Str:::lab Studio

A zero-dependency, modular browser SQL Studio for Flink SQL — built for engineers who want a real query interface without leaving their laptop.

Apache Flink is a trademark of the Apache Software Foundation. Str:::lab Studio is an independent open-source project that uses the Flink SQL Gateway REST API. It is not affiliated with or endorsed by the Apache Software Foundation.

⭐ Support This Project

If you find Str:::lab Studio useful, please consider:

Starring this repository on GitHub
Starring it on Docker Hub
Sharing it with your colleagues and team

Your support helps drive continued development and maintenance. Thank you! 🙏

What is it?

Str:::lab Studio is a self-hosted web IDE that connects to the Flink SQL Gateway and lets you:

Write and run Flink SQL in a multi-tab editor with live streaming results
Upload Java, Python, or Scala JARs and register custom UDFs (ScalarFunction, TableFunction, AggregateFunction) directly from the browser — no SSH, no CLI
Build SQL Views with computed columns and CASE WHEN expressions using the visual View Builder — no JAR needed
Build streaming pipelines visually using the ◈ Pipeline Manager — drag operators onto a canvas, connect them, and submit to Flink without writing SQL by hand
Register and test external system connections using the ⊙ Systems Manager — configure Kafka, PostgreSQL, Elasticsearch, MinIO, Hive Metastore, and more with a ⊙ Test Connectivity check before saving
Register persistent external catalogs using the ⊕ Catalog Manager — JDBC, Hive, Iceberg (Hive/REST/Glue), and Delta Lake catalogs with live connectivity testing
Visualise running job DAGs with live operator metrics, backpressure indicators, and throughput charts
Plot streaming results as bar, line, area, scatter, pie, donut, histogram, or heatmap charts with explicit X/Y axis selection
Highlight result rows in real time using Colour Describe — a rules engine that applies colour to matching rows as they stream in
Upload and Submit packaged Flink application written and compiled as jar
Manage multiple sessions — each with its own isolated workspace (tabs, logs, history, jobs, UDFs)
Monitor cluster health: backpressure, checkpoints, slot utilisation, JVM heap, records/s per operator
Use Admin Session for full cluster visibility, cross-session oversight, and pipeline inspection
Generate PDF reports — standard session reports, or admin-grade Technical / Business reports
Organise work as named Projects — save, load, run, and export your pipelines
Compatible with Flink 1.16 through 2.x — including the Flink 2.0 release

Quickstart Cluster Setup

Need a Flink cluster to test with? We provide a ready-to-use Docker Compose setup that includes:

Flink JobManager and TaskManagers (2 Task managers(10 slots per manager), 1 Jobmanager) (v1.19). customizable
SQL Gateway with pre-downloaded connectors (Kafka, JDBC, Elasticsearch, etc.)
Nginx for CORS handling
All connectors used by Str:::lab Studio pre-installed

git clone https://github.com/coded-streams/FLINK-CLUSTER.git
cd FLINK-CLUSTER
docker compose up -d

Note: This cluster setup uses Flink 1.19. You can easily update it to newer Flink versions by changing the image tags in docker-compose.yml. The setup includes scripts to pre-download all connectors supported by Str:::lab Studio, making it the perfect companion for testing and development.

What do I need to run this?

The only hard requirement is a running Flink SQL Gateway. The Studio talks to the Gateway REST API exclusively — it does not connect directly to Kafka, ZooKeeper, or any other infrastructure.

What you have	What to do
Nothing yet	`git clone` + `docker compose up -d` — starts everything
Flink cluster, no SQL Gateway	Add `flink-sql-gateway` to your compose (snippet in Option 3)
Flink cluster + SQL Gateway	Add Studio only — Option 2
Cloud Flink (Confluent, Ververica, AWS)	Option 5 — Direct Gateway or Remote mode with token
Kubernetes	Option 6 — Helm operator

No CORS proxy required. The Studio image includes nginx which proxies all browser requests to the gateway and adds CORS headers itself. A separate CORS proxy (flink-gateway-cors-proxy) is provided in the repo but is optional — you only need it if you want browsers to call the gateway directly on port 8084 without going through the Studio.

