Welcome to Gorox!
Gorox is a pragmatic Network Proxy, Webapp Server, and RPC Framework. It can be used as a:
- Web Reverse Proxy (HTTP, TLS, WebSocket, Caching, Load Balancing)
- Layer 7 Reverse Proxy (Various Protocols, with or without Load Balancing)
- Layer 4 Reverse Proxy (TCP, UDP, QUIC, TLS, UDS, Load Balancing)
- Web Server (HTTP, TLS, WebSocket, static, FCGI, uwsgi)
- Go Web Application Server (Frameworks, Applications)
- RPC Framework for Go (gRPC, HRPC)
- ... and more through its highly extensible compoments design!
Gorox is currently under heavy development. For more details about Gorox, please see our project site: https://gorox.org/.
Gorox works on these operating systems:
- Linux kernel >= 3.9
- Microsoft Windows >= 10
- Apple macOS >= Catalina
- FreeBSD >= 12.0
And these 64-bit CPU architectures:
- AMD64, a.k.a. x86-64 or x64
- ARM64, a.k.a. AArch64
- RISCV64, a.k.a. RV64
- Loong64, a.k.a. LoongArch64
Other platforms are currently not tested and probably don't work.
If you would like to use Gorox as a Network Proxy, you can download the official binary distribution and read the "Start and stop Gorox" section below. But if you prefer to build it from source, please read below.
When using Gorox as a Webapp Server or RPC Framework, you have to build it from source. Before building, please ensure you have Go >= 1.20 installed:
shell> go version
Then download the source code tarball, uncompress it, and build it with Go:
shell> cd gorox-x.y.z
shell> go build
If build failed, set CGO_ENABLED to 0 and try again:
shell> go env -w CGO_ENABLED=0
shell> go build
On succeed, a single "gorox" or "gorox.exe" binary will be generated.
Now, you can try to change some webapps under "apps/examples/", and build again. To follow Gorox's convention, put your new Web applications under "apps/" and your new RPC services under "svcs/". If you need to write components of your own, put them into specific sub directories under "exts/".
After you have downloaded and uncompressed the official binary distribution, or have successfully built the binary from source, you can run Gorox as a daemon (or remove the "-daemon" option if you don't like to run it as a daemon):
shell> ./gorox serve -daemon
Then ensure the leader and the worker process have both been started:
shell> ./gorox pids
Now visit http://localhost:3080 to see if it works. To exit server gracefully:
shell> ./gorox quit
For more actions and options, run:
shell> ./gorox help
To install Gorox, simply move the whole Gorox directory to where you like. You may also add the directory to your $PATH so you can run "gorox" without "./".
To uninstall, simply remove the whole Gorox directory and remove it from $PATH.
We provide some example configs for Gorox to use, see them under conf/examples. For example, if you are using Gorox as an HTTP reverse proxy, there is a demo config in conf/examples/http_proxy.conf, you can modify it and start Gorox like:
shell> ./gorox -daemon -config conf/examples/http_proxy.conf
To be written.
Gorox is fast. You can use your favorite HTTP benchmarking tool (like wrk) to perform a benchmark against the following URLs:
Generally, the result is about 80% of nginx and slightly faster than fasthttp.
View Gorox documentation online:
- English version: https://gorox.org/docs
- Chinese version: https://www.gorox.org/docs
Or view locally (ensure your local Gorox is started):
- English version: http://gorox.net:3080/docs
- Chinese version: http://www.gorox.net:3080/docs
By default, Gorox uses these directories:
- apps/ - Place your Web applications (including source code),
- bins/ - Place source code of your auxiliary commands,
- conf/ - Place configs for Gorox and your commands,
- docs/ - Place docs of your project,
- exts/ - Place extended components written for your project,
- hemi/ - The Hemi Engine,
- libs/ - Place libs written or generated by you for your project,
- misc/ - Place misc resource of your project,
- svcs/ - Place your RPC services,
- test/ - Place tests for your project.
After Gorox is started, an extra directory called "data/" will be created, with 3 sub directories in it:
- data/log/ - Place running logs,
- data/tmp/ - Place files which are safe to remove after Gorox is shutdown,
- data/var/ - Place dynamic data files used by Gorox.
A typical deployment architecture using Gorox might looks like this:
mobile pc iot
| | |
| | | public internet
| | |
v v v
+------------+
+------------| edgeProxy1 |--------------+ gorox cluster
| +--+---+--+--+ |
| http | | | tcp |
| +--------+ | +--------+ |
| | | | |
| v rpc v |
| +------+ | +---------+ |
| | app1 +----+ | +--+ server1 | |
| +------+ | | | +----+----+ |
| | | | | | stateless layer
| v v v v |
| +------+ +-----------+ +--------+ | +------------------+
| | svc1 |<->| svcProxy1 | | proxy2 |--+-->| php-fpm / tomcat |
| +------+ +-----+-----+ +--------+ | +------------------+
| | |
| v |
| +------+ +-----------+ +--------+ |
| | svc2 |<--+ svcProxy2 | |cronjob1| |
| +------+ +-----------+ +--------+ |
| |
+-----------+------+---------+-----------+
| | |
v v v
+----------------------------------------+
| +-------+ +-----+ +--------+ |
| ... | db1 | | mq1 | | cache1 | ... | stateful layer
| +-------+ +-----+ +--------+ |
+----------------------------------------+
In this typical architecture, with various configurations, Gorox can play ALL of the roles in "gorox cluster":
- edgeProxy1: The Edge Proxy, also works as an API Gateway or WAF,
- app1 : A Web application implemented directly on Gorox,
- server1 : A TCP server implemented directly on Gorox,
- svc1 : A public RPC service implemented directly on Gorox,
- svcProxy1 : A service proxy for svc1,
- proxy2 : A gateway proxy passing requests to PHP-FPM or Tomcat server,
- svc2 : A private RPC service implemented directly on Gorox,
- svcProxy2 : A service proxy for svc2,
- cronjob1 : A background application in Gorox doing something periodically.
The whole Gorox cluster can alternatively be managed by a Myrox instance, which behaves like the control plane in Service Mesh. In this configuration, all Gorox instances in the cluster connect to Myrox and are under its control.
A Gorox instance has two processes: a leader process, and a worker process:
+----------------+ +----------------+ business traffic
cmdConn | | admConn | |<===============>
operator--------->| leader process |<------->| worker process |<===============>
| | | |<===============>
+----------------+ +----------------+
The leader process manages the worker process, which do the real and heavy work.
A Gorox instance can be controlled by operators through the cmdui interface of the leader process. Operators connect to the leader, send commands, and the leader executes the commands. Some commands are delivered to the worker process through admConn.
Alternatively, Gorox instances can connect to a Myrox instance and delegate their administration to Myrox. In this way, the cmdui interface in the leader process is disabled.
Currently Github Discussions is used for discussing:
https://github.com/hexinfra/gorox/discussions
Gorox is originally written by Zhang Jingcheng [email protected]. You can also contact him through Twitter: @diogin.
The official website of the Gorox project is at:
- English version: https://gorox.org/
- Chinese version: https://www.gorox.org/
Gorox is licensed under a 2-clause BSD License. See LICENSE file.
Gorox is hosted at Github:
https://github.com/hexinfra/gorox
Fork this repository and contribute your patch through Github Pull Requests.
By contributing to Gorox, you MUST agree to release your code under the BSD License that you can find in the LICENSE file.
Please report any security issue or crash report to:
Zhang Jingcheng [email protected]
Your issue will be triaged and coped with appropriately.
Thank you in advance for helping to keep Gorox secure!