[OM-88241] preventing cache from recursively accessing itself
[OM-84781] prevent cache crash if per-server repository init failed; log exceptions before crash
[OM-54730] listen_port configuration was not changeable through the toml config file
Updated the Jinja2 package version to 3.1.2
Updated the idl.py package version to 0.16+master
Updated the websockets package version to 10.3
Clean up compliance tooling
[OM-56139] remove ‘Host’ header when accessing upstreams (except if the upstream is a cache)
Fixed compliance pipeline elements
[OM-53614] Fixing the linux build
[OM-51454] Trying to resolve the problem “The specified network name is no longer available”
[OM-48540] Avoid logging exceptions failing to read toml configs when running inside a container
[OM-48746] Disable upstream cache upon startup if Nucleus Cache can’t connect to the upstream/cache_speedtest endpoint
[FDN-190] Open telemetry tracing.
Support for X-Cache header, make cache warmer download only not cached files
[Windows] Added system SSL certs in addition to certifi certs
Use total data size in cache cleaner checks
update omni-config-py package to 2.1.3 version (fix for possibility of not initialized control port number)
Fixed caching when remote server doesn’t return Content-Length header, fixed issue when content downloaded not completelly
Added least recently used cached data cleanup
Package SSL certificates together with the cache application by using certifi package
Removed whitespaces from proxy_server config option
Added CORS headers for /ping endpoint
Added fix for LookupError bug
Added “/ping” endpoint for checking if cache is alive by HEAD request
Disabled metrics calculation if OmniverseConfig.metrics_enabled == False
When using upstream proxy/cache with a port not specified, use the default one from the config (8891)
Fixed bug with undefined proxy_server
Made cache more secure by upgrading websockets
Removed unsecure npm packages from docker image
Added proxy_server_compression, client_compression and other_upstreams_compression config options
Added cache_speedtest endpoint and speed measurement to upstream proxy
Added speed measurement to other upstreams like LFT
Metrics have been changed to match Prometheus Metrics RFC
Added cache_type configuration option
Switched to approved ubuntu image
Fixed caching of non-S3/non-API requests
Added cache_head_requests_enable option
Added redirects support for HEAD requests
Added cache_without_etag_enable option to config (for caching requests without etag header)
Supporting https caching
Added ‘stale’ API cache status in the access log to indicate that the cached response was returned, but it was stale and was scheduled to be updated
Added SSL support.
Fixed API caching when accessing S3 bucket list
Updated service to use Python 3.8.7.
Updated omni-config-py with bugfixes for paths on Windows.
Updated omniverse.discovery.client.py with detailed error logs.
Fixing Linux packaging
API cache: do not cache error-responses (with HTTP status >= 400)
Removed the legacy IDL transport and service implementations.
Updated the service to generate client libraries and service stubs using email@example.com.
Allow configuring used ports via System Monitor.
Register the service in Nucleus Discovery.
Changed the default paths for the cache and logs root using omni-config library.
Disable metrics on non-Docker setups.
Fixing a bug where some responses don’t have ‘Transfer-Encoding’ header and that was causing an exception
Disabling aiohttp.ClientSession cache (it leads to a rare bug)
Fixing packaging to include platform name
Do not consume 100% of the core when removing a folder fails, improve logging when it fails
OM-18418: initializing metrics to 0 on startup
Properly suppport compression
Server on localhost only by default
IDL interface fixes
Support for helm charts.
Teamcity build tools.
Fix: windows path length issues.
Fix: long running coroutines crashing.
Fix: access logging from the dashboard in the main log
Feat: cache HEAD requests for S3 resources.
Feat: supervise consumer and producer coroutines in http resolver
Feat: enable caching Nucleus server S3 API requests
Initial Cache Server Release