Quickstart

Option 1 — Start everything from scratch

git clone https://github.com/coded-streams/strlabstudio
cd strlabstudio
docker compose up -d
open http://localhost:3030

Starts: JobManager, TaskManagers, SQL Gateway, Studio, and an optional CORS proxy. Select Via Studio on the connect screen — done.

Option 2 — I already have a Flink cluster and SQL Gateway

Add Studio to your existing docker-compose.yml. Point it at your gateway and jobmanager. No CORS proxy needed. Via Studio mode routes through the Studio nginx which handles CORS.

services:

  flink-studio:
    image: codedstreams/strlabstudio:latest
    container_name: flink-studio
    restart: unless-stopped
    ports:
      - "3030:80"
    environment:
      FLINK_GATEWAY_HOST: flink-sql-gateway      # your gateway container name
      FLINK_GATEWAY_PORT: "8083"
      JOBMANAGER_HOST:    your-jobmanager         # your jobmanager container name
      JOBMANAGER_PORT:    "8081"
    volumes:
      - udf-jars:/var/www/udf-jars               # required for UDF JAR upload
    networks:
      - your-network                             # same network as your cluster

volumes:
  udf-jars:

networks:
  your-network:
    external: true                               # joins your existing network

For UDF JAR upload also add - udf-jars:/var/www/udf-jars to your flink-sql-gateway volumes. See the UDF JAR Upload section.

docker compose up -d flink-studio
open http://localhost:3030

Option 3 — I have a Flink cluster but no SQL Gateway

The SQL Gateway is the API layer between the Studio and your Flink cluster. Add it alongside the Studio:

services:

  flink-sql-gateway:
    image: flink:1.19.1-scala_2.12-java11
    container_name: flink-sql-gateway
    depends_on:
      your-jobmanager:
        condition: service_healthy
    command: >
      /bin/bash -c "
        cp /opt/flink/plugins/connectors/*.jar /opt/flink/lib/ 2>/dev/null || true &&
        exec /opt/flink/bin/sql-gateway.sh start-foreground
      "
    environment:
      FLINK_PROPERTIES: |
        jobmanager.rpc.address: your-jobmanager
        sql-gateway.endpoint.rest.address: 0.0.0.0
        sql-gateway.endpoint.rest.port: 8083
    volumes:
      - ./connectors:/opt/flink/plugins/connectors
      - udf-jars:/var/www/udf-jars
    networks:
      - your-network
    healthcheck:
      test: ["CMD-SHELL", "wget -qO- http://localhost:8083/v1/info || exit 1"]
      interval: 15s
      timeout: 10s
      retries: 20
      start_period: 45s

  flink-studio:
    image: codedstreams/strlabstudio:latest
    container_name: flink-studio
    depends_on:
      flink-sql-gateway:
        condition: service_healthy
    ports:
      - "3030:80"
    environment:
      FLINK_GATEWAY_HOST: flink-sql-gateway
      FLINK_GATEWAY_PORT: "8083"
      JOBMANAGER_HOST:    your-jobmanager
      JOBMANAGER_PORT:    "8081"
    volumes:
      - udf-jars:/var/www/udf-jars
    networks:
      - your-network

volumes:
  udf-jars:

Put your Kafka/Flink connector JARs in ./connectors/. They are copied to /opt/flink/lib/ at Gateway startup.

Option 4 — Standalone `docker run`

docker run -d -p 3030:80 \
  --name flink-studio \
  --network <your-flink-network> \
  -e FLINK_GATEWAY_HOST=flink-sql-gateway \
  -e FLINK_GATEWAY_PORT=8083 \
  -e JOBMANAGER_HOST=your-jobmanager \
  -e JOBMANAGER_PORT=8081 \
  codedstreams/strlabstudio:latest

--network is required for Via Studio mode. Without it, nginx cannot resolve the gateway hostname by container name and returns HTTP 500. Use Direct Gateway mode if you cannot join the network.

