Some checks failed
codecov / upload (push) Failing after 13m59s
CodeQL / Analyze (push) Failing after 1s
govulncheck / govulncheck (.) (push) Failing after 1s
govulncheck / govulncheck (./cmd/protoc-gen-go-grpc) (push) Failing after 1s
govulncheck / govulncheck (./gcp/observability) (push) Failing after 1s
govulncheck / govulncheck (./security/advancedtls) (push) Failing after 1s
govulncheck / govulncheck (./stats/opencensus) (push) Failing after 1s
Testing / vet-proto (push) Failing after 1s
Testing / tests (arm, latest) (push) Waiting to run
Testing / extras (latest) (push) Failing after 1s
Testing / static checks (latest-1) (push) Failing after 1s
Testing / tests (-race, latest) (push) Failing after 1s
Testing / tests (i386, latest) (push) Failing after 1s
Testing / tests (latest) (push) Failing after 1s
Testing / tests (latest-1) (push) Failing after 1s
Lock Threads / lock (push) Failing after 1s
Stale bot / stale (push) Failing after 1s
49 lines
1.6 KiB
Markdown
49 lines
1.6 KiB
Markdown
# ORCA Load Reporting
|
|
|
|
ORCA is a protocol for reporting load between servers and clients. This
|
|
example shows how to implement this from both the client and server side. For
|
|
more details, please see [gRFC
|
|
A51](https://github.com/grpc/proposal/blob/master/A51-custom-backend-metrics.md)
|
|
|
|
## Try it
|
|
|
|
```
|
|
go run server/main.go
|
|
```
|
|
|
|
```
|
|
go run client/main.go
|
|
```
|
|
|
|
## Explanation
|
|
|
|
gRPC ORCA support provides two different ways to report load data to clients
|
|
from servers: out-of-band and per-RPC. Out-of-band metrics are reported
|
|
regularly at some interval on a stream, while per-RPC metrics are reported
|
|
along with the trailers at the end of a call. Both of these mechanisms are
|
|
optional and work independently.
|
|
|
|
The full ORCA API documentation is available here:
|
|
https://pkg.go.dev/google.golang.org/grpc/orca
|
|
|
|
### Out-of-band Metrics
|
|
|
|
The server registers an ORCA service that is used for out-of-band metrics. It
|
|
does this by using `orca.Register()` and then setting metrics on the returned
|
|
`orca.Service` using its methods.
|
|
|
|
The client receives out-of-band metrics via the LB policy. It receives
|
|
callbacks to a listener by registering the listener on a `SubConn` via
|
|
`orca.RegisterOOBListener`.
|
|
|
|
### Per-RPC Metrics
|
|
|
|
The server is set up to report query cost metrics in its RPC handler. For
|
|
per-RPC metrics to be reported, the gRPC server must be created with the
|
|
`orca.CallMetricsServerOption()` option, and metrics are set by calling methods
|
|
on the returned `orca.CallMetricRecorder` from
|
|
`orca.CallMetricRecorderFromContext()`.
|
|
|
|
The client performs one RPC per second. Per-RPC metrics are available for each
|
|
call via the `Done()` callback returned from the LB policy's picker.
|