Releases: arayabrain/araya-optinist
v1.1.4
Summary
Stabilization release centered on the premium instance lifecycle. Premium manager and ECS startup are restored to a proven baseline to eliminate v1.1.3 regressions, and CloudWatch alarms are tuned to reduce noise from transient states. The Platform service info view also gets a more compact, readable layout.
Improvements
Application
- Refine Platform service info display with compact task IDs, single-line layout, and task-to-instance mapping #544
Platform/Infrastructure
- Stabilize premium instance lifecycle and ECS startup behavior #551
- Tune CloudWatch alarms to reduce noise from transient disconnect and startup states #540 #541
Bug Fixes
Platform/Infrastructure
- Fix IAM permissions for premium manager Lambda #536
Full Changelog: v1.1.3...v1.1.4
DEPRECATED - v1.1.3
Summary
DEPRECATED — do not use. This release introduced regressions from #503 and #504. Superseded by v1.1.4, which contains the caiman fix from v1.1.3 with #503/#504 rolled back. See #551.
A stabilization release focused on platform reliability. Key fixes address premium instance lifecycle management, agent recovery, and scaling behavior. Also includes a CaImAn environment upgrade to Python 3.11.
Improvements
Platform/Infrastructure
- Upgrade CaImAn conda environment to Python 3.11 #507
Bug Fixes
Application
- Fix premium manager permission issue #509
Platform/Infrastructure
- Resolve stale agent detection with new Lambda and startup health check #503
- Prevent Docker errors and port conflicts from crashing the agent #504
- Fix agent-recovery scripts interfering with ASG scaling management #522
Full Changelog: v1.1.2...v1.1.3
v1.1.2
Summary
This release focuses on application reliability and cross-environment isolation, with key fixes for premium user routing errors, visualization data handling, and Lambda scope contamination between production and development environments.
Improvements
Application
Display "Backend unavailable" screen when the backend is down #502
Platform/Infrastructure
Untagged resources report for AWS cost attribution #490
Bug Fixes
Application
Fix visualize panel showing stale data after intermediate data deletion #467
Fix discrepancy between HDF5 input data and thumbnail #473
Fix discrepancy between S3 bucket list and Record screen list #448
Platform/Infrastructure
Fix Lambda scope and ghost instance issue causing 503 errors for premium users #477 #436
Fix cross-environment contamination in Lambda packages and CloudWatch namespaces #471 #457
Fix development environment outage caused by dev config impact on production #468
And, other minor bug fixes
Full Changelog: v1.1.1...v1.1.2
v1.1.1
Summary
This release focuses on infrastructure reliability, particularly around premium instance lifecycle management and environment isolation for the dev environment.
Improvements
Application
Display platform information in Log viewer #462
Platform/Infrastructure
Increase active users threshold before scale-up #461
Event-driven cleanup on EC2 instance termination #460
Add environment filtering to premium instance discovery #456
Isolate environment variables to prevent cross-env contamination #459
Bug Fixes
Platform/Infrastructure
Fix SSM Agent timing and S3 permissions for instance cleanup #444 #446
Terminate inactive premium instances reliably #464
Add ALB creation retry to handle race condition #465
Fix RDS Proxy target registration after dev scheduler restore #469
Fix IAM permissions and monitoring email #474 #475
And, other minor bug fixes
Full Changelog: v1.1.0...v1.1.1
v1.1.0
Summary
The main focus is on bug fixes after the v1.0 release.
- In particular, improvements to stabilize the allocation of user-dedicated instances.
Features
Platform/Infrastructure
- Expired subscription data deletion policy and implementation #380
Improvements
Application
Platform/Infrastructure
- Fix cost tracker, add to monthly-maintenance #386
- Maintenance procedure documentation, scripts and email alerts #382
- Terraform Develop Environment #383
- Centralize instance type configuration into variables #404
- Separate ECR with development and production #409
- Add dev environment cost savings automatic shutdown/start #407
- Assigned to the autoscaling-pool fallback can get stuck #423
Bug Fixes
Application
- Public experiment input files display #364
- Fix/thumbnail placeholder type image #376
- update code subscription checkout bug #388
- Mitigate snakemake yaml sync error #391
- Edit ROI to use absolute path #428
- Move DB update after observe_overall success #433
Platform/Infrastructure
- Fix false assigning popup on refresh because hasAttemptedAutoAssignment #378
- Add sticky sessions to premium target groups #379
- Add terminate_aged_stopped_instances to run during cleanup #400
- ExperimentsSynced, SyncErrors, and SyncErrorRate CloudWatch metrics #403
- fix: resolve dev environment login/registration failures and Inject infrastructure values to terraform #413
- Fixes for dev env #414
- Premium users stuck when instance stopped, but they log in again #424
- Update Dev Scheduler #431
- Check instances are lively to avoid stale db reocrds #435
- Add 503 error fix #438
- Enhance premium manager instance health checks #437
- Split error messages based on error type #439
- Fix premium manager stale assignment bugs and reorder assignment priorities (fixed) #440
And, other minor bug fixes
Full Changelog
v0.9.0
Summary
This is a pre-verification version prior to the introduction of features in v1.0.0.
- Main unimplemented features
- Cloud Platform: Only minimal AWS S3 access functionality is implemented
- Subscription Plans: Not implemented
v1.0.0
Araya-OptiNiSt Cloud
Cloud-based calcium imaging pipeline tool for open, reproducible neuroscience.
Built by Araya Inc. on top of OptiNiSt, originally developed in collaboration with OIST.
Cloud Platform
- Public Data Sharing — Publish experiments to a public page; anyone can view results and reproduce workflows without an
account - Cloud Computing — Run analysis pipelines on cloud infrastructure with no local hardware required
- Cloud Storage — Secure Amazon S3-backed storage with on-demand synchronization
Subscription Plans
- Free — Shared compute, limited storage
- Premium — Dedicated compute, expanded storage
Analysis Pipeline
- GUI Workflow Editor — Build analysis pipelines visually using a node-and-edge graph; no coding required
- Interactive Visualization — One-click result visualization powered by Plotly with multiple plot styles
- Workflow Management — Record, reproduce, and share complete pipeline configurations
Supported Libraries
- ROI Detection: Suite2p, CaImAn, LCCD
- Postprocessing analysis libraray including: ETA, PCA, TSNE, correlation analysis
- Data Format: NWB (Neurodata Without Borders)
Technical Stack
- Backend: Python 3.11+, FastAPI, SQLModel, Firebase, AWS (Boto3), Snakemake
- Frontend: React 18, TypeScript, Material-UI