UDF JAR upload requires a shared Docker named volume between Studio and the gateway. Use docker compose if you need that feature.

Option 5 — Cloud / Remote Flink

Run Studio anywhere and point it at your remote gateway:

docker run -d -p 3030:80 codedstreams/strlabstudio:latest
open http://localhost:3030

On the connect screen select Remote / Cloud, enter your gateway URL (e.g. https://your-gateway.example.com), and provide your Bearer token or Basic auth credentials.

Works with: Confluent Cloud, Ververica Platform, Amazon Managed Service for Apache Flink, or any self-hosted gateway with a public endpoint.

Option 6 — Kubernetes

helm repo add strlabstudio https://coded-streams.github.io/strlabstudio-operator/charts
helm install strlabstudio-operator strlabstudio/strlabstudio-operator \
  --namespace flinksql-system --create-namespace

Apply a StrlabStudio custom resource and the operator manages the Deployment, Service, and PVC:

apiVersion: codedstreams.io/v1alpha1
kind: StrlabStudio
metadata:
  name: my-studio
  namespace: flink
spec:
  image: codedstreams/strlabstudio:latest
  gateway:
    host: flink-sql-gateway
    port: 8083
  jobmanager:
    host: flink-jobmanager
    port: 8081
  service:
    type: ClusterIP
    port: 80

How the Studio connects — architecture

Browser (http://localhost:3030)
         │
         ▼
┌─────────────────────────────────────────────────────────────────┐
│  flink-studio  (codedstreams/strlabstudio:latest)               │
│                                                                 │
│  nginx serves the IDE at port 80                                │
│  docker-entrypoint.sh writes nginx.conf at container startup    │
│                                                                 │
│  /flink-api/*       → proxy → FLINK_GATEWAY_HOST:PORT           │
│                               + CORS headers added by nginx     │
│  /jobmanager-api/*  → proxy → JOBMANAGER_HOST:PORT              │
│  /udf-jars/         → WebDAV PUT → /var/www/udf-jars/           │
└─────────────────────────────────────────────────────────────────┘
         │
         ├─── flink-sql-gateway:8083      SQL sessions, statements, results
         │           └─── your-jobmanager:8081   job submission
         │                     ├── taskmanager-1
         │                     └── taskmanager-2
         │
         └─── [optional] flink-gateway-cors-proxy:8084
                    Only for Direct Gateway mode — not required for Via Studio

Connection Modes

Mode	When to use	How it works
Via Studio	Docker Compose or Kubernetes — Studio on same network as cluster	Browser → Studio nginx → SQL Gateway. Same-origin. No CORS issues.
Direct Gateway	Studio not on same network, or browser → gateway directly	Browser → `flink-gateway-cors-proxy:8084`. Needs CORS proxy.
Remote / Cloud	Confluent Cloud, Ververica, AWS, any remote cluster	Browser → your cloud gateway URL. Supports Bearer token + Basic auth.
🛡 Admin Session	Platform operators — full cluster visibility	Any of the above + admin passcode.

Environment Variables

Variable	Default	Description
`FLINK_GATEWAY_HOST`	`localhost`	Container name or hostname of the Flink SQL Gateway
`FLINK_GATEWAY_PORT`	`8083`	SQL Gateway REST port
`JOBMANAGER_HOST`	`localhost`	Container name or hostname of the Flink JobManager
`JOBMANAGER_PORT`	`8081`	JobManager REST API port

Substituted into nginx.conf at container startup by docker-entrypoint.sh.

What is the CORS proxy and do I need it?

