launch-igv-studio

name	launch-igv-studio
description	Launch IGV Studio instances on Seqera Platform. Use when deploying IGV Studio, starting genomics visualization studios, or mounting data links to studios.
compatibility	Requires tw CLI, TOWER_ACCESS_TOKEN, and optionally wave CLI for building images.
metadata	{"author":"Seqera/edmundmiller","version":"1.0"}

Launch IGV Studio

Launch IGV Studio instances on Seqera Platform with optional data link mounting.

Prerequisites

export TOWER_ACCESS_TOKEN=$(op read "op://Employee/Seqera Platform Prod/password")

Quick Start

Build with Wave and launch immediately:

# Build and capture image URL
IMAGE_URL=$(wave -f Dockerfile --context . --platform linux/amd64 --await --tower-token "$TOWER_ACCESS_TOKEN")

# Launch studio
tw studios add \
  --name "IGV Studio" \
  --workspace "scidev/testing" \
  --custom-template "$IMAGE_URL" \
  --compute-env "seqera_aws_london_fusion_nvme" \
  --auto-start

Build Options

Option 1: Wave Temporary URL (~24h)

No registry push permissions required:

wave -f Dockerfile --context . --platform linux/amd64 --await --tower-token "$TOWER_ACCESS_TOKEN"
# Returns: wave.seqera.io/wt/xxxxx/wave/build:yyyyy

Option 2: Wave Persistent Image

Requires push permissions to target registry:

wave -f Dockerfile --context . --build-repo cr.seqera.io/seqera-services/igv-studio \
  --platform linux/amd64 --freeze --await --tower-token "$TOWER_ACCESS_TOKEN"

Option 3: Docker buildx (Local)

Requires local Docker + registry auth:

gh auth refresh -h github.com -s write:packages
docker buildx build --platform linux/amd64 -t cr.seqera.io/seqera-services/igv-studio:latest --push .

Launch with Data Links

Mount data links for genomic data access:

tw studios add \
  --name "IGV Studio - My Project" \
  --workspace "scidev/testing" \
  --custom-template "cr.seqera.io/seqera-services/igv-studio:latest" \
  --compute-env "seqera_aws_london_fusion_nvme" \
  --mount-data "my-data-link-name" \
  --auto-start

Available Compute Environments (scidev/testing)

• seqera_aws_london_fusion_nvme - AWS Batch with Fusion + NVMe (recommended)
• seqera_gcp_finland_fusion - Google Batch with Fusion
• seqera_azure_virginia_fusion - Azure Batch with Fusion

Manage Studios

# List studios
tw studios list -w scidev/testing

# View studio details
tw studios view --name "IGV Studio" -w scidev/testing

# Stop studio by session ID
tw studios stop --session-id <SESSION_ID> -w scidev/testing

Image Requirements

• Connect client: v0.9.0+ (set via CONNECT_CLIENT_VERSION build arg)
• Architecture: linux/amd64 for cloud compute environments
• User: Must run as root (no USER directive) for Fusion loop device setup

See references/TROUBLESHOOTING.md for common issues. See references/WORKFLOW.md for iterative development workflow.