ossrs / oryx Goto Github PK

See https://blog.ossrs.io/how-to-setup-a-video-streaming-service-by-1-click-e9fe6f314ac6

Hi,

I'm family with GG cloud, so I want to install the srs-cloud on GG cloud instead Digital Ocean! I tried to GG search for the guideline but haven't got anything...

Appreciate someone would shares the way to setup on GG cloud...

Lighthouse has commands for logging in and viewing passwords. Baota users are generally used to getting the interface, so you can add a page to display the login password on Baota.

Before supporting this feature, you can use this command to get it first:

cat /usr/local/srs-cloud/mgmt/.env |grep MGMT_PASSWORD

In a virtual machine environment, SRS is installed in BaoTa, and virtual live streaming cannot upload files, prompting an error.

The virtual machine is installed with Ubuntu 20.0, and then the local browser installs BaoTa and SRS. When uploading local files in the virtual live streaming section, an error is prompted.
You got an error!
Request: http://192.168.161.100/terraform/v1/ffmpeg/vlive/upload/attach-file-0
Status: 500 Internal Server Error
Data: filename=2.mp4 mismatch attach-file-0

After the recording is completed, call back the user's API to enable the user to manage their own recording tasks.

During automatic upgrades, if the SRS image is also updated, it will cause the SRS server to restart and update.

A better method is to support specifying the upgrade time period.

In addition, it is necessary to enable SRS's Gracefully Quit feature for a smooth exit to avoid affecting the business.

Describe the bug
A clear and concise description of what the bug is.

v5.10.11
Desribe your SRS Stack version here.
docker run -it --rm -v /home/live/srs-stack/data:/data -p 2022:2022/tcp -p 443:2443/tcp -p 1935:1935/tcp -p 1985:1985/tcp -p 80:8080/tcp -p 8000:8000/udp -p 10080:10080/udp -it --name srs-stack ossrs/srs-stack:v5.10.11

To Reproduce
Steps to reproduce the behavior:

1. When accessing via https://, the displayed address is 127.0.0.1 and it is followed by http.
2. Certificate application shows a 500 error.
3. The player page fails to load, unable to play.
4. See error

Expected behavior
A clear and concise description of what you expected to happen.

Screenshots

Additional context
Add any other context about the problem here.

TRANS_BY_GPT4

Describe the bug
I try to get SSL on page /mgmt/en/routers-settings?tab=https , it showed error message:

You got an error!
Request: http://**.***.**.***:2022/terraform/v1/mgmt/letsencrypt
Status: 500 Internal Server Error
Data: updateSslFiles domain=live.75.*****.cn: run 
lego [--email [email protected] --domains live.75.*****.cn 
--http.webroot /usr/local/srs-stack/platform/containers/data --http 
--accept-tos run] in /usr/local/srs-stack/platform/containers/data/lego,

stdout , 
stderr 
2023/09/24 11:25:36 [INFO] [live.75.*****.cn] acme: Obtaining bundled SAN certificate 
2023/09/24 11:25:37 [INFO] [live.75.*****.cn] AuthURL: https://acme-v02.api.letsencrypt.org/acme/authz-v3/267515225066 
2023/09/24 11:25:37 [INFO] [live.75.*****.cn] acme: Could not find solver for: tls-alpn-01 
2023/09/24 11:25:37 [INFO] [live.75.*****.cn] acme: use http-01 solver 
2023/09/24 11:25:37 [INFO] [live.75.*****.cn] acme: Trying to solve HTTP-01 
2023/09/24 11:25:41 [INFO] Deactivating auth: https://acme-v02.api.letsencrypt.org/acme/authz-v3/267515225066 
2023/09/24 11:25:41 Could not obtain certificates: error: one or more domains had a problem: [live.75.*****.cn] acme: error: 400 :: urn:ietf:params:acme:error:connection :: **.***.**.***: 

Fetching http://live.75.*****.cn/.well-known/acme-challenge/hzYJ4mhAkEa62O9DJtdy0qctchO0pM1D-YDx0FZMzv4: 
Connection refused : exit status 1

thinks!

Version
©srs-stack/v5.10.2 (docker)

To Reproduce
Steps to reproduce the behavior:

1. Go to '...'
2. Click on '....'
3. Scroll down to '....'
4. See error

Expected behavior
A clear and concise description of what you expected to happen.

Screenshots
If applicable, add screenshots to help explain your problem.

Additional context
Add any other context about the problem here.

Supports merging two streams, usually one audio stream and one video stream.

