Merge branch 'main' into fix/use-default-endpoint-for-empty-string

Author: cijothomas

docs/release_0.30.md

@@ -0,0 +1,68 @@
+# Release Notes 0.30
+
+OpenTelemetry Rust 0.30 introduces a few breaking changes to the
+`opentelemetry_sdk` crate in the `metrics` feature. These changes were essential
+to drive the Metrics SDK towards stability. With this release, the Metrics SDK
+is officially declared stable. The Metrics API was declared stable last year,
+and previously, the Logs API, SDK, and OTel-Appender-Tracing were also marked
+stable. Importantly, no breaking changes have been introduced to components
+already marked as stable.
+
+It is worth noting that the `opentelemetry-otlp` crate remains in a
+Release-Candidate state and is not yet considered stable. With the API and SDK
+for Logs and Metrics now stable, the focus will shift towards further refining
+and stabilizing the OTLP Exporters in upcoming releases. Additionally,
+Distributed Tracing is expected to progress towards stability, addressing key
+interoperability challenges.
+
+For detailed changelogs of individual crates, please refer to their respective
+changelog files. This document serves as a summary of the main changes.
+
+## Key Changes
+
+### Metrics SDK Improvements
+
+1. **Stabilized "view" features**: Previously under an experimental feature
+   flag, views can now be used to modify the name, unit, description, and
+   cardinality limit of a metric. Advanced view capabilities, such as changing
+   aggregation or dropping attributes, remain under the experimental feature
+   flag.
+
+2. **Cardinality capping**: Introduced the ability to cap cardinality and
+   configure limits using views.
+
+3. **Polished public API**: Refined the public API to hide implementation
+   details from exporters, enabling future internal optimizations and ensuring
+   consistency. Some APIs related to authoring custom metric readers have been
+   moved behind experimental feature flags. These advanced use cases require
+   more time to finalize the API surface before being included in the stable
+   release.
+
+### Context-Based Suppression
+
+Added the ability to suppress telemetry based on Context. This feature prevents
+telemetry-induced-telemetry scenarios and addresses a long-standing issue. Note
+that suppression relies on proper context propagation. Certain libraries used in
+OTLP Exporters utilize `tracing` but do not adopt OpenTelemetry's context
+propagation. As a result, not all telemetry is automatically suppressed with
+this feature. Improvements in this area are expected in future releases.
+
+## Next Release
+
+In the [next
+release](https://github.com/open-telemetry/opentelemetry-rust/milestone/22), the
+focus will shift to OTLP Exporters and Distributed Tracing, specifically
+resolving
+[interoperability](https://github.com/open-telemetry/opentelemetry-rust/issues/2420)
+issues with `tokio-tracing` and other fixes required to drive Distributed
+Tracing towards stability.
+
+## Acknowledgments
+
+Thank you to everyone who contributed to this milestone. We welcome your
+feedback through GitHub issues or discussions in the OTel-Rust Slack channel
+[here](https://cloud-native.slack.com/archives/C03GDP0H023).
+
+We are also excited to announce that [Anton Grübel](https://github.com/gruebel)
+and [Björn Antonsson](https://github.com/bantonsson) have joined the OTel Rust
+project as Approvers.

examples/metrics-advanced/README.md

@@ -12,6 +12,3 @@ Run the following, and the Metrics will be written out to stdout.
 ```shell
 $ cargo run
 ```
-
-
-

examples/metrics-advanced/src/main.rs

@@ -85,15 +85,23 @@ async fn main() -> Result<(), Box<dyn Error + Send + Sync + 'static>> {
         .with_description("My histogram example description")
         .build();
 
-    // Record measurements using the histogram instrument.
-    // This metric will have a cardinality limit of 2,
-    // as set in the view. Because of this, only the first two
-    // measurements will be recorded, and the rest will be folded
-    // into the overflow attribute.
+    // Record measurements using the histogram instrument. This metric will have
+    // a cardinality limit of 2, as set in the view. Because of this, only the
+    // first two distinct attribute combinations will be recorded, and the rest
+    // will be folded into the overflow attribute. Any number of measurements
+    // can be recorded as long as they use the same or already-seen attribute
+    // combinations.
     histogram2.record(1.5, &[KeyValue::new("mykey1", "v1")]);
-
     histogram2.record(1.2, &[KeyValue::new("mykey1", "v2")]);
 
+    // Repeatedly emitting measurements for "v1" and "v2" will not
+    // trigger overflow, as they are already seen attribute combinations.
+    histogram2.record(1.7, &[KeyValue::new("mykey1", "v1")]);
+    histogram2.record(1.8, &[KeyValue::new("mykey1", "v2")]);
+
+    // Emitting measurements for new attribute combinations will trigger
+    // overflow, as the cardinality limit of 2 has been reached.
+    // All the below measurements will be folded into the overflow attribute.
     histogram2.record(1.23, &[KeyValue::new("mykey1", "v3")]);
 
     histogram2.record(1.4, &[KeyValue::new("mykey1", "v4")]);
@@ -104,9 +112,9 @@ async fn main() -> Result<(), Box<dyn Error + Send + Sync + 'static>> {
 
     histogram2.record(1.8, &[KeyValue::new("mykey1", "v7")]);
 