You do NOT need it if you use Via Studio connection mode (the default). The Studio nginx proxies all requests to the gateway and adds CORS headers itself. This covers the vast majority of use cases.

You need it only if you want a browser tab to call the SQL Gateway directly on port 8084 without going through the Studio (Direct Gateway mode).

On Kubernetes — CORS is not relevant. In-cluster traffic is same-origin. The proxy is never deployed by the operator.

Pipeline Manager

The Pipeline Manager is a visual Flink SQL pipeline builder built into the Studio. Open it with ◈ Pipeline in the topbar.

What it does:

Drag operators from a categorised palette onto an infinite canvas
Connect operators with typed edges (FORWARD, HASH, BROADCAST, REBALANCE, RESCALE)
Configure each operator with a modal — the config generates the exact SQL WITH (...) clause
Animate data flow with particle physics running along Bézier edges
Validate the pipeline before submission — detects unconfigured nodes, disconnected transforms, and sourceless sinks
Generate and submit the full CREATE TABLE + INSERT INTO SQL in one click
Save, load, export (JSON), and import pipelines across sessions

Operator groups:

Group	Operators
Sources	Kafka, Datagen, JDBC, Filesystem, Pulsar, Kinesis
Transformations	Filter, Project, UDF Map, Enrich, Union, Split
Windows	Tumble, Hop, Session, Cumulate
Aggregations	Aggregate, Dedup, Top-N
Joins	Interval Join, Temporal Join, Regular Join
CEP	Match Recognize, CEP Alert
Sinks	Kafka, JDBC, Filesystem, Elasticsearch, Print, Blackhole, MongoDB
Output	Result Output, AI Model
My UDFs	UDF Node (any registered UDF)

Key behaviours:

Operators that require connector JARs are flagged with a ⚡ badge and trigger a Systems Manager warning when dropped
Print and Blackhole sinks use CREATE TABLE ... WITH (...) LIKE source (EXCLUDING ALL) — they inherit the exact schema from the upstream source automatically, avoiding column mismatch errors
Filter and Project operators have no table_name — they inject a WHERE clause and SELECT column list respectively into the generated INSERT INTO
Multiple sinks generate EXECUTE STATEMENT SET BEGIN ... END with one INSERT INTO per sink
Auto-layout uses Kahn's topological sort to arrange nodes left-to-right by data flow layer

Systems Manager

The Systems Manager centralises connector JAR management and external system integration. Open it with ⊙ Systems in the topbar.

Tabs:

📦 Connector JARs — browse all supported connectors with Maven coordinates, version notes, and SQL examples
⬆ Upload JAR — drag-and-drop connector JARs; Studio saves them via WebDAV PUT to /opt/flink/lib/ (requires shared volume)
⊙ Integrations — configure named connections to Kafka, PostgreSQL, MySQL, Elasticsearch, MinIO/S3, Schema Registry, and Hive Metastore. Each form has a ⊙ Test Connectivity button that probes the service before saving
💾 Saved — browse saved integrations; load them back into the form or insert their generated SQL directly into the editor
? Guide — deployment cheatsheet and connectivity test explanation

⊙ Test Connectivity behaviour by system type:

System	Test method
Elasticsearch, MinIO/S3, Schema Registry	Direct browser `fetch()` — shows version/status if reachable
Kafka, PostgreSQL, MySQL, Hive Metastore	Probes Flink cluster `/v1/info`; provides exact `nc -zv host port` command for container-side testing

Saved integrations are stored in localStorage and appear as a prefill banner in Pipeline Manager JDBC Sink nodes — one click fills all connection fields.

Catalog Manager

The Catalog Manager registers persistent external catalogs in the active Flink SQL Gateway session. Open it with ⊕ Catalogs in the topbar.

