](https://img.shields.io/badge/License-Apache%202.0-blue.svg)](LICENSE)
 

dbt_project.yml

@@ -8,27 +8,33 @@ model-paths: ["models"]
 macro-paths: ["macros"]
 
 vars:
-  # Source configuration - override these to match your data location
-  # otel_database: defaults to target.database (handled in sources.yml)
-  otel_schema: "raw"
-  otel_logs_table: "logs"
-  otel_spans_table: "spans"
-  otel_gauge_table: "gauge"
-  otel_sum_table: "sum"
+    # Source configuration - override these to match your data location
+    # otel_database: defaults to target.database (handled in sources.yml)
+    otel_schema: "raw"
+    otel_logs_table: "logs"
+    otel_spans_table: "spans"
+    otel_gauge_table: "gauge"
+    otel_sum_table: "sum"
 
-  # Feature flags for optional domain models
-  # Set to true in your dbt_project.yml to enable
-  otel_enable_k8s: false
-  otel_enable_ai_agents: false
-  otel_enable_apm: false
+    # Feature flags for optional domain models
+    # Set to true in your dbt_project.yml to enable
+    otel_enable_k8s: false
+    otel_enable_ai_agents: false
+    otel_enable_apm: false
 
 models:
-  dbt_opentelemetry:
-    staging:
-      +materialized: view
-    k8s:
-      +materialized: table
-    ai_agents:
-      +materialized: table
-    apm:
-      +materialized: table
+    dbt_opentelemetry:
+        staging:
+            +materialized: view
+        core:
+            +materialized: view
+            dim_otel_resources:
+                +materialized: incremental
+        reporting:
+            +materialized: table
+        k8s:
+            +materialized: table
+        ai_agents:
+            +materialized: table
+        apm:
+            +materialized: table

integration_tests/macros/test_helpers.sql

@@ -30,6 +30,7 @@
   {{ log("Created table " ~ logs_table, info=True) }}
 
   {# Spans table - transform to OTLP schema #}
+  {# Note: end_time computed from start + duration since parquet end_timestamp units vary #}
   {% set spans_table = schema_name ~ '.' ~ var('otel_spans_table', 'spans') %}
   {{ log("Creating table " ~ spans_table ~ "...", info=True) }}
   {% set spans_sql %}
@@ -42,7 +43,7 @@
       span_name as name,
       span_kind as kind,
       epoch_ns(timestamp) as start_time_unix_nano,
-      end_timestamp as end_time_unix_nano,
+      epoch_ns(timestamp) + duration as end_time_unix_nano,
       resource_attributes,
       scope_name,
       scope_version,

integration_tests/testdata/gauge.parquet

@@ -2,6 +2,10 @@
   $."{{ key | replace('"', '\\"') }}"
 {%- endmacro %}
 
+{% macro resource_hash(json_col) -%}
+  md5({{ json_col }}::varchar)
+{%- endmacro %}
+
 {% macro attr_str(json_col, key, default=None) -%}
   coalesce(
     json_extract_string({{ json_col }}::json, '{{ dbt_opentelemetry.json_path(key) }}'),

integration_tests/testdata/logs.parquet

@@ -1,17 +1,52 @@
 {{ config(enabled=var('otel_enable_ai_agents', false)) }}
 
+-- AI agent events with resource fields already promoted from dim_otel_resources
+
 with logs as (
   select *
-  from {{ ref('stg_otel_logs') }}
+  from {{ ref('fct_otel_logs') }}
 )
 
 select
-  logs.*,
-  {{ dbt_opentelemetry.coalesce_attr_str(['resource_attributes', 'attributes'], 'agent.id') }} as agent_id,
-  {{ dbt_opentelemetry.coalesce_attr_str(['resource_attributes', 'attributes'], 'agent.session_id') }} as session_id,
-  {{ dbt_opentelemetry.coalesce_attr_str(['resource_attributes', 'attributes'], 'tool.name') }} as tool_name,
-  {{ dbt_opentelemetry.coalesce_attr_str(['resource_attributes', 'attributes'], 'llm.model') }} as llm_model,
-  {{ dbt_opentelemetry.attr_int('attributes', 'llm.tokens_in') }} as tokens_in,
-  {{ dbt_opentelemetry.attr_int('attributes', 'llm.tokens_out') }} as tokens_out,
-  {{ dbt_opentelemetry.attr_double('attributes', 'llm.cost_usd') }} as cost_usd
+  -- Core log fields
+  resource_id,
+  trace_id,
+  span_id,
+  logged_at,
+  observed_at,
+  event_ts,
+  severity_number,
+  severity_text,
+  severity_class,
+  is_error,
+  body,
+  scope_name,
+  attributes,
+
+  -- Service identity (from resource dimension)
+  service_name,
+  service_namespace,
+  service_version,
+
+  -- Agent-specific fields (typically in attributes)
+  {{ coalesce_attr_str(['resource_attributes', 'attributes'], 'agent.id') }} as agent_id,
+  {{ coalesce_attr_str(['resource_attributes', 'attributes'], 'agent.session_id') }} as session_id,
+  {{ coalesce_attr_str(['resource_attributes', 'attributes'], 'agent.name') }} as agent_name,
+  {{ attr_str('attributes', 'agent.event.type') }} as event_type,
+
+  -- Tool usage
+  {{ attr_str('attributes', 'tool.name') }} as tool_name,
+  {{ attr_str('attributes', 'tool.result') }} as tool_result,
+
+  -- LLM usage
+  {{ coalesce_attr_str(['resource_attributes', 'attributes'], 'llm.model') }} as llm_model,
+  {{ attr_str('attributes', 'llm.provider') }} as llm_provider,
+  {{ attr_int('attributes', 'llm.tokens_in') }} as tokens_in,
+  {{ attr_int('attributes', 'llm.tokens_out') }} as tokens_out,
+  {{ attr_double('attributes', 'llm.cost_usd') }} as cost_usd,
+  {{ attr_double('attributes', 'llm.latency_ms') }} as llm_latency_ms,
+
+  -- Full resource blob
+  resource_attributes
+
 from logs

integration_tests/testdata/spans.parquet

@@ -9,11 +9,18 @@ with events as (
 select
   session_id,
   max(agent_id) as agent_id,
+  max(agent_name) as agent_name,
+  max(service_name) as service_name,
   min(event_ts) as session_start_ts,
   max(event_ts) as session_end_ts,
+  extract(epoch from max(event_ts) - min(event_ts)) as session_duration_seconds,
   count(*) as event_count,
+  count(distinct tool_name) filter (where tool_name is not null) as unique_tools_used,
+  count(*) filter (where tool_name is not null) as tool_invocations,
+  count(*) filter (where is_error) as error_count,
   sum(coalesce(tokens_in, 0)) as tokens_in_total,
   sum(coalesce(tokens_out, 0)) as tokens_out_total,
+  sum(coalesce(tokens_in, 0) + coalesce(tokens_out, 0)) as tokens_total,
   sum(coalesce(cost_usd, 0)) as cost_usd_total
 from events
 group by 1