Add common guidance on recording errors on spans and metrics, clarify DB conventions (#1716)

Co-authored-by: Trask Stalnaker <[email protected]>

Author: Liudmila Molkova

.chloggen/1716.yaml

@@ -0,0 +1,4 @@
+change_type: enhancement
+component: docs, db
+note: Add common guidance for recording errors on spans and metrics, clarify DB conventions.
+issues: [1516, 1536, 1716]

docs/attributes-registry/exception.md

@@ -6,30 +6,23 @@
 
 # Exception
 
+- [Exception Attributes](#exception-attributes)
+- [Deprecated Exception Attributes](#deprecated-exception-attributes)
+
 ## Exception Attributes
 
 This document defines the shared attributes used to report a single exception associated with a span or log.
 
 | Attribute | Type | Description | Examples | Stability |
 |---|---|---|---|---|
-| <a id="exception-escaped" href="#exception-escaped">`exception.escaped`</a> | boolean | SHOULD be set to true if the exception event is recorded at a point where it is known that the exception is escaping the scope of the span. [1] |  | ![Stable](https://img.shields.io/badge/-stable-lightgreen) |
 | <a id="exception-message" href="#exception-message">`exception.message`</a> | string | The exception message. | `Division by zero`; `Can't convert 'int' object to str implicitly` | ![Stable](https://img.shields.io/badge/-stable-lightgreen) |
 | <a id="exception-stacktrace" href="#exception-stacktrace">`exception.stacktrace`</a> | string | A stacktrace as a string in the natural representation for the language runtime. The representation is to be determined and documented by each language SIG. | `Exception in thread "main" java.lang.RuntimeException: Test exception\n at com.example.GenerateTrace.methodB(GenerateTrace.java:13)\n at com.example.GenerateTrace.methodA(GenerateTrace.java:9)\n at com.example.GenerateTrace.main(GenerateTrace.java:5)` | ![Stable](https://img.shields.io/badge/-stable-lightgreen) |
 | <a id="exception-type" href="#exception-type">`exception.type`</a> | string | The type of the exception (its fully-qualified class name, if applicable). The dynamic type of the exception should be preferred over the static type in languages that support it. | `java.net.ConnectException`; `OSError` | ![Stable](https://img.shields.io/badge/-stable-lightgreen) |
 
-**[1] `exception.escaped`:** An exception is considered to have escaped (or left) the scope of a span,
-if that span is ended while the exception is still logically "in flight".
-This may be actually "in flight" in some languages (e.g. if the exception
-is passed to a Context manager's `__exit__` method in Python) but will
-usually be caught at the point of recording the exception in most languages.
-
-It is usually not possible to determine at the point where an exception is thrown
-whether it will escape the scope of a span.
-However, it is trivial to know that an exception
-will escape, if one checks for an active exception just before ending the span,
-as done in the [example for recording span exceptions](https://opentelemetry.io/docs/specs/semconv/exceptions/exceptions-spans/#recording-an-exception).
-
-It follows that an exception may still escape the scope of the span
-even if the `exception.escaped` attribute was not set or set to false,
-since the event might have been recorded at a time where it was not
-clear whether the exception will escape.
+## Deprecated Exception Attributes
+
+Deprecated exception attributes.
+
+| Attribute | Type | Description | Examples | Stability |
+|---|---|---|---|---|
+| <a id="exception-escaped" href="#exception-escaped">`exception.escaped`</a> | boolean | Indicates that the exception is escaping the scope of the span. |  | ![Deprecated](https://img.shields.io/badge/-deprecated-red)<br>It's no longer recommended to record exceptions that are handled and do not escape the scope of a span. |

docs/cli/cli-spans.md

@@ -13,7 +13,8 @@ Span kind SHOULD be `INTERNAL` when the traced program is the callee or `CLIENT`
 The span name SHOULD be set to `{process.executable.name}`.
 Instrumentations that have additional context about executed commands MAY use a different low-cardinality span name format and SHOULD document it.
 