Supported catalog types:

Type	Backend	JAR required
Generic In-Memory	Flink built-in	No
PostgreSQL (JDBC)	PostgreSQL via JDBC	flink-connector-jdbc + postgresql driver
MySQL / MariaDB (JDBC)	MySQL via JDBC	flink-connector-jdbc + mysql-connector-j
Apache Hive	Hive Metastore	flink-connector-hive
Apache Iceberg (Hive)	Iceberg + Hive Metastore	iceberg-flink-runtime
Apache Iceberg (REST)	Nessie, Polaris, Tabular, Gravitino	iceberg-flink-runtime
AWS Glue (Iceberg)	AWS Glue Data Catalog	iceberg-flink-runtime + iceberg-aws-bundle
Delta Lake	Delta Lake storage	delta-flink + delta-standalone

Tabs:

⊕ Create Catalog — select type, fill form, run ⊙ Test Connectivity, click ⚡ Create Catalog. Generated CREATE CATALOG SQL is previewed in real time as you type.
◎ Active Catalogs — lists all catalogs in the current session; click USE to switch active catalog or Drop to remove
🕑 History — last 20 catalogs created; Insert or Copy their SQL
📖 Setup Guide — JAR placement instructions and connectivity test explanation

After creation, the catalog appears in the Studio sidebar catalog tree. Tables and columns are live — click a column name to insert it at the cursor in the SQL editor.

UDF JAR Upload

You are not limited to what Flink SQL ships with. Upload your own Java, Python, or Scala functions directly from the browser.

How it works:

Open ⨍ UDFs → ⬆ Upload JAR — drag your shaded JAR and click Upload
Studio nginx saves the file to /var/www/udf-jars/ via WebDAV PUT
Studio runs ADD JAR '/var/www/udf-jars/yourjar.jar' in the active Gateway session
Go to ＋ Register UDF → fill in class path, method, parameter types, scope → Execute Registration
Call your function in any SELECT query against a live streaming source

Supported UDF types: ScalarFunction · TableFunction · AggregateFunction

Requirement: the udf-jars named volume must be mounted on both flink-studio and flink-sql-gateway at the same path:

flink-studio:
  volumes:
    - udf-jars:/var/www/udf-jars

flink-sql-gateway:
  volumes:
    - udf-jars:/var/www/udf-jars

volumes:
  udf-jars:

ADD JAR runs inside the Gateway JVM and reads from the gateway container's own filesystem. The shared volume makes /var/www/udf-jars/ identical on both containers.

UDF deployment paths

Environment	JAR path
Docker (Studio + Gateway compose)	`/var/www/udf-jars/yourjar.jar`
Kubernetes (operator-provisioned PVC)	`/opt/flink/usrlib/yourjar.jar`
Cloud / remote cluster	Upload to cluster node, use absolute path

View Builder

No JAR? No problem. Open ⨍ UDFs → View Builder to create a TEMPORARY VIEW with computed columns, CASE WHEN expressions, and optional WHERE filters — entirely in SQL. The Studio generates and runs the DDL automatically.

Chart Report

Plot your streaming results without leaving the IDE.

Click 📊 Chart Report in the results toolbar
Select the query slot to visualise
Choose your X axis (categories, timestamps, asset names — any column)
Add one or more Y axis fields (scores, counts, aggregates)
Pick a chart type: Bar · Line · Area · Scatter · Pie · Donut · Histogram · Heatmap
Charts refresh live every 2 seconds as new rows arrive
Export as PDF in one click

Colour Describe

A rules engine for your result table. Rows highlight in real time as they stream in — no code required.

Click 🎨 Colour Describe in the results toolbar
Select the live query slot to apply highlighting to
Build rules: pick a field, operator, and value
- Operators: == != > >= < <= contains starts with ends with regex
Choose a highlight colour and style: row background · left border accent · text colour
Click ⚡ Apply & Activate — matching rows highlight immediately and continue as rows stream in

Rules are evaluated top-to-bottom. First match per row wins. A colour legend appears below the table. Toggle off at any time to clear all highlighting.

