feldera · anandbraman · Jun 18, 2026 · Jun 17, 2026
diff --git a/docs.feldera.com/docs/connectors/sources/nats.md b/docs.feldera.com/docs/connectors/sources/nats.md
@@ -6,12 +6,6 @@ with the `nats_input` connector.
 The NATS input connector supports exactly-once [fault
 tolerance](/pipelines/fault-tolerance) using JetStream's ordered pull consumer.
 
-:::warning
-
-NATS support is still experimental, and it may be substantially modified in the future.
-
-:::
-
 ## How it works
 
 The NATS input connector uses JetStream's **ordered pull consumer**, which provides:

diff --git a/docs.feldera.com/docs/formats/json.md b/docs.feldera.com/docs/formats/json.md
@@ -196,8 +196,8 @@ data change event formats in JSON: (1) the raw format and (2) the insert/delete
 ### The insert/delete format
 
 A data change event in this format is represented as a JSON object with a
-single key, which must be equal to `insert` or `delete`.  The associated value
-represents the table row to be inserted or deleted, encoded using the format
+single key, which must be equal to `insert`, `update`, or `delete`.  The associated value
+represents the table row to be inserted, updated, or deleted, encoded using the format
 documented above.  Example row insertion event:
 
 ```json
@@ -210,6 +210,12 @@ Example row deletion event:
 {"delete": {"part": 1, "vendor": 2, "price": 10000}}
 ```
 
+Full or partial updates are also supported if the table has a [primary key](/sql/grammar/#creating-tables). Example update event for a vendors table:
+```json
+{"update": {"vendor_id": 2, "vendor_address": "123 Feldera Place, San Francisco, CA"}}
+```
+The `vendor_id` column must be a primary key. The vendors table may have more columns, but only the `vendor_address` column will be updated by the update event above. An `update` payload that contains all of the table columns is equivalent to an `insert`.
+
 ### The raw format
 
 This format is applicable to append-only event streams where rows can only

diff --git a/docs.feldera.com/docs/pipelines/checkpoint-sync.md b/docs.feldera.com/docs/pipelines/checkpoint-sync.md
@@ -1,9 +1,5 @@
 # Checkpoint Sync to Object Store
 
-:::caution Experimental feature
-Synchronizing checkpoints to object store is a highly experimental feature.
-:::
-
 Feldera can synchronize pipeline checkpoints to an S3-compatible object store
 and restore them at startup. This enables disaster recovery scenarios where
 local storage may be lost, and supports running standby pipelines that can

diff --git a/docs.feldera.com/docs/pipelines/modifying.md b/docs.feldera.com/docs/pipelines/modifying.md
@@ -4,12 +4,6 @@
 This feature is only available in Feldera Enterprise Edition.
 :::
 
-:::warning
-This is an experimental feature and may undergo significant
-changes, including non-backward-compatible modifications, in future releases of
-Feldera.
-:::
-
 ## Overview
 
 Feldera supports **incremental pipeline modification** through a feature called **bootstrapping**.

diff --git a/docs.feldera.com/docs/pipelines/transactions.md b/docs.feldera.com/docs/pipelines/transactions.md
@@ -3,12 +3,6 @@ import TabItem from '@theme/TabItem';
 
 # Efficient Bulk Data Processing using Transactions
 
-:::warning
-Transaction support is an experimental feature and may undergo significant
-changes, including non-backward-compatible modifications in future releases of
-Feldera.
-:::
-
 Transactions enable Feldera pipelines to ingest and process large volumes of data atomically—in
 one logical unit of work rather than piece-by-piece.  Transactions are used to achieve:
 

diff --git a/docs.feldera.com/docs/sql/grammar.md b/docs.feldera.com/docs/sql/grammar.md
@@ -73,7 +73,6 @@ columnConstraint
   :   PRIMARY KEY
   |   FOREIGN KEY REFERENCES identifier '(' identifier ')'
   |   LATENESS expression
-  |   WATERMARK expression
   |   DEFAULT expression
 
 tableConstraint
@@ -272,9 +271,9 @@ latenessStatement
 
 See [Streaming SQL Extensions, LATENESS](streaming.md#lateness-expressions)
 
-### WATERMARKS
+<!-- ### WATERMARKS
 
-See [Streaming SQL Extensions, WATERMARKS](streaming.md#watermark-expressions)
+See [Streaming SQL Extensions, WATERMARKS](streaming.md#watermark-expressions) -->
 
 ## Creating user-defined functions.
 
@@ -545,7 +544,7 @@ hintOption
 
 :::warning
 
-These hints are considered still experimental, and they may change
+These hints are considered experimental, and they may change
 
 :::
 

diff --git a/docs.feldera.com/docs/sql/streaming.md b/docs.feldera.com/docs/sql/streaming.md
@@ -18,7 +18,7 @@ lateness time units out of order.
 See the [Time Series Analysis Guide](/tutorials/time-series#timestamp-columns-and-lateness)
 for details.
 
-## `WATERMARK` expressions
+<!-- ## `WATERMARK` expressions
 
 :::warning
 
@@ -31,7 +31,7 @@ or substantially modified in the future.
 of the input rows by a constant amount of time.
 