For example, many slow live video sources nowadays come from cameras and have no sound. If I want to personalize it, I need to match different music to each video frame.

For v1.0.292+

Run srs-cloud in one docker, so please mount volumes if want to save data to local disk:

docker run --rm -it -p 2022:2022 -p 1935:1935/tcp \
  -p 8080:8080/tcp -p 8000:8000/udp -p 10080:10080/udp -v $HOME/data:/data \
  --name srs-cloud ossrs/srs-cloud:5

Note: Please use registry.cn-hangzhou.aliyuncs.com/ossrs/srs-cloud in China.

Then open http://localhost:2022/mgmt in browser.

The volumes for srs-cloud:

• /data The global data directory.
  • redis The redis data directory, the publish secret and record configuration.
  • config The mgmt password and cloud configuration.
  • record The record storage directory, save record files.
  • vlive The storage directory for virtual live, save video files.

You can change the volumes to other directories.

Currently, Cloud SRS supports single-stream retransmission, with a simple overall solution and easy operation.

However, if multiple streams are pushed to SRS, only one stream can be randomly selected, which can achieve hot backup.

But some friends have reported that hot backup is not that important. Supporting multi-stream retransmission is more useful, and currently, multiple Cloud SRS instances need to be opened, making management cumbersome.

We need to carefully consider the following:

1. Some people have single-stream requirements, so we cannot increase operational complexity.

2. Whether there is a hot backup requirement, if hot backup is needed, supporting multi-stream retransmission becomes very complex.

3. If we consider transcoding in the future, it's hard to imagine, as it would be extremely complex.

If you have used multi-stream retransmission services, please feel free to share your thoughts.

WordPress plugin to play stream of SRS:

• HLS plugin.
• HTTP-FLV plugin.
• WebRTC plugin.

Usage: https://github.com/ossrs/WordPress-Plugin-SrsPlayer

Currently, when using the SRS Stack to record audio or video content, there is no option to set a maximum recording time. This can be problematic when we need files of specific durations for various applications, such as podcast episodes, video clips, or other media content. Therefore, I kindly request the implementation of a "Maximum Length" setting for recorded files within the SRS Stack.

I would like to ask for guidance on two issues I encountered while using a clean CentOS 7_2009 with a clean BaoTa panel 7.9.8 to install srs-cloud:

1. After the deployment is complete, the first time I open the management page, it asks me to set an initial password. I cannot click "Set Administrator Password" using either Chrome or Edge browsers.

2. After the server restarts, port 2022 cannot be found, causing the srs-cloud:platform-v1.0.232 in Docker to start with an error.

Currently, the HLS cache is not cleaned up, causing the playback stream to persist, making it seem like the live broadcast hasn't ended.

SRS supports deleting HLS files after stopping the push stream for a certain period of time, but this feature is turned off by default.

In the business scenario of camera recording, if multiple cloud SRS share a storage, it can be more convenient to manage.

TRANS_BY_GPT4

For machines located overseas, accessing updates from the Gitee repository can be slow, so it is recommended to connect to the GitHub repository instead.

An automatic detection can be implemented to determine which source is faster and fetch from that source.

FAQ

Before submitting any issue or feature request, please ensure to search through existing issues, discussions, and discord channels.

Please choose the language for the FAQ:

• English: https://ossrs.io/lts/en-us/faq-srs-cloud
• Chinese: https://ossrs.net/lts/zh-cn/faq-srs-cloud

Features:

• English: https://ossrs.io/lts/en-us/faq-srs-cloud#features
• Chinese: https://ossrs.net/lts/zh-cn/faq-srs-cloud#features

Features:

• English: https://ossrs.io/lts/en-us/faq-srs-cloud#changelog
• Chinese: https://ossrs.net/lts/zh-cn/faq-srs-cloud#changelog

Community

If you encounter any issues, feel free to reach out to us on Discord. Please be aware that all communications on Discord must be in English, as we are a worldwide open source community.

中文用户，可以加微信群（Chinese users are also welcome to join the WeChat group）：

TRANS_BY_GPT3

What is the business background? Please provide a description.
We have a scenario where multiple srt streams are received.
Currently we output those streams to a SDI matrix and have the content available on our facilities.
In case we wish to forward a srt stream that was received via srs-stack, we are reingesting the content via sdi, reencoding, repackaging and pushing the stream using 3rd party tools.

Is your feature request related to a problem? Please describe.
Ability to forward streams via srt protocol to 3rd parties via internet.

Describe the solution you'd like
It is currently possible to easily forward streams via rtmp{s}, it would be great to have the possibility to push via srt.

