mirror of
https://github.com/LizardByte/Sunshine.git
synced 2025-01-29 18:32:38 +00:00
0b70d46bef
Some checks are pending
CI / GitHub Env Debug (push) Waiting to run
CI / Setup Release (push) Waiting to run
CI / Setup Flatpak Matrix (push) Waiting to run
CI / Linux Flatpak (push) Blocked by required conditions
CI / Linux ${{ matrix.type }} (-DSUNSHINE_BUILD_APPIMAGE=ON, 22.04, AppImage) (push) Blocked by required conditions
CI / Homebrew (${{ matrix.os_name }}-${{ matrix.os_version }}${{ matrix.release == true && ' (Release)' || '' }}) (macos, 12) (push) Blocked by required conditions
CI / Homebrew (${{ matrix.os_name }}-${{ matrix.os_version }}${{ matrix.release == true && ' (Release)' || '' }}) (macos, 13) (push) Blocked by required conditions
CI / Homebrew (${{ matrix.os_name }}-${{ matrix.os_version }}${{ matrix.release == true && ' (Release)' || '' }}) (macos, 14) (push) Blocked by required conditions
CI / Homebrew (${{ matrix.os_name }}-${{ matrix.os_version }}${{ matrix.release == true && ' (Release)' || '' }}) (ubuntu, latest) (push) Blocked by required conditions
CI / Homebrew (${{ matrix.os_name }}-${{ matrix.os_version }}${{ matrix.release == true && ' (Release)' || '' }}) (ubuntu, latest, true) (push) Blocked by required conditions
CI / Macports (macOS-${{ matrix.os_version }}) (12, true) (push) Blocked by required conditions
CI / Macports (macOS-${{ matrix.os_version }}) (13) (push) Blocked by required conditions
CI / Macports (macOS-${{ matrix.os_version }}) (14) (push) Blocked by required conditions
CI / Windows (push) Blocked by required conditions
CI Docker / Check Dockerfiles (push) Waiting to run
CI Docker / Setup Release (push) Blocked by required conditions
CI Docker / Lint Dockerfile${{ matrix.tag }} (push) Blocked by required conditions
CI Docker / Docker${{ matrix.tag }} (push) Blocked by required conditions
CodeQL / Get language matrix (push) Waiting to run
CodeQL / Analyze (${{ matrix.name }}) (push) Blocked by required conditions
Build GH-Pages / update_pages (push) Waiting to run
Co-authored-by: ReenigneArcher <42013603+ReenigneArcher@users.noreply.github.com>
98 lines
3.9 KiB
ReStructuredText
98 lines
3.9 KiB
ReStructuredText
General
|
|
=======
|
|
|
|
Forgotten Credentials
|
|
---------------------
|
|
If you forgot your credentials to the web UI, try this.
|
|
.. tab:: General
|
|
|
|
.. code-block:: bash
|
|
|
|
sunshine --creds {new-username} {new-password}
|
|
|
|
.. tab:: AppImage
|
|
|
|
.. code-block:: bash
|
|
|
|
./sunshine.AppImage --creds {new-username} {new-password}
|
|
|
|
.. tab:: Flatpak
|
|
|
|
.. code-block:: bash
|
|
|
|
flatpak run --command=sunshine dev.lizardbyte.app.Sunshine --creds {new-username} {new-password}
|
|
|
|
|
|
Web UI Access
|
|
-------------
|
|
Can't access the web UI?
|
|
#. Check firewall rules.
|
|
|
|
Controller works on Steam but not in games
|
|
------------------------------------------
|
|
One trick might be to change Steam settings and check or uncheck the configuration to support Xbox/Playstation
|
|
controllers and leave only support for Generic controllers.
|
|
|
|
Also, if you have many controllers already directly connected to the host, it might help to disable them so that the
|
|
Sunshine provided controller (connected to the guest) is the "first" one. In Linux this can be accomplished on USB
|
|
devices by finding the device in `/sys/bus/usb/devices/` and writing `0` to the `authorized` file.
|
|
|
|
Network performance test
|
|
------------------------
|
|
For real-time game streaming the most important characteristic of the network
|
|
path between server and client is not pure bandwidth but rather stability and
|
|
consistency (low latency with low variance, minimal or no packet loss).
|
|
|
|
The network can be tested using the multi-platform tool `iPerf3 <https://iperf.fr>`__.
|
|
|
|
On the Sunshine host ``iperf3`` is started in server mode:
|
|
|
|
.. code-block:: bash
|
|
|
|
iperf3 -s
|
|
|
|
On the client device iperf3 is asked to perform a 60-second UDP test in reverse
|
|
direction (from server to client) at a given bitrate (e.g. 50 Mbps):
|
|
|
|
.. code-block:: bash
|
|
|
|
iperf3 -c {HostIpAddress} -t 60 -u -R -b 50M
|
|
|
|
Watch the output on the client for packet loss and jitter values. Both should be
|
|
(very) low. Ideally packet loss remains less than 5% and jitter below 1ms.
|
|
|
|
For Android clients use `PingMaster <https://play.google.com/store/apps/details?id=com.appplanex.pingmasternetworktools>`__.
|
|
|
|
For iOS clients use `HE.NET Network Tools <https://apps.apple.com/us/app/he-net-network-tools/id858241710>`__.
|
|
|
|
If you are testing a remote connection (over the internet) you will need to
|
|
forward the port 5201 (TCP and UDP) from your host.
|
|
|
|
Packet loss (Buffer overrun)
|
|
----------------------------
|
|
If the host PC (running Sunshine) has a much faster connection to the network
|
|
than the slowest segment of the network path to the client device (running
|
|
Moonlight), massive packet loss can occur: Sunshine emits its stream in bursts
|
|
every 16ms (for 60fps) but those bursts can't be passed on fast enough to the
|
|
client and must be buffered by one of the network devices inbetween. If the
|
|
bitrate is high enough, these buffers will overflow and data will be discarded.
|
|
|
|
This can easily happen if e.g. the host has a 2.5 Gbit/s connection and the
|
|
client only 1 Gbit/s or Wifi. Similarly a 1 Gbps host may be too fast for a
|
|
client having only a 100 Mbps interface.
|
|
|
|
As a workaround the transmission speed of the host NIC can be reduced: 1 Gbps
|
|
instead of 2.5 or 100 Mbps instead of 1 Gbps. (A technically more advanced
|
|
solution would be to configure traffic shaping rules at the OS-level, so that
|
|
only Sunshine's traffic is slowed down.)
|
|
|
|
Sunshine versions > 0.23.1 include improved networking code that should
|
|
alleviate or even solve this issue (without reducing the NIC speed).
|
|
|
|
Packet loss (MTU)
|
|
-----------------
|
|
Albeit unlikely, some guests might work better with a lower `MTU
|
|
<https://en.wikipedia.org/wiki/Maximum_transmission_unit>`__ from the host. For example, a LG TV was found to have 30-60%
|
|
packet loss when the host had MTU set to 1500 and 1472, but 0% packet loss with a MTU of 1428 set in the network card
|
|
serving the stream (a Linux PC). It's unclear how that helped precisely so it's a last resort suggestion.
|