Job Graph & Resource Monitoring

Open the Job Graph tab while a pipeline is running to see:

Live operator DAG with SOURCE / PROCESS / SINK node classification
Records in/out per second, backpressure %, and parallelism per node
Animated edges showing data flow direction and shipping strategy
Fault highlighting with error message overlay on failed vertices
Zoom, pan, and drag to navigate large graphs

Double-click any operator node to open a drill-down modal:

Metrics grid: records/s, bytes/s, backpressure, duration, subtask count
Subtask table: per-subtask status, host, and record counts
All Metrics tab: full live metric list fetched from the JobManager API
Live Events stream: continuous throughput polling with a mini throughput chart

Performance Benchmarking

The Performance tab tracks every query and pipeline submitted in your session:

Execution time (ms) per query
Row throughput across the session
Job comparison chart — plot multiple jobs side by side on any metric
Per-job checkbox toggles to show/hide individual jobs from the comparison

Admin Session

Connect using the 🛡 Admin button. Default passcode: admin1234 — change it immediately via the 🛡 badge in the topbar after first login.

Feature	Regular Session	Admin Session
Jobs visible	Own session only	All cluster jobs
Cancel job	Own jobs only	Any job
Session Inspector	—	Full cross-session breakdown
Audit trail	—	Timestamped log of admin actions
Report type	Standard session report	Technical or Business/Management PDF

Report Generation

Generate a formatted PDF report from any result set directly from the results toolbar.

Options:

Row range filter (e.g. rows 1–500)
Value filter — search across all columns
Custom report title
Automatic field descriptions for known column patterns
Session metadata: catalog, database, session ID, parallelism, job name
Colour Describe rules included in the report output

Reports are designed to be shared with both business stakeholders (Business/Management PDF) and technical reviewers (Technical PDF via Admin Session).

Flink 2.0 Compatibility

Str:::lab Studio is compatible with Flink 1.16 through 2.x — including the Flink 2.0.0 release (March 2025).

The Studio connects exclusively via the SQL Gateway REST API, which is preserved and enhanced in Flink 2.0. All session management, ADD JAR, CREATE FUNCTION, and streaming SQL features work identically.

Things to update when moving to Flink 2.0:

Area	Change
Config file	`flink-conf.yaml` → `config.yaml` (strict YAML)
Connector JARs	Use `3.4.x-2.0` series — the `3.3.x-1.x` series will not work
Java version	Java 8 dropped — use Java 11+ on your cluster
DataSet / Scala APIs	Removed — no impact on Studio (SQL only)
State compatibility	Not cross-version — take a savepoint before upgrading
ML_PREDICT (2.1+)	New built-in SQL function for ML inference — callable from any Studio tab

Connector JARs

Connector JARs must match your Flink version exactly. Wrong version → NoClassDefFoundError on every INSERT job.

Flink version	Kafka connector JAR
2.0.x / 2.1.x	Check Maven Central for latest `*-2.0.jar`
1.20.x	`flink-sql-connector-kafka-3.3.0-1.20.jar`
1.19.x ← recommended	`flink-sql-connector-kafka-3.3.0-1.19.jar`
1.18.x	`flink-sql-connector-kafka-3.3.0-1.18.jar`
1.17.x	`flink-sql-connector-kafka-3.2.0-1.17.jar`

Place JARs in ./connectors/. The compose file copies them to /opt/flink/lib/ at Gateway startup.

