LMNR CLI

LMNR is a kubectl-style command-line tool for managing Laminar pipelines, connection profiles, and tables. It communicates with the Laminar API via HTTP and uses YAML manifests for declarative resource management.

Quick Reference

Command	Description
`lmnr apply -f <file>`	Create or update resources from YAML
`lmnr list <type>`	List resources (pipelines, profiles, tables, jobs)
`lmnr get <type> <id>`	Get detailed resource information
`lmnr describe <type> <id>`	Show comprehensive resource details
`lmnr delete <type> <id>`	Delete a resource by ID
`lmnr delete -f <file>`	Delete resources defined in YAML
`lmnr summary`	Display cluster overview
`lmnr cleanup`	Remove all resources

Configuration

Environment Variables

# API endpoint (default: <laminar_backend>/api/v1)
export LAMINAR_API_URL="<laminar_backend>/api/v1"
 
# Optional: Authentication token
export LAMINAR_TOKEN="your-auth-token"

Global Options

All commands support these options:

Option	Description	Default
`--api-url <URL>`	Laminar API URL	`<laminar_backend>/api/v1`
`--token <TOKEN>`	Authentication token	-
`-o, --output <FORMAT>`	Output format: `table`, `yaml`, `json`	`table`
`-h, --help`	Print help	-
`-V, --version`	Print version	-

Configuration Precedence

Options are resolved in this order (highest to lowest priority):

Command-line flags (--api-url, --token)
Environment variables (LAMINAR_API_URL, LAMINAR_TOKEN)
Default values

Commands

apply

Create or update resources from YAML manifest files. Supports both individual files and directories.

lmnr apply -f <FILE_OR_DIRECTORY>

Options:

Option	Description
`-f, --file <PATH>`	Path to YAML manifest file or directory (required)

Features:

Directory support: When given a directory, recursively processes all .yaml and .yml files
Profile name resolution: Tables can reference profiles by name instead of ID - LMNR automatically resolves the name to the correct ID
Multi-resource files: Process multiple resources in a single file separated by ---
Idempotent: Safely re-run without creating duplicates

Examples:

# Apply a single manifest file
lmnr apply -f pipeline.yaml
 
# Apply all manifests in a directory
lmnr apply -f ./manifests/
 
# Apply with custom API URL
lmnr apply -f pipeline.yaml --api-url http://laminar.example.com:8000/api/v1

Output:

Parsed 3 resource(s) from pipeline.yaml

[1/3] Applying Table 'events'...
  ✓ Created Table 'events' (ID: ct_7DYJlexDEz)

[2/3] Applying Table 'output'...
  ✓ Created Table 'output' (ID: ct_Uyukf2ajPK)

[3/3] Applying Pipeline 'my-first-pipeline'...
  ✓ Created Pipeline 'my-first-pipeline' (ID: pl_Irnc9kACAC)

All resources applied successfully

list

List Laminar resources of a specific type. Automatically handles pagination for large result sets.

lmnr list <RESOURCE_TYPE> [OPTIONS]

Resource Types:

Type	Description
`pipelines`	List all pipelines
`profiles`	List connection profiles
`tables`	List connection tables
`jobs`	List jobs (optionally filtered by pipeline)

Options:

Option	Description
`--pipeline-id <ID>`	Filter jobs by pipeline ID (only for `jobs`)

Examples:

# List all pipelines
lmnr list pipelines
 
# List all tables
lmnr list tables
 
# List all profiles
lmnr list profiles
 
# List jobs for a specific pipeline
lmnr list jobs --pipeline-id pl_Irnc9kACAC
 
# Output as YAML
lmnr list pipelines -o yaml
 
# Output as JSON for scripting
lmnr list pipelines -o json

Output Examples:

Pipelines:

┌───────────────────┬───────────────┬─────────────┬────────────┬─────────┐
│ NAME              ┆ ID            ┆ PARALLELISM ┆ CHECKPOINT ┆ STATUS  │
╞═══════════════════╪═══════════════╪═════════════╪════════════╪═════════╡
│ my-first-pipeline ┆ pl_Irnc9kACAC ┆ 1           ┆ 1s         ┆ Running │
└───────────────────┴───────────────┴─────────────┴────────────┴─────────┘

Tables:

┌────────┬───────────────┬───────────┬─────────┬─────────┐
│ NAME   ┆ ID            ┆ CONNECTOR ┆ PROFILE ┆ CREATED │
╞════════╪═══════════════╪═══════════╪═════════╪═════════╡
│ output ┆ ct_Uyukf2ajPK ┆ preview   ┆ -       ┆ 5m      │
├╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌┤
│ events ┆ ct_7DYJlexDEz ┆ mock      ┆ -       ┆ 5m      │
└────────┴───────────────┴───────────┴─────────┴─────────┘