-Span status SHOULD be set to `Error` if `{process.exit.code}` is not 0.
+Span status SHOULD be set to `Error` if `{process.exit.code}` is not 0. Refer to the [Recording Errors](/docs/general/recording-errors.md) document for
+additional details on how to record span status.
 
 <!-- TODO: context propagation https://github.com/open-telemetry/semantic-conventions/issues/1612 -->
 

docs/database/cassandra.md

@@ -69,8 +69,7 @@ system specific term if more applicable.
 
 **[5] `db.operation.name`:** If readily available and if there is a single operation name that describes the database call. The operation name MAY be parsed from the query text, in which case it SHOULD be the single operation name found in the query.
 
-**[6] `db.response.status_code`:** The status code returned by the database. Usually it represents an error code, but may also represent partial success, warning, or differentiate between various types of successful outcomes.
-Semantic conventions for individual database systems SHOULD document what `db.response.status_code` means in the context of that system.
+**[6] `db.response.status_code`:** All Cassandra protocol error codes SHOULD be considered errors.
 
 **[7] `db.response.status_code`:** If the operation failed and status code is available.
 

docs/database/cosmosdb.md

@@ -193,8 +193,7 @@ additional values when introducing new operations.
 
 **[5] `db.operation.name`:** If readily available and if there is a single operation name that describes the database call. The operation name MAY be parsed from the query text, in which case it SHOULD be the single operation name found in the query.
 
-**[6] `db.response.status_code`:** The status code returned by the database. Usually it represents an error code, but may also represent partial success, warning, or differentiate between various types of successful outcomes.
-Semantic conventions for individual database systems SHOULD document what `db.response.status_code` means in the context of that system.
+**[6] `db.response.status_code`:** Response codes in the 4xx and 5xx range SHOULD be considered errors.
 
 **[7] `error.type`:** The `error.type` SHOULD match the `db.response.status_code` returned by the database or the client library, or the canonical name of exception that occurred.
 When using canonical exception type name, instrumentation SHOULD do the best effort to report the most relevant type. For example, if the original exception is wrapped into a generic one, the original exception SHOULD be preferred.

docs/database/couchdb.md

@@ -23,16 +23,15 @@ The Semantic Conventions for [CouchDB](https://couchdb.apache.org/) extend and o
 |---|---|---|---|---|---|
 | [`db.namespace`](/docs/attributes-registry/db.md) | string | The name of the database, fully qualified within the server address and port. | `customers`; `test.users` | `Conditionally Required` If available. | ![Release Candidate](https://img.shields.io/badge/-rc-mediumorchid) |
 | [`db.operation.name`](/docs/attributes-registry/db.md) | string | The HTTP method + the target REST route. [1] | `GET /{db}/{docid}` | `Conditionally Required` If readily available. | ![Release Candidate](https://img.shields.io/badge/-rc-mediumorchid) |
-| [`db.response.status_code`](/docs/attributes-registry/db.md) | string | The HTTP response code returned by the Couch DB. [2] | `200`; `201`; `429` | `Conditionally Required` [3] | ![Release Candidate](https://img.shields.io/badge/-rc-mediumorchid) |
+| [`db.response.status_code`](/docs/attributes-registry/db.md) | string | The HTTP response code returned by the Couch DB recorded as a string. [2] | `200`; `201`; `429` | `Conditionally Required` [3] | ![Release Candidate](https://img.shields.io/badge/-rc-mediumorchid) |
 | [`error.type`](/docs/attributes-registry/error.md) | string | Describes a class of error the operation ended with. [4] | `timeout`; `java.net.UnknownHostException`; `server_certificate_invalid`; `500` | `Conditionally Required` If and only if the operation failed. | ![Stable](https://img.shields.io/badge/-stable-lightgreen) |
 | [`server.port`](/docs/attributes-registry/server.md) | int | Server port number. [5] | `80`; `8080`; `443` | `Conditionally Required` [6] | ![Stable](https://img.shields.io/badge/-stable-lightgreen) |
 | [`db.operation.batch.size`](/docs/attributes-registry/db.md) | int | The number of queries included in a batch operation. [7] | `2`; `3`; `4` | `Recommended` | ![Release Candidate](https://img.shields.io/badge/-rc-mediumorchid) |
 | [`server.address`](/docs/attributes-registry/server.md) | string | Name of the database host. [8] | `example.com`; `10.1.2.80`; `/tmp/my.sock` | `Recommended` | ![Stable](https://img.shields.io/badge/-stable-lightgreen) |
 