Describe alternatives you've considered
ffmpeg can do this:

ffmpeg -i srt://127.0.0.1:1234?mode=listener -v:c -a:c -f mpegts "srt://127.0.0.1:4321=mode?caller"

ARM64 architecture servers, the docker [redis] created by the plugin cannot be started, and the log reports an exec format error. This causes the console to fail to start.
https://hub.docker.com/r/ossrs/redis This image only has one AMD64 tag,

Among these images, only AMD64 tags are available for all except srs. I hope to add support for ARM64 to make deployment more convenient.

1. Background on using SRS-Stack: Our company has a camera that records non-stop, 24 hours a day. We need to record and store the audio and video signals locally using SRS-Stack.

2. Problem encountered: In the recording mode of SRS Stack, the local storage space is limited. What happens when the storage space is full? Will SRS Stack continue recording in a loop, or will it stop recording?

3. Suggested solution: Can we set up a recording mode, like a loop mode, or have it stop and send an alert when the storage space is full?

Similarly: How does SRS-Stack handle things in the cloud? If a user's cloud recording space is limited, what happens when the effective space for cloud recording reaches zero?

TRANS_BY_GPT4

Describe the bug
502 error occurred after modifying the '/data/record' directory using 'ln'.
A 502 error occurred after modifying the "/data/record" directory using "ln".

Version

• OS: Ubuntu 20.04.6 LTS x86_64(Py3.7.9)
• Pagoda: Linux Official Version (Free Version) 8.0.2 (2023/08/23)
• srs-stack/v5.8.31

To Reproduce

1. Install the system, install Pagoda.
2. Pagoda homepage pop-up box installs Nginx 1.22.1, MySQL 5.7.43, PHP-7.4.33, phpMyAdmin 5.2, etc.
3. After success, install Docker Manager 3.9.2 from the Pagoda software store.
4. After success, install SRS Audio and Video Server from the Pagoda software store, click on settings => input domain name => install, wait for completion.
5. Pagoda homepage scans a bunch of "security risks" and fixes them according to the prompts.
6. After clicking on "SRS Stack Management Panel" in the "SRS Audio and Video Server" panel, enter the password to log in to /mgmt/zh/routers-scenario normally.
7. Application Scenario => Recording => Set Recording Rules => Check "Record All Streams" and submit.

• Phenomenon 1:

8. After setting the streaming address with EV screen recording, start streaming, and stop streaming after about 1 minute.
9. Relevant recording files can appear in the system "/data/record" directory, this phenomenon is normal.

• Phenomenon 2:

10. After setting the "Service Status" to "Stop" in the "SRS Audio and Video Server" panel,
11. In the Baota "File", cut and paste the "/data/record" directory as the "/www/wwwroot/srsDir/record" directory,
12. In the Baota "File", use the Baota's "New => Soft Link File" function in the "/data/" directory to create a "record" link to "/www/wwwroot/srsDir/record".
13. After setting the "Service Status" to "Start" in the "SRS Audio and Video Server" panel,
14. Click on the "SRS Stack Management Panel" to open the mgmt page, an error message "You got an error!" appears,
15. Click the browser refresh button to refresh the page, a "502" error message appears, other pages also have errors,
16. Start streaming with EV screen recording, an error message "Error number -10053 occurred" appears indicating streaming failure.

• Phenomenon 3:

17. After setting the "Service Status" to "Stop" in the "SRS Audio and Video Server" panel,
18. In the Baota "File", delete the "/data/record" directory and cut and paste the "/www/wwwroot/srsDir/record" directory back to the "/data/record" directory.
19. In the Baota "File", cut and paste the "/data" directory as the "/www/wwwroot/srsDir/data" directory,
20. In the Baota "File", use the Baota's "New => Soft Link File" function in the "/" directory to create a "data" link to the "/www/wwwroot/srsDir/data" directory.
21. After setting the "Service Status" to "Start" in the "SRS Audio and Video Server" panel,
22. Click on the "SRS Stack Management Panel" to open the mgmt page normally, refresh the page normally, and click on other pages normally,
23. Streaming with EV screen recording is normal, stop streaming after about 1 minute,
24. In the Baota "File", there is a recorded folder in the "/www/wwwroot/srsDir/data/record" directory and the recorded video files can be played normally.

Expected behavior
The 14th, 15th, and 16th steps of the expected Phenomenon 2 should be consistent with the 22nd, 23rd, and 24th steps of Phenomenon 3.

