Awesome
Description
grpc-client-cli
is a generic gRPC
command line client - call any gRPC
service. If your service exposes gRPC Reflection service the tool will discover all services and methods automatically. If not, please specify --proto
parameter with the path to proto files.
Installation
Download the binary and install it to /usr/local
directory:
- Linux:
curl -L https://github.com/vadimi/grpc-client-cli/releases/download/v1.21.1/grpc-client-cli_linux_x86_64.tar.gz | tar -C /usr/local/bin -xz
(you might need to addsudo
before tar) - macOS:
curl -L https://github.com/vadimi/grpc-client-cli/releases/download/v1.21.1/grpc-client-cli_darwin_x86_64.tar.gz | tar -C /usr/local/bin -xz
For go 1.17+
use this command to install the app to $GOPATH/bin
directory:
go install github.com/vadimi/grpc-client-cli/cmd/grpc-client-cli@v1.21.1
go install github.com/vadimi/grpc-client-cli/cmd/grpc-client-cli@latest
Usage
Just specify a connection string to a service in host:port
format and follow instructions to select service, method and enter request message in json
or proto
text format.
grpc-client-cli localhost:4400
or grpc-client-cli --address localhost:4400
In this case the service needs to expose gRPC Reflection service.
For full list of supported command line args please run grpc-client-cli -h
.
To provide the list of services to call specify --proto
parameter and --protoimports
in case an additional directory for imports is required:
grpc-client-cli --proto /path/to/proto/files localhost:5050
The tool also supports :authority
header override.
grpc-client-cli --authority localhost:9090 localhost:5050
It's also possible to capture some of the diagnostic information like request and response sizes, call duration:
grpc-client-cli -V localhost:4400
Proto text format for input and output:
grpc-client-cli --informat text --outformat text localhost:5050
Eureka Support
grpc-client-cli provides integrated support for services published to a Eureka service registry.
Connecting to a service published to Eureka running on http://localhost:8761/eureka/
grpc-client-cli eureka://application-name/
Connecting to a service running remotely on http://example.com:8761/eureka/
grpc-client-cli eureka://example.com/eureka/application-name/
Connecting to a service running remotely on http://example.com:9000/not-eureka/
grpc-client-cli eureka://example.com:9000/not-eureka/application-name/
The Eureka currently connects to services using the IP Addresses published in the service registry and the following published ports, in order:
- Metadata key "grpc"
- Metadata key "grpc.port"
- Default insecure port
If you require a different default port, please file an issue, and that port will be considered for inclusion.
Subcommands
discover - print service protobuf contract
grpc-client-cli discover localhost:5050
grpc-client-cli -s User discover localhost:5050
health - call health check service, this command returns non-zero exit code in case health check returns NOT_SERVING
response or the call fails for any other reason, so it's useful for example in kubernetes health probes
grpc-client-cli health localhost:5050
grpc-client-cli --address localhost:5050 health
Non-interactive mode
In non-interactive mode grpc-client-cli
expects all parameters to be passed to execute gRPC service.
Pass message json through stdin
echo '{"user_id": "12345"}' | grpc-client-cli -service UserService -method GetUser localhost:5050
cat message.json | grpc-client-cli -service UserService -method GetUser localhost:5050
On windows this could be achieved using type
command
type message.json | grpc-client-cli -service UserService -method GetUser localhost:5050
Input file
Another option of providing a file with message json is -input
(or -i
) parameter:
grpc-client-cli -service UserService -method GetUser -i message.json localhost:5050
Autocompletion
To enable autocompletion in your terminal add the following commands to your .bashrc
or .zshrc
files.
ZSH
PROG=grpc-client-cli
_CLI_ZSH_AUTOCOMPLETE_HACK=1
source autocomplete/zsh_autocomplete
Bash
PROG=grpc-client-cli
source autocomplete/bash_autocomplete
autocomplete
directory is located in the root of the repo. Please find more details here.
JSON format specifics
Most of the fields in proto message can be intuitively mapped to json
types. There are some exclusions though:
Timestamp
mapped to a string inISO 8601
format.
For example:
{
"flight_start_date": "2018-03-19T00:00:00.0Z"
}
Duration
mapped to a string representation in seconds.
For example:
{
"start_time": "72000s",
"some_other_duration": "1s"
}