Recommended session setup

Available as a snippet: CONFIG → ⚡ Recommended Streaming Config

SET 'execution.runtime-mode'              = 'streaming';
SET 'parallelism.default'                 = '2';
SET 'execution.checkpointing.interval'    = '10000';
SET 'execution.checkpointing.mode'        = 'EXACTLY_ONCE';
SET 'table.exec.state.ttl'                = '3600000';
SET 'table.exec.source.idle-timeout'      = '10000';
SET 'table.exec.mini-batch.enabled'       = 'true';
SET 'table.exec.mini-batch.allow-latency' = '500 ms';
SET 'table.exec.mini-batch.size'          = '5000';
SET 'table.optimizer.agg-phase-strategy'  = 'TWO_PHASE';

📚 Additional Resources

Resource	Link
Docker Hub	codedstreams/strlabstudio
Official Website	coded-streams.github.io/strlabstudio
Kubernetes Operator	codedstreams/strlab-kube-operator
Quickstart Cluster	FLINK-CLUSTER repo — Pre-configured Flink 1.19 with all connectors
GitHub Repository	coded-streams/strlabstudio

Changelog

v1.3.2

Pipeline Manager v2.0 — Visual Flink SQL pipeline builder. 35+ operators across 9 groups (Sources, Transformations, Windows, Aggregations, Joins, CEP, Sinks, Output, My UDFs). Operators rendered as SVG shapes (rect, diamond, hexagon, circle, stadium, parallelogram). Bézier edges with typed shipping strategies (FORWARD, HASH, BROADCAST, REBALANCE, RESCALE). ✏ edit button on every node for reliable config access. Particle animation along edges on ▶ Run. Kahn's topological sort for auto-layout and SQL generation order. EXECUTE STATEMENT SET BEGIN...END for multi-sink pipelines. Print and Blackhole sinks use LIKE source (EXCLUDING ALL) to inherit schema automatically.
Systems Manager v1.1 — Connector JAR browser and integration registry. Manages Kafka, JDBC, S3/MinIO, Elasticsearch, Iceberg, Hive, Redis, Datagen, and Blackhole connectors. ⊙ Test Connectivity button on every integration form. HTTP-accessible services probed directly from the browser; TCP services probed via Flink REST with nc -zv command provided. Saved integrations prefill JDBC Sink nodes in Pipeline Manager.
Catalog Manager v1.1 — External catalog registration for JDBC (PostgreSQL, MySQL), Hive, Iceberg (Hive/REST/Glue), and Delta Lake. ⊙ Test Connectivity before catalog creation. Generated CREATE CATALOG SQL previewed in real time. History of last 20 created catalogs. Live catalog tree in sidebar after creation.
AI Model operator — Pipeline Manager output operator for OpenAI, Anthropic, Bedrock, Vertex AI, Azure OpenAI, Hugging Face, and Custom HTTP endpoints. Auth types: Bearer, AWS SigV4, Azure AD, API Key Header, Basic Auth.
Kafka auth fields — SASL/SSL, SASL_PLAINTEXT, Confluent Cloud, and Schema Registry credentials on both Kafka Source and Kafka Sink operators.
JDBC auth fields — username, password, driver class, and fetch size on JDBC Source and Sink operators.
Print/Blackhole schema fix — fixed column mismatch validation error (Column types of query result and sink do not match) by switching from hardcoded fallback schema to LIKE source (EXCLUDING ALL).
Four standalone pipeline demo pages — step-by-step interactive guides: Datagen fanout, Kafka producer + consumer, Systems Manager connector → pipeline, and Catalog Manager → pipeline.

