Breaking Changes
* Now it requires Python 3.10 or higher to run. ([52](https://github.com/lablup/backend.ai-webserver/issues/52))
Features
* Adds configurations on maximum resources per session (cores, CUDA devices, shared memory, and upload size limit). ([28](https://github.com/lablup/backend.ai-webserver/issues/28))
* Prevents too many login attempts, usually an indication of abuse and/or malicious robots. Block time and allowed fail count are configurable. ([29](https://github.com/lablup/backend.ai-webserver/issues/29))
* Add an option to support creating a compute session by manually typing the name of an image. ([37](https://github.com/lablup/backend.ai-webserver/issues/37))
* Add ping API to to check webserver health. ([38](https://github.com/lablup/backend.ai-webserver/issues/38))
* Send X-Forwarded-For header from token login handler for token authentication. ([40](https://github.com/lablup/backend.ai-webserver/issues/40))
* Add webui_debug option to enable debug mode in Web-UI. ([41](https://github.com/lablup/backend.ai-webserver/issues/41))
* General image allowlist support on configuration file ([44](https://github.com/lablup/backend.ai-webserver/issues/44))
* Add allow_signup_without_confirmation in configuration file to disable token input and email confirmation ([50](https://github.com/lablup/backend.ai-webserver/issues/50))
* Add mask_user_info in configuration file to hiding user information ([51](https://github.com/lablup/backend.ai-webserver/issues/51))
Fixes
* Use persistent manager endpoint for websocket streaming (app) per session and per app by storing the endpoint information in the browser session. ([30](https://github.com/lablup/backend.ai-webserver/issues/30))
* Send cookies for an api request. This is required to cookie-based authentication for each request. ([35](https://github.com/lablup/backend.ai-webserver/issues/35))
* Add max_memory_per_container in configuration file to limit memory allocation for each container. ([39](https://github.com/lablup/backend.ai-webserver/issues/39))
* Deliver the client IP by referencing request.remote when X-Forwarded-For is empty. ([43](https://github.com/lablup/backend.ai-webserver/issues/43))
* Update dependencies for 21.09 release ([45](https://github.com/lablup/backend.ai-webserver/issues/45))
* Update dependencies for Backend.AI common 21.9, aiohttp 3.8, and aioredis 2.0 ([47](https://github.com/lablup/backend.ai-webserver/issues/47))
* Upgrade aiohttp-session to a pre-release version compatible with aioredis v2 ([48](https://github.com/lablup/backend.ai-webserver/issues/48))
* Update Jinja2 package from 2.11.x to 3.0.x ([49](https://github.com/lablup/backend.ai-webserver/issues/49))
Miscellaneous
* Improve GitHub Actions CI workflows to run as single pipeline and make the deployment to be triggered via tagged commits like other repositories ([31](https://github.com/lablup/backend.ai-webserver/issues/31))
* Update the name of console-server to webserver ([32](https://github.com/lablup/backend.ai-webserver/issues/32))
Older changelogs
* [21.09](https://github.com/lablup/backend.ai-webserver/blob/21.09/CHANGELOG.md)
* [21.03](https://github.com/lablup/backend.ai-webserver/blob/21.03/CHANGELOG.md)
* [20.09](https://github.com/lablup/backend.ai-webserver/blob/20.09/CHANGELOG.md)
* [20.03](https://github.com/lablup/backend.ai-webserver/blob/20.03/CHANGELOG.md)