Profiles:

┌────────────┬────────────────┬───────────┐
│ NAME       ┆ ID             ┆ CONNECTOR │
╞════════════╪════════════════╪═══════════╡
│ kafka-prod ┆ prof_abc123def ┆ kafka     │
└────────────┴────────────────┴───────────┘

Jobs:

┌─────────────────┬─────────────────┬─────────┬───────┬─────────┐
│ ID              ┆ RUN_ID          ┆ STATUS  ┆ TASKS ┆ AGE     │
╞═════════════════╪═════════════════╪═════════╪═══════╪═════════╡
│ job_xyz789      ┆ run_001         ┆ Running ┆ 4     ┆ 2h      │
└─────────────────┴─────────────────┴─────────┴───────┴─────────┘

get

Get detailed information about a specific resource.

lmnr get <RESOURCE_TYPE> <ID>

Resource Types: pipelines, profiles, tables, jobs

Examples:

# Get pipeline details
lmnr get pipelines pl_Irnc9kACAC
 
# Get table details as YAML
lmnr get tables ct_7DYJlexDEz -o yaml
 
# Get profile details as JSON
lmnr get profiles prof_456 -o json
 
# Get job details
lmnr get jobs job_xyz789

Output (pipeline):

Pipeline Details
==================================================
Name:                my-first-pipeline
ID:                  pl_Irnc9kACAC
Status:              Running
Parallelism:         1
Checkpoint Interval: 1s
Preview:             false
Created:             2025-12-03 02:24:50 UTC (5m)

Query:
--------------------------------------------------
INSERT INTO output SELECT id, value * 2 as doubled_value, timestamp FROM events WHERE value > 50

Execution Graph:
--------------------------------------------------
Nodes: 3
  Mock<SchemaDriven<3 fields, 100 eps>> (mock) - parallelism: 1
  events -> watermark -> sink projection (chained_op) - parallelism: 1
  PreviewSink (preview) - parallelism: 1
Edges: 2
  0 -> 1
  1 -> 4

Output (job):

Job Details
==================================================
ID:              job_xyz789
Run ID:          run_001
State:           Running
Running Desired: true
Start Time:      2025-12-03 02:24:50 UTC
Tasks:           4
Created:         2025-12-03 02:24:50 UTC (2h)

describe

Show comprehensive details about a resource including configuration, status, execution graph, and related resources.

lmnr describe <RESOURCE_TYPE> <ID>

Resource Types: pipelines, jobs (only these two types are supported)

Examples:

# Describe pipeline (includes recent jobs, operators, execution graph)
lmnr describe pipelines pl_Irnc9kACAC
 
# Describe job (includes execution details, failure messages if any)
lmnr describe jobs job_xyz789

Output (pipeline describe):

Pipeline Details
==================================================
Name:                my-first-pipeline
ID:                  pl_Irnc9kACAC
Status:              Running
Parallelism:         1
Checkpoint Interval: 1s
Preview:             false
Created:             2025-12-03 02:24:50 UTC (5m)

Query:
--------------------------------------------------
INSERT INTO output SELECT id, value * 2 as doubled_value, timestamp FROM events WHERE value > 50

Execution Graph:
--------------------------------------------------
Nodes: 3
  Mock<SchemaDriven<3 fields, 100 eps>> (mock) - parallelism: 1
  events -> watermark -> sink projection (chained_op) - parallelism: 1
  PreviewSink (preview) - parallelism: 1
Edges: 2
  0 -> 1
  1 -> 4

Recent Jobs:
--------------------------------------------------
┌─────────────────┬─────────────────┬─────────┬───────┬─────────┐
│ ID              ┆ RUN_ID          ┆ STATUS  ┆ TASKS ┆ AGE     │
╞═════════════════╪═════════════════╪═════════╪═══════╪═════════╡
│ job_xyz789      ┆ run_001         ┆ Running ┆ 4     ┆ 5m      │
└─────────────────┴─────────────────┴─────────┴───────┴─────────┘

delete

Delete Laminar resources by ID or from a manifest file.