-    // Metrics are exported by default every 30 seconds when using stdout exporter,
+    // Metrics are exported by default every 60 seconds when using stdout exporter,
     // however shutting down the MeterProvider here instantly flushes
-    // the metrics, instead of waiting for the 30 sec interval.
+    // the metrics, instead of waiting for the 60 sec interval.
     meter_provider.shutdown()?;
     Ok(())
 }

examples/metrics-basic/README.md

@@ -11,6 +11,3 @@ Run the following, and the Metrics will be written out to stdout.
 ```shell
 $ cargo run
 ```
-
-
-

examples/metrics-basic/src/main.rs

@@ -136,9 +136,9 @@ async fn main() -> Result<(), Box<dyn Error>> {
         })
         .build();
 
-    // Metrics are exported by default every 30 seconds when using stdout
+    // Metrics are exported by default every 60 seconds when using stdout
     // exporter, however shutting down the MeterProvider here instantly flushes
-    // the metrics, instead of waiting for the 30 sec interval. Shutdown returns
+    // the metrics, instead of waiting for the 60 sec interval. Shutdown returns
     // a result, which is bubbled up to the caller The commented code below
     // demonstrates handling the shutdown result, instead of bubbling up the
     // error.

opentelemetry-appender-log/CHANGELOG.md

@@ -2,6 +2,12 @@
 
 ## vNext
 
+## 0.30.0
+
+Released 2025-May-23
+
+- Updated `opentelemetry` and `opentelemetry-semantic-conventions` dependencies to version 0.30.0.
+
 ## 0.29.0
 
 Released 2025-Mar-21

opentelemetry-appender-log/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "opentelemetry-appender-log"
-version = "0.29.0"
+version = "0.30.0"
 description = "An OpenTelemetry appender for the log crate"
 homepage = "https://github.com/open-telemetry/opentelemetry-rust/tree/main/opentelemetry-appender-log"
 repository = "https://github.com/open-telemetry/opentelemetry-rust/tree/main/opentelemetry-appender-log"
@@ -15,12 +15,12 @@ autobenches = false
 bench = false
 
 [dependencies]
-opentelemetry = { version = "0.29", path = "../opentelemetry", features = [
+opentelemetry = { version = "0.30", path = "../opentelemetry", features = [
   "logs",
 ] }
 log = { workspace = true, features = ["kv", "std"] }
 serde = { workspace = true, optional = true, features = ["std"] }
-opentelemetry-semantic-conventions = { version = "0.29", path = "../opentelemetry-semantic-conventions", optional = true, features = [
+opentelemetry-semantic-conventions = { version = "0.30", path = "../opentelemetry-semantic-conventions", optional = true, features = [
   "semconv_experimental",
 ] }
 

opentelemetry-appender-tracing/CHANGELOG.md

@@ -2,6 +2,13 @@
 
 ## vNext
 
+## 0.30.0
+
+Released 2025-May-23
+
+- Updated `opentelemetry` dependency to version 0.30.0.
+
+
 ## 0.29.1
 
 Released 2025-Mar-24

opentelemetry-appender-tracing/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "opentelemetry-appender-tracing"
-version = "0.29.1"
+version = "0.30.0"
 edition = "2021"
 description = "An OpenTelemetry log appender for the tracing crate"
 homepage = "https://github.com/open-telemetry/opentelemetry-rust/tree/main/opentelemetry-appender-tracing"
@@ -13,12 +13,12 @@ autobenches = false
 
 [dependencies]
 log = { workspace = true, optional = true }
-opentelemetry = { version = "0.29", path = "../opentelemetry", features = ["logs"] }
+opentelemetry = { version = "0.30", path = "../opentelemetry", features = ["logs"] }
 tracing = { workspace = true, features = ["std"]}
 tracing-core = { workspace = true }
 tracing-log = { workspace = true, optional = true }
 tracing-subscriber = { workspace = true, features = ["registry", "std"] }
-tracing-opentelemetry = { workspace = true, optional = true }
+# tracing-opentelemetry = { workspace = true, optional = true }
 
 [dev-dependencies]
 log = { workspace = true }
@@ -37,7 +37,8 @@ pprof = { version = "0.14", features = ["flamegraph", "criterion"] }
 default = []
 experimental_metadata_attributes = ["dep:tracing-log"]
 spec_unstable_logs_enabled = ["opentelemetry/spec_unstable_logs_enabled"]
-experimental_use_tracing_span_context = ["tracing-opentelemetry"]
+# TODO: Enable this back in 0.30.1 after tracing-opentelemetry is released
+# experimental_use_tracing_span_context = ["tracing-opentelemetry"]
 
 
 [[bench]]