* [installer] make sure dashboard is deployed after server and papi-server
* fix build
* Add unit tests
* address feedback
* wait feature flag until get actual value of timed out
* default config cat client nil
* log avg fetch time
* 1
* mock feature flag hang
* Add metric
* fixup
* [public-api] add dummy service for testing
* [public-api] proxy dummy to server
* [public-api] hello service server impl
* [server] fix API contribution bindings
* [dashboard] emulate unary call
* only if actually called
* [dummy] auth
* fix tests
* [server] add interceptor to public api
* add server side observability
* fix port name
* change to unimplemented for unknown methods
* [public-api] client metrics
* fix metrics imports
* align server metrics
* actually fix metrics
* add feature flags
* fix server side streams
* [dashboard] hook error reporting
* rebase and fix imports
* feature flagged metrics from dashboard
* revert GRPC_TYPE
* address feedback
They are too noisy right now. We still report observability. Later we are going to reconsider to report only from our components to make it more useful.
Co-authored-by: mustard <mhqnwt@gmail.com>
Co-authored-by: Anton Kosyakov <anton@gitpod.io>
Co-authored-by: Jean Pierre <jeanp413@hotmail.com>
Co-authored-by: iQQBot <tianshi8650@gmail.com>
Move remainder of IDE metrics to under OpenVSCode Server workbench
Co-authored-by: Anton Kosyakov <anton@gitpod.io>
Co-authored-by: mustard <mhqnwt@gmail.com>