# Delete by ID
lmnr delete <RESOURCE_TYPE> <ID>
 
# Delete from manifest file
lmnr delete -f <FILE>

Resource Types: pipelines, profiles, tables, jobs

Options:

Option	Description
`-f, --file <FILE>`	Delete resources defined in YAML file

Deletion Order (file mode):

When deleting from a manifest file, resources are deleted in reverse dependency order:

Pipelines (first)
Tables
Profiles (last)

Examples:

# Delete a pipeline by ID
lmnr delete pipelines pl_Irnc9kACAC
 
# Delete a table by ID
lmnr delete tables ct_7DYJlexDEz
 
# Delete a profile by ID
lmnr delete profiles prof_456
 
# Delete a job by ID
lmnr delete jobs job_xyz789
 
# Delete all resources defined in a manifest
lmnr delete -f pipeline.yaml

summary

Display a cluster overview showing all resources.

lmnr summary

Output:

Cluster Summary

Connection Profiles
  No profiles found

Connection Tables
┌────────┬───────────────┬───────────┬─────────┬─────────┐
│ NAME   ┆ ID            ┆ CONNECTOR ┆ PROFILE ┆ CREATED │
╞════════╪═══════════════╪═══════════╪═════════╪═════════╡
│ output ┆ ct_Uyukf2ajPK ┆ preview   ┆ -       ┆ 5m      │
├╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌┤
│ events ┆ ct_7DYJlexDEz ┆ mock      ┆ -       ┆ 5m      │
└────────┴───────────────┴───────────┴─────────┴─────────┘

Pipelines
┌───────────────────┬───────────────┬─────────────┬────────────┬─────────┐
│ NAME              ┆ ID            ┆ PARALLELISM ┆ CHECKPOINT ┆ STATUS  │
╞═══════════════════╪═══════════════╪═════════════╪════════════╪═════════╡
│ my-first-pipeline ┆ pl_Irnc9kACAC ┆ 1           ┆ 1s         ┆ Running │
└───────────────────┴───────────────┴─────────────┴────────────┴─────────┘

Jobs
  No jobs found

cleanup

Remove all resources from the cluster. Use with caution as this operation cannot be undone.

lmnr cleanup [--force]

Options:

Option	Description
`--force`	Skip confirmation prompt

Operation Order:

The cleanup command removes resources in the following order to respect dependencies:

Stop all running pipelines (with retry logic)
Delete all pipelines (with retry logic)
Delete all tables
Delete all profiles

Examples:

# With confirmation prompt
lmnr cleanup
 
# Skip confirmation (useful for scripts)
lmnr cleanup --force

Output:

⚠️  This will delete ALL resources from the cluster.
Are you sure you want to continue? [y/N]: y

Stopping pipelines...
  ✓ Stopped pipeline 'my-first-pipeline'

Deleting pipelines...
  ✓ Deleted pipeline 'my-first-pipeline'

Deleting tables...
  ✓ Deleted table 'output'
  ✓ Deleted table 'events'

Deleting profiles...
  No profiles to delete

Cleanup complete

YAML Manifest Format

LMNR uses Kubernetes-style YAML manifests for defining resources.

Resource Structure

---
apiVersion: laminar.io/v1
kind: <ResourceKind>
spec:
  name: <resource-name>
  # ... resource-specific configuration

Supported Resource Kinds:

Profile - Connection profiles
Table - Connection tables
Pipeline - Data pipelines

Multiple Resources in One File

Combine multiple resources using --- separator:

---
apiVersion: laminar.io/v1
kind: Profile
spec:
  name: kafka-local
  # ...
 
---
apiVersion: laminar.io/v1
kind: Table
spec:
  name: events
  # ...
 
---
apiVersion: laminar.io/v1
kind: Pipeline
spec:
  name: my-pipeline
  # ...

Profile

Connection profiles define how to connect to external systems like Kafka, Iceberg, or cloud services.

---
apiVersion: laminar.io/v1
kind: Profile
spec:
  name: kafka-prod
  connector: kafka
  config:
    bootstrap_servers: kafka.example.com:9092
    authentication:
      type: sasl
      sasl_config:
        protocol: SASL_SSL
        mechanism: SCRAM-SHA-512
        username: my-user
        password: my-password

Supported Connectors:

Connector	Description	Requires Profile
`kafka`	Apache Kafka	Yes
`confluent`	Confluent Cloud	Yes
`iceberg`	Apache Iceberg	Yes
`kinesis`	AWS Kinesis	Yes
`delta`	Delta Lake	Yes
`filesystem`	Local/S3/GCS filesystem	Yes
`mock`	Mock data generator	No
`preview`	Preview sink	No
`stdout`	Standard output sink	No

Table

Tables define source and sink connections with their schemas.