v1.3.0

UDF Manager — Upload Java/Python/Scala JARs directly from the browser. Guided 3-step registration form (class path, method, parameter types, scope). Supports ScalarFunction, TableFunction, AggregateFunction. ADD JAR browser URL bug fixed — always uses container filesystem path.
View Builder — Create TEMPORARY VIEWs with computed columns and CASE WHEN expressions via a visual form. No JAR required.
Chart Report — Plot streaming results as bar, line, area, scatter, pie, donut, histogram, or heatmap. Explicit X axis (labels/categories) and Y axis (values) selection. Live 2s refresh. PDF export.
Colour Describe — Rules engine for live row highlighting. Pick any column, operator, value, and colour. Applies to streaming rows as they arrive. Supports background, border accent, and text colour styles.
SHOW queries → shared Statements slot — SHOW TABLES, SHOW JARS, SHOW VIEWS, DESCRIBE, EXPLAIN results no longer create a new result badge per query. Appended to the shared Statements slot.
SHOW VIEWS fix — TEMPORARY views are now tracked in session state and merged into SHOW VIEWS results. Views appear correctly even though Flink doesn't list TEMPORARY views natively.
ADD JAR path resolution — "Use →" button now sets the container filesystem path, not the browser URL.
Language validation — UDF language check now runs after SHOW JARS (real classpath, not stale cache).
Session JAR state clearing — JAR state cleared on session expiry or disconnect.
DESCRIBE on functions — DESCRIBE/DESC on a function name is auto-rewritten to SHOW CREATE FUNCTION.
Tab isolation fix — UDF Manager Step 1–3 panels no longer leak onto non-register tabs.
Flink 2.0 compatibility — Documented and verified. Studio works with Flink 1.16 through 2.x.

v1.2.4

Studio image is now self-contained — nginx, proxy config, WebDAV JAR storage, and entrypoint bundled.
No CORS proxy required for Via Studio mode.
UDF JAR upload via shared Docker volume — no Hadoop required.
nginx startup crash fixed.

v1.2.0

UDF Manager overhaul — 3-step guided wizard, ClassNotFoundException diagnosis.

v1.0.22

Brand rename to Str:::lab Studio; Apache Flink trademark attribution added.

v1.0.21

Project Manager, JAR upload fix, Maven/Gradle config generator.

v1.0.19 – v1.0.20

Admin Session, UDF Manager, Colour Describe, duplicate submission guard.

v1.0.1 – v1.0.18

Initial build: SQL editor, sessions, results, Job Graph, performance, themes, workspace.

Contributing

We welcome contributions! The project is built with plain HTML/CSS/JS to keep it dependency-free.

Fork the repo and create a feature branch.
Each JS module has a single clear responsibility — keep it that way.
No build step, no bundler, no dependencies — plain HTML/CSS/JS only.
Test against Flink SQL Gateway 1.17+.
Open a PR with a clear description of what changed and why.

License

Apache License 2.0 — see LICENSE.

Created by Nestor Martourez A. A · coded-streams

Name		Name	Last commit message	Last commit date
Latest commit History 77 Commits
.github/workflows		.github/workflows
docs		docs
examples		examples
k8s		k8s
nginx		nginx
scripts		scripts
studio		studio
.gitignore		.gitignore
CONTRIBUTING.md		CONTRIBUTING.md
Dockerfile		Dockerfile
LICENSE		LICENSE
README.md		README.md
docker-compose.yml		docker-compose.yml
docker-entrypoint.sh		docker-entrypoint.sh

Folders and files

Latest commit

History

Repository files navigation

Str:::lab Studio

⭐ Support This Project

What is it?

Quickstart Cluster Setup

What do I need to run this?

Quickstart

Option 1 — Start everything from scratch

Option 2 — I already have a Flink cluster and SQL Gateway

Option 3 — I have a Flink cluster but no SQL Gateway

Option 4 — Standalone docker run

Option 5 — Cloud / Remote Flink

Option 6 — Kubernetes

How the Studio connects — architecture

Connection Modes

Environment Variables

What is the CORS proxy and do I need it?

Pipeline Manager

Systems Manager

Catalog Manager

UDF JAR Upload

UDF deployment paths

View Builder

Chart Report

Colour Describe

Job Graph & Resource Monitoring

Performance Benchmarking

Admin Session

Report Generation

Flink 2.0 Compatibility

Connector JARs

Recommended session setup

📚 Additional Resources

Changelog

v1.3.2

v1.3.0

v1.2.4

v1.2.0

v1.0.22

v1.0.21

v1.0.19 – v1.0.20

v1.0.1 – v1.0.18

Contributing

License

About

Resources

License

Contributing

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

Option 4 — Standalone `docker run`

Packages