syncthing/cmd/relaysrv
Jakob Borg 4a69f3987f cmd/relaysrv: Fix lint warnings
GitHub-Pull-Request: https://github.com/syncthing/syncthing/pull/3274
2016-06-07 07:31:43 +00:00
..
etc/linux-systemd cmd/relaysrv: Missed changes in the repo merge (README, systemd) 2016-06-02 14:42:57 +02:00
testutil Update packages, fix testutil. Goddamit godep. 2015-11-23 21:29:23 +00:00
LICENSE Initial commit 2015-06-24 00:34:16 +01:00
README.md cmd/relaysrv: Missed changes in the repo merge (README, systemd) 2016-06-02 14:42:57 +02:00
listener.go cmd/relaysrv: Copyright headers 2016-06-02 14:16:02 +02:00
main.go cmd/relaysrv: Fix lint warnings 2016-06-07 07:31:43 +00:00
pool.go cmd/relaysrv: Copyright headers 2016-06-02 14:16:02 +02:00
session.go cmd/relaysrv: Copyright headers 2016-06-02 14:16:02 +02:00
status.go cmd/relaysrv: Add number of routines 2016-06-02 14:39:19 +02:00
utils.go cmd/relaysrv: Copyright headers 2016-06-02 14:16:02 +02:00

README.md

relaysrv

Latest Build

This is the relay server for the syncthing project.

To get it, run go get github.com/syncthing/relaysrv or download the latest build from the build server.

Warnings - Read or regret

By default, all relay servers will join the default public relay pool, which means that the relay server will be availble for public use, and will consume your bandwidth helping others to connect.

If you wish to disable this behaviour, please specify -pools="" argument.

Please note that relaysrv is only usable by syncthing version v0.12 and onwards.

To run relaysrv you need to have port 22067 available to the internet, which means you might need to allow it through your firewall if you have a public IP, or setup a port-forwarding (22067 to 22067) if you are behind a router.

Furthermore, by default relaysrv will also expose a /status HTTP endpoint on port 22070, which is used by the pool servers to peek at metrics of the relaysrv, such as what are the current transfer rates, how many clients are connected, etc, etc. If you wish this information to be available, similarlly you might want to allow it through your firewall, or port-forward it (22070 to 22070) on your NAT device.

This is not mandatory for the relaysrv to function, and is used only to gather metrics and present them in the overview page of the pool server, displaying stats about the specific relay.

At the point of writing the endpoint output looks as follows:

{
    "bytesProxied": 0,
    "goArch": "amd64",
    "goMaxProcs": 1,
    "goNumRoutine": 13,
    "goOS": "linux",
    "goVersion": "go1.6",
    "kbps10s1m5m15m30m60m": [
        0,
        0,
        0,
        0,
        0,
        0
    ],
    "numActiveSessions": 0,
    "numConnections": 0,
    "numPendingSessionKeys": 2,
    "numProxies": 0,
    "options": {
        "global-rate": 0,
        "message-timeout": 60,
        "network-timeout": 120,
        "per-session-rate": 0,
        "ping-interval": 60,
        "pools": [
            "https://relays.syncthing.net/endpoint"
        ],
        "provided-by": ""
    },
    "startTime": "2016-03-06T12:53:07.090847749-05:00",
    "uptimeSeconds": 17
}

If you wish to disable the /status endpoint, provide -status-srv="" as one of the arguments when starting the relaysrv.

Running for public use

Make sure you have a public IP with port 22067 open, or make sure you have port-forwarding (22067 to 22067) if you are behind a router.

Run the relaysrv with no arguments (or -debug if you want more output), and that should be enough for the server to join the public relay pool. You should see a message saying:

2015/09/21 22:45:46 pool.go:60: Joined https://relays.syncthing.net/endpoint rejoining in 48m0s

See relaysrv -help for other options, such as rate limits, timeout intervals, etc.

Running for private use

Once you've started the relaysrv, it will generate a key pair and print an URI:

relay://:22067/?id=EZQOIDM-6DDD4ZI-DJ65NSM-4OQWRAT-EIKSMJO-OZ552BO-WQZEGYY-STS5RQM&pingInterval=1m0s&networkTimeout=2m0s&sessionLimitBps=0&globalLimitBps=0&statusAddr=:22070

This URI contains partial address of the relay server, as well as it's options which in the future may be taken into account when choosing the best suitable relay out of multiple available.

Because -listen option was not used, the relaysrv does not know it's external IP, therefore you should replace the host part of the URI with your public IP address on which the relaysrv will be available:

relay://123.123.123.123:22067/?id=EZQOIDM-6DDD4ZI-DJ65NSM-4OQWRAT-EIKSMJO-OZ552BO-WQZEGYY-STS5RQM&pingInterval=1m0s&networkTimeout=2m0s&sessionLimitBps=0&globalLimitBps=0&statusAddr=:22070

If you do not care about certificate pinning (improved security) or do not care about passing verbose settings to the clients, you can shorten the URL to just the host part:

relay://123.123.123.123:22067

This URI can then be used in syncthing as one of the relay servers.

See relaysrv -help for other options, such as rate limits, timeout intervals, etc.

Other items available in this repo

testutil

A test utility which can be used to test connectivity of a relay server. You need to generate two x509 key pairs (key.pem and cert.pem), one for the client, another one for the server, in separate directories. Afterwards, start the client:

./testutil -relay="relay://uri.of.relay" -keys=certs/client/ -join

This prints out the client ID:

2015/09/21 23:00:52 main.go:42: ID: BG2C5ZA-W7XPFDO-LH222Z6-65F3HJX-ADFTGRT-3SBFIGM-KV26O2Q-E5RMRQ2

In the other terminal run the following:

 ./testutil -relay="relay://uri.of.relay" -keys=certs/server/ -connect=BG2C5ZA-W7XPFDO-LH222Z6-65F3HJX-ADFTGRT-3SBFIGM-KV26O2Q-E5RMRQ2

Which should then give you an interactive prompt, where you can type things in one terminal, and they get relayed to the other terminal.

Relay protocol definition.

Available here

Relay client

Only used by the testutil.

Available here