Profile Reference:

Tables can reference profiles by name or ID. When using a name, LMNR automatically resolves it to the correct profile ID:

# Reference by name (recommended)
connection_profile_id: kafka-prod
 
# Reference by ID
connection_profile_id: prof_abc123def

Examples

Source Table (Mock)

---
apiVersion: laminar.io/v1
kind: Table
spec:
  name: events
  connector: mock
  config:
    SchemaDriven:
      name: events
      fields:
        - name: id
          type: int64
          generator:
            generator_type: sequence
            start: 1
        - name: value
          type: float64
          generator:
            generator_type: range
            min: 0
            max: 100
        - name: timestamp
          type: timestamp
          generator:
            generator_type: datetime
      generation:
        mode: streaming
        rate: 100.0
  schema:
    format:
      json: {}
    fields:
      - field_name: id
        field_type:
          type:
            primitive: Int64
        nullable: false
      - field_name: value
        field_type:
          type:
            primitive: F64
        nullable: false
      - field_name: timestamp
        field_type:
          type:
            primitive: DateTime
        nullable: false

Source Table (Kafka)

---
apiVersion: laminar.io/v1
kind: Table
spec:
  name: orders-source
  connector: kafka
  connection_profile_id: kafka-prod
  config:
    topic: orders
    type:
      source_config:
        offset: earliest
  schema:
    format:
      json:
        timestampFormat: rfc3339
    fields:
      - field_name: order_id
        field_type:
          type:
            primitive: String
        nullable: false
      - field_name: amount
        field_type:
          type:
            primitive: F64
        nullable: false
      - field_name: timestamp
        field_type:
          type:
            primitive: DateTime
        nullable: false

Sink Table (Preview)

---
apiVersion: laminar.io/v1
kind: Table
spec:
  name: output
  connector: preview
  config: {}
  schema:
    format:
      json: {}
    fields:
      - field_name: id
        field_type:
          type:
            primitive: Int64
        nullable: false
      - field_name: result
        field_type:
          type:
            primitive: F64
        nullable: false

Sink Table (Iceberg)

---
apiVersion: laminar.io/v1
kind: Table
spec:
  name: events-sink
  connector: iceberg
  connection_profile_id: iceberg-lakehouse
  config:
    type: sink
    sink_table_config:
      namespace: my_database
      table_name: events
      rolling_policy:
        interval_seconds: 10
  schema:
    format:
      parquet: {}
    fields:
      - field_name: id
        field_type:
          type:
            primitive: Int64
        nullable: false
      - field_name: value
        field_type:
          type:
            primitive: F64
        nullable: false

Pipeline

Pipelines define SQL transformations that process data from source tables to sink tables.

---
apiVersion: laminar.io/v1
kind: Pipeline
spec:
  name: my-pipeline
  query: |
    INSERT INTO output
    SELECT id, value * 2 as doubled_value, timestamp
    FROM events
    WHERE value > 50
  parallelism: 1
  checkpoint_interval_micros: 1000000

Pipeline Fields:

Field	Type	Required	Description
`name`	string	Yes	Pipeline name
`query`	string	Yes	SQL query with INSERT INTO...SELECT
`parallelism`	integer	No	Number of parallel tasks (default: 1)
`checkpoint_interval_micros`	integer	No	Checkpoint interval in microseconds (default: 1000000 = 1s)

Complete Example

1. Create the manifest

Create my-pipeline.yaml:

---
apiVersion: laminar.io/v1
kind: Table
spec:
  name: events
  connector: mock
  config:
    SchemaDriven:
      name: events
      fields:
        - name: id
          type: int64
          generator:
            generator_type: sequence
            start: 1
        - name: value
          type: float64
          generator:
            generator_type: range
            min: 0
            max: 100
        - name: timestamp
          type: timestamp
          generator:
            generator_type: datetime
      generation:
        mode: streaming
        rate: 100.0
  schema:
    format:
      json: {}
    fields:
      - field_name: id
        field_type:
          type:
            primitive: Int64
        nullable: false
      - field_name: value
        field_type:
          type:
            primitive: F64
        nullable: false
      - field_name: timestamp
        field_type:
          type:
            primitive: DateTime
        nullable: false
 
---
apiVersion: laminar.io/v1
kind: Table
spec:
  name: output
  connector: preview
  config: {}
  schema:
    format:
      json: {}
    fields:
      - field_name: id
        field_type:
          type:
            primitive: Int64
        nullable: false
      - field_name: doubled_value
        field_type:
          type:
            primitive: F64
        nullable: false
      - field_name: timestamp
        field_type:
          type:
            primitive: DateTime
        nullable: false
 
---
apiVersion: laminar.io/v1
kind: Pipeline
spec:
  name: my-first-pipeline
  query: |
    INSERT INTO output
    SELECT id, value * 2 as doubled_value, timestamp
    FROM events
    WHERE value > 50
  parallelism: 1

2. Deploy

lmnr apply -f my-pipeline.yaml

3. Verify

# Check cluster status
lmnr summary
 
# Get pipeline details
lmnr list pipelines
lmnr get pipelines <pipeline_id>
 
# View comprehensive details
lmnr describe pipelines <pipeline_id>

4. Clean up

# Delete resources from manifest
lmnr delete -f my-pipeline.yaml
 
# Or delete everything
lmnr cleanup --force