Screenshots
Screenshot of Step 12:

Screenshot of Step 13:

Screenshot of Step 14:

Screenshot of Step 15:

Screenshot of Step 16:

TRANS_BY_GPT4

Baota has installed SRS;
During the push stream, it was found that the server CPU could not break through 30%. After further analysis, is it because the push stream to SRS is limited by the CPU?
Where can I change the CPU traffic limit value?

Add a scene for a wedding live broadcast, which will be divided into several segments. The first day, the afternoon will be live for a few hours, the evening will be live for a few hours, and the next day, the wedding will be live for a few more hours. After each live broadcast, a replay video of the current live broadcast can be automatically generated. The end of the live broadcast can be manually generated, and all replay videos from the first day to the end can be generated. It is best to set the start and end time points, which is very convenient to use.

Add a live streaming feature for a wedding event, which will be divided into several sessions. The live stream will occur for a few hours in the afternoon and a few hours in the evening the day before the wedding, and then for a few hours on the wedding day itself. After each live streaming session ends, a replay video of that session should be automatically generated. The live stream can be manually ended, and upon completion, a compilation of all replay videos from the first day to the last should be created. Ideally, the ability to set start and end times for the live stream would be included, making the feature very convenient to use.

TRANS_BY_GPT4

Describe the bug

Version
docker run --rm -p 2022:2022 -p 2443:2443 -p 1935:1935/tcp -p 1985:1985/tcp
-p 8080:8080/tcp -p 8000:8000/udp -p 10080:10080/udp --name srs-stack
-v $HOME/db:/data ossrs/srs-stack:5

To Reproduce

1. docker run --rm -p 2022:2022 -p 2443:2443 -p 1935:1935/tcp -p 1985:1985/tcp
   -p 8080:8080/tcp -p 8000:8000/udp -p 10080:10080/udp --name srs-stack
   -v $HOME/db:/data ossrs/srs-stack:5

2. Use OBS To Set SRT Addr:
   

3. Press "Begin Living" In OBS

4. See Connect Error In OBS and The Error Msg Above : srt serve error code=6006

Expected behavior
A bout 2 month ago the srt living feature is work.

Screenshots
None

Additional context
None

Operating system: Ubuntu 22.04.1 LTS (pure)
Platform: BaoTa panel (pure) + SRS plugin

On the /mgmt/zh/routers-scenario?tab=live page, click "Play WebRTC Stream" and get a 404 error.
Try changing the players symlink in the BaoTa panel directory to players2, and this page can be opened.

Solution:
In the /usr/local/lighthouse/softwares/srs-cloud/mgmt/containers/conf/default.d/20.srs.conf file,
change location ~ ^/(console|players)/ { to location ~ ^/(console)/ {

You encountered an error!
Name: NotFoundError message
: Cannot execute "insertBefore" on "Node": The node before which the new node is to be inserted is not a child of this node.
Error: Failed to execute 'insertBefore' on 'Node': The node before which the new node is to be inserted is not a child of this node.
at kl (http://8.218.33.142/mgmt/static/js/main.d54a682c.js:2:1141971)
at kl (http://8.218.33.142/mgmt/static/js/main.d54a682c.js:2:1142044)
at wl (http://8.218.33.142/mgmt/static/js/main.d54a682c.js:2:1141469)
at Tc (http://8.218.33.142/mgmt/static/js/main.d54a682c.js:2:1154228)
at t.unstable_runWithPriority (http://8.218.33.142/mgmt/static/js/main.d54a682c.js:2:1245445)
at Ki (http://8.218.33.142/mgmt/static/js/main.d54a682c.js:2:1097521)
at Fc (http://8.218.33.142/mgmt/static/js/main.d54a682c.js:2:1152389)
at xc (http://8.218.33.142/mgmt/static/js/main.d54a682c.js:2:1148475)
at http://8.218.33.142/mgmt/static/js/main.d54a682c.js:2:1097744
at t.unstable_runWithPriority (http://8.218.33.142/mgmt/static/js/main.d54a682c.js:2:1245445)

This error occurs after multi-platform forwarding. What is the reason?

This includes images and comments.

The current version of the SRS (Simple Realtime Server) stack is experiencing issues where the M3U8 files from live broadcast recordings, when called by other programs, are not playable on Apple and WeChat PC clients. Please address this problem.

TRANS_BY_GPT4

After testing, the console /console/ng_index.html#/vhosts?port=[port]&http=[port] can be used without login, with permissions including kicking streams, previewing, etc. Developers are advised to assess the risks.

For example:