 See the [Time Series Analysis Guide](/tutorials/time-series#delaying-inputs-with-watermark)
-for details.
+for details. -->
 
 ## `append_only` tables
 

diff --git a/docs.feldera.com/docs/sql/system.md b/docs.feldera.com/docs/sql/system.md
@@ -5,13 +5,6 @@ built-in.  Their contents is pre-populated by the runtime.
 
 ## `ERROR_VIEW`
 
-:::warning
-
-The `ERROR_VIEW` is still experimental; the schema and name of this
-view may change.
-
-:::
-
 `ERROR_VIEW` is a view that signals runtime errors.  The schema is
 given by the following (illegal) SQL declaration:
 

diff --git a/docs.feldera.com/docs/sql/udf.md b/docs.feldera.com/docs/sql/udf.md
@@ -15,17 +15,10 @@ AS (str LIKE ('%' || CAST(value AS VARCHAR) || '%'));
 CREATE VIEW V0 AS SELECT contains_number(CAST('YES: 10 NO:5' AS VARCHAR), 5)
 ```
 
-## User-defined functions written in Rust
-
-:::warning Experimental feature
-
-Rust UDF support is currently experimental and may undergo significant changes, including
-non-backward-compatible modifications, in future releases of Feldera.
-
-:::
-
 :::danger
 
+## User-defined functions written in Rust
+
 * Feldera's incremental query engine assumes that all computations are deterministic. Using
   a non-deterministic UDF is likely to result in incorrect outputs. The SQL compiler cannot verify
   that Rust functions meet this requirement. Therefore, it is the responsibility of the UDF
@@ -254,10 +247,6 @@ curl -i -X PUT http://127.0.0.1:8080/v0/pipelines/udf_api_test \
 
 ### SQL type representation in Rust
 
-:::caution Experimental feature
-
-:::
-
 The following table shows the Rust representation of standard SQL data
 types.  A nullable SQL type is represented by the corresponding rust
 `Option<>` type.  Notice that some of these types are not standard

diff --git a/docs.feldera.com/docs/tutorials/monitoring/index.md b/docs.feldera.com/docs/tutorials/monitoring/index.md
@@ -70,39 +70,6 @@ A Feldera Instance Monitoring dashboard will be created in Grafana.
 The dashboard is a template and may need to be adjusted to fit your specific needs.
 Look for the `feldera_*` metrics in Grafana to add more metrics to the dashboard.
 
-## Tracing with Jaeger
-
-A Feldera pipeline can be traced using Jaeger. Tracing is useful if you need to analyze
-throughput or latency bottlenecks of your pipeline, as it instruments every step of the
-pipeline execution and provides detailed information about the execution time of each
-step.
-
-### Setup
-
-1. **Jaeger:** You must have [Jaeger installed](https://www.jaegertracing.io).
-2. Start Jaeger using the `jaeger-all-in-one` script:
-    ```bash
-    ./jaeger-all-in-one
-    ```
-3. **Tracing a Feldera pipeline**
-    - Enable/disable tracing and specify the Jaeger endpoint in the pipeline `runtime_config`:
-   ```bash
-   curl -i -X PATCH http://127.0.0.1:8080/v0/pipelines/tracing-pipeline \
-   -H 'Content-Type: application/json' \
-   -d '{
-     "runtime_config": {"tracing": true, "tracing_endpoint_jaeger": "host.docker.internal:6831", <other config settings> }
-   }'
-   ```
-    - Use `host.docker.internal:6831` if you are running Feldera in docker or
-      `127.0.0.1:6831` if you run Feldera directly.
-    - Make sure to specify other settings you changed from non-default values in the `runtime_config` as well.
-      `runtime_config` can be retrieved with:
-   ```bash
-   curl -X 'GET' \
-   'http://127.0.0.1:8080/v0/pipelines/tracing-pipeline' \
-   -H 'accept: application/json'
-   ```
-
 ## DBSP Profiles
 
 A DBSP profile is a graph of the pipeline's circuit where each node represents an

diff --git a/docs.feldera.com/docs/tutorials/time-series.md b/docs.feldera.com/docs/tutorials/time-series.md
@@ -147,8 +147,9 @@ tables and views:
 * **LATENESS does NOT delay computation**.  Feldera does not delay computation
   until all out-of-order records have been received.  It always computes the
   output of the queries given all inputs received so far and incrementally updates
-  these outputs as new out-of-order inputs arrive.  (See [below](#delaying-inputs-with-watermark)
-  for an experimental feture that allows delaying inputs on demand).
+  these outputs as new out-of-order inputs arrive.
+  <!-- (See [below](#delaying-inputs-with-watermark)
+  for an experimental feature that allows delaying inputs on demand). -->
 
 * **Inputs that violate lateness are discarded.**  When a program receives a record
   that is more than lateness time units behind the most recent timestamp value
@@ -399,7 +400,7 @@ The use of `emit_final` is subject to the following restrictions:
   for the view used as output.  If the view is used in defining
   other views, these derived views will receive the non-delayed data.
 
-## Delaying inputs with `WATERMARK`
+<!-- ## Delaying inputs with `WATERMARK`
 
 :::warning
 
@@ -442,7 +443,7 @@ CREATE TABLE purchase_watermark (
 The current version of the SQL compiler does not support multiple
 `WATERMARK` columns in a single table.
 
-:::
+::: -->
 
 ## Append-only tables