 **[1] `db.operation.name`:** In **CouchDB**, `db.operation.name` should be set to the HTTP method + the target REST route according to the API reference documentation. For example, when retrieving a document, `db.operation.name` would be set to (literally, i.e., without replacing the placeholders with concrete values): [`GET /{db}/{docid}`](https://docs.couchdb.org/en/stable/api/document/common.html#get--db-docid).
 
-**[2] `db.response.status_code`:** The status code returned by the database. Usually it represents an error code, but may also represent partial success, warning, or differentiate between various types of successful outcomes.
-Semantic conventions for individual database systems SHOULD document what `db.response.status_code` means in the context of that system.
+**[2] `db.response.status_code`:** HTTP response codes in the 4xx and 5xx range SHOULD be considered errors.
 
 **[3] `db.response.status_code`:** If response was received and the HTTP response code is available.
 

docs/database/database-spans.md

@@ -12,7 +12,6 @@ linkTitle: Client Calls
 
 - [Name](#name)
 - [Status](#status)
-  - [Recording exception events](#recording-exception-events)
 - [Common attributes](#common-attributes)
   - [Notes and well-known identifiers for `db.system`](#notes-and-well-known-identifiers-for-dbsystem)
 - [Sanitization of `db.query.text`](#sanitization-of-dbquerytext)
@@ -89,59 +88,11 @@ For example, for an operation describing SQL query on an anonymous table like `S
 
 ## Status
 
-[Span Status Code][SpanStatus] MUST be left unset if the operation has ended without any errors.
+Refer to the [Recording Errors](/docs/general/recording-errors.md) document for
+details on how to record span status.
 
-Instrumentation SHOULD consider the operation as failed if any of the following is true:
-
-- the `db.response.status_code` value indicates an error
-
-  > [!NOTE]
-  >
-  > The classification of status code as an error depends on the context.
-  > For example, a SQL STATE `02000` (`no_data`) indicates an error when the application
-  > expected the data to be available. However, it is not an error when the
-  > application is simply checking whether the data exists.
-  >
-  > Instrumentations that have additional context about a specific operation MAY use
-  > this context to set the span status more precisely.
-  > Instrumentations that don't have any additional context MUST follow the
-  > guidelines in this section.
-
-- an exception is thrown by the instrumented method call
-- the instrumented method returns an error in another way
-
-When the operation ends with an error, instrumentation:
-
-- SHOULD set the span status code to `Error`
-- SHOULD set the `error.type` attribute
-- SHOULD set the span status description when it has additional information
-  about the error which is not expected to contain sensitive details and aligns
-  with [Span Status Description][SpanStatus] definition.
-
-  It's NOT RECOMMENDED to duplicate `db.response.status_code` or `error.type`
-  in span status description.
-
-  When the operation fails with an exception, the span status description SHOULD be set to
-  the exception message.
-
-### Recording exception events
-
-**Status**: [Experimental][DocumentStatus]
-
-When the operation fails with an exception, instrumentation SHOULD record
-an [exception event](../exceptions/exceptions-spans.md) by default if, and only if,
-the span being recorded is a local root span (does not have a local parent).
-
-> [!NOTE]
->
-> Exception stack traces could be very long and are expensive to capture and store.
-> Exceptions which are not handled by instrumented libraries are likely to be handled
-> and logged by the caller.
-> Exceptions that are not handled will be recorded by the outermost (local root)
-> instrumentation such as HTTP or gRPC server.
-
-Instrumentation MAY provide a configuration option to record exceptions that
-escape the surface of the instrumented API.
+Semantic conventions for individual systems SHOULD specify which values of `db.response.status_code`
+classify as errors.
 
 ## Common attributes
 
@@ -466,4 +417,3 @@ More specific Semantic Conventions are defined for the following database techno
 * [SQL](sql.md): Semantic Conventions for *SQL* databases.
 
 [DocumentStatus]: https://opentelemetry.io/docs/specs/otel/document-status
-[SpanStatus]: https://github.com/open-telemetry/opentelemetry-specification/tree/v1.39.0/specification/trace/api.md#set-status

docs/database/elasticsearch.md

@@ -82,8 +82,7 @@ When a query string value is redacted, the query string key SHOULD still be pres
 
 **[4] `db.elasticsearch.path_parts`:** Many Elasticsearch url paths allow dynamic values. These SHOULD be recorded in span attributes in the format `db.elasticsearch.path_parts.<key>`, where `<key>` is the url path part name. The implementation SHOULD reference the [elasticsearch schema](https://raw.githubusercontent.com/elastic/elasticsearch-specification/main/output/schema/schema.json) in order to map the path part values to their names.
 
-**[5] `db.response.status_code`:** The status code returned by the database. Usually it represents an error code, but may also represent partial success, warning, or differentiate between various types of successful outcomes.
-Semantic conventions for individual database systems SHOULD document what `db.response.status_code` means in the context of that system.
+**[5] `db.response.status_code`:** HTTP response codes in the 4xx and 5xx range SHOULD be considered errors.
 
 **[6] `error.type`:** The `error.type` SHOULD match the `db.response.status_code` returned by the database or the client library, or the canonical name of exception that occurred.
 When using canonical exception type name, instrumentation SHOULD do the best effort to report the most relevant type. For example, if the original exception is wrapped into a generic one, the original exception SHOULD be preferred.

docs/database/hbase.md

@@ -24,7 +24,7 @@ The Semantic Conventions for [HBase](https://hbase.apache.org/) extend and overr
 | [`db.collection.name`](/docs/attributes-registry/db.md) | string | The HBase table name. [1] | `mytable`; `ns:table` | `Conditionally Required` If applicable. | ![Release Candidate](https://img.shields.io/badge/-rc-mediumorchid) |
 | [`db.namespace`](/docs/attributes-registry/db.md) | string | The HBase namespace. [2] | `mynamespace` | `Conditionally Required` If applicable. | ![Release Candidate](https://img.shields.io/badge/-rc-mediumorchid) |
 | [`db.operation.name`](/docs/attributes-registry/db.md) | string | The name of the operation or command being executed. [3] | `findAndModify`; `HMSET`; `SELECT` | `Conditionally Required` If readily available. | ![Release Candidate](https://img.shields.io/badge/-rc-mediumorchid) |
-| [`db.response.status_code`](/docs/attributes-registry/db.md) | string | Protocol-specific response code recorded as string. [4] | `200`; `409`; `14` | `Conditionally Required` If response was received. | ![Release Candidate](https://img.shields.io/badge/-rc-mediumorchid) |
+| [`db.response.status_code`](/docs/attributes-registry/db.md) | string | Protocol-specific response code recorded as a string. [4] | `200`; `409`; `14` | `Conditionally Required` If response was received. | ![Release Candidate](https://img.shields.io/badge/-rc-mediumorchid) |
 | [`error.type`](/docs/attributes-registry/error.md) | string | Describes a class of error the operation ended with. [5] | `timeout`; `java.net.UnknownHostException`; `server_certificate_invalid`; `500` | `Conditionally Required` If and only if the operation failed. | ![Stable](https://img.shields.io/badge/-stable-lightgreen) |
 | [`server.port`](/docs/attributes-registry/server.md) | int | Server port number. [6] | `80`; `8080`; `443` | `Conditionally Required` [7] | ![Stable](https://img.shields.io/badge/-stable-lightgreen) |
 | [`db.operation.batch.size`](/docs/attributes-registry/db.md) | int | The number of queries included in a batch operation. [8] | `2`; `3`; `4` | `Recommended` | ![Release Candidate](https://img.shields.io/badge/-rc-mediumorchid) |

