first cut

monolith_instrumenting/slides.Rmd

@@ -51,6 +51,36 @@ Chapter 3 and 5 in the book
 
 ## Pipeline Semantics
 
+- signals are logs, traces and metrics
+- add new signals like sessions in the Future
+- signals have attributes
+
+## Attributes (from the sqlquery)
+
+```yaml
+      - sql: "SELECT sum(xact_rollback) as count_test FROM pg_stat_database; "
+        metrics:
+          - metric_name: query_errors # E from REDS
+            value_column: count_test
+```
+
+- the value_column is count_test but other columns will be stored as attributes
+
+## Attributes (from today's lab)
+
+```python
+# Now create a counter instrument to make measurements with
+roll_counter = meter.create_counter(
+    "dice.rolls",
+    description="The number of rolls by roll value",
+)
+roll_counter.add(1, {"roll.value": result})
+```
+
+- roll.value is an attribute
+
+## Future of Open Telemetry
+
 ![OpenTelemetry’s long-term support guarantees](images/otelSupport.png)
 
 ## Foundation of OpenTelemetry is built on two points: 
@@ -67,4 +97,73 @@ Chapter 3 and 5 in the book
 
 - can provide schema transformations
 - support changes in semantic conventions (and their associated support in analysis tools) 
-- without having to reinstrument or redefine output from existing services
+- without having to reinstrument or redefine output from existing services
+
+## setting up OpenTelemetry is twofold
+
+- installing the software development kit (SDK)
+- for python we do this with pip packages
+- installing instrumentation
+- some instrumentation is automatic with the SDK, some is custom
+
+## requirements.txt
+
+```txt
+opentelemetry-distro
+opentelemetry-exporter-otlp
+```
+
+- then run `opentelemetry-bootstrap -a install`
+
+## logs
+
+```python
+import logging
+logging.basicConfig(level=logging.INFO)
+logger = logging.getLogger(__name__)
+logger.warn("{} is rolling the dice: {}", "Rich", "6")
+```
+
+- built in to python
+
+## Logger Provider
+
+![The LoggerProvider framework](images/logger.png)
+
+## custom traces
+
+```python
+from opentelemetry import trace
+
+# Acquire a tracer
+tracer = trace.get_tracer("diceroller.tracer")
+with tracer.start_as_current_span("roll") as rollspan:
+    rollspan.set_attribute("roll.value", 6) 
+```
+
+## Tracer Provider
+
+![The TracerProvider framework](images/tracer.png)
+
+## custom metrics
+
+```python
+from opentelemetry import metrics
+meter = metrics.get_meter("diceroller.meter")
+# Now create a counter instrument to make measurements with
+roll_counter = meter.create_counter(
+    "dice.rolls",
+    description="The number of rolls by roll value",
+)
+roll_counter.add(1, {"roll.value": "6"})
+```
+
+## Meter Provider
+
+![The MeterProvider framework](images/metric.png)
+
+## Conclusion
+
+- work required to reinstrument a large system a form of vendor lock-in
+- advantage of OpenTelemetry is that once it’s done
+- OpenTelemetry is a standard that works with every observability system.