docs/database/mariadb.md

@@ -42,41 +42,9 @@ Instrumentation SHOULD document if `db.namespace` reflects the database provided
 
 It is RECOMMENDED to capture the value as provided by the application without attempting to do any case normalization.
 
-**[2] `db.response.status_code`:** SQL defines [SQLSTATE](https://wikipedia.org/wiki/SQLSTATE) as a database
-return code which is adopted by some database systems like PostgreSQL.
-See [PostgreSQL error codes](https://www.postgresql.org/docs/current/errcodes-appendix.html)
-for the details.
-
-Other systems like MySQL, Oracle, or MS SQL Server define vendor-specific
-error codes. Database SQL drivers usually provide access to both properties.
-For example, in Java, the [`SQLException`](https://docs.oracle.com/javase/8/docs/api/java/sql/SQLException.html)
-class reports them with `getSQLState()` and `getErrorCode()` methods.
-
-Instrumentations SHOULD populate the `db.response.status_code` with the
-the most specific code available to them.
-
-Here's a non-exhaustive list of databases that report vendor-specific
-codes with granularity higher than SQLSTATE (or don't report SQLSTATE
-at all):
-
-- [DB2 SQL codes](https://www.ibm.com/docs/db2-for-zos/12?topic=codes-sql).
-- [Maria DB error codes](https://mariadb.com/kb/en/mariadb-error-code-reference/)
-- [Microsoft SQL Server errors](https://docs.microsoft.com/sql/relational-databases/errors-events/database-engine-events-and-errors)
-- [MySQL error codes](https://dev.mysql.com/doc/mysql-errors/9.0/en/error-reference-introduction.html)
-- [Oracle error codes](https://docs.oracle.com/cd/B28359_01/server.111/b28278/toc.htm)
-- [SQLite result codes](https://www.sqlite.org/rescode.html)
-
-These systems SHOULD set the `db.response.status_code` to a
-known vendor-specific error code. If only SQLSTATE is available,
-it SHOULD be used.
-
-When multiple error codes are available and specificity is unclear,
-instrumentation SHOULD set the `db.response.status_code` to the
-concatenated string of all codes with '/' used as a separator.
-
-For example, generic DB instrumentation that detected an error and has
-SQLSTATE `"42000"` and vendor-specific `1071` should set
-`db.response.status_code` to `"42000/1071"`."
+**[2] `db.response.status_code`:** MariaDB uses vendor-specific error codes on all errors and reports [SQLSTATE](https://mariadb.com/kb/en/sqlstate/) in some cases.
+MariaDB error codes are more granular than SQLSTATE, so MariaDB instrumentations SHOULD set the `db.response.status_code` to this known error code.
+When SQLSTATE is available, SQLSTATE of "Class 02" or higher SHOULD be considered errors. When SQLSTATE is not available, all MariaDB error codes SHOULD be considered errors.
 
 **[3] `error.type`:** The `error.type` SHOULD match the `db.response.status_code` returned by the database or the client library, or the canonical name of exception that occurred.
 When using canonical exception type name, instrumentation SHOULD do the best effort to report the most relevant type. For example, if the original exception is wrapped into a generic one, the original exception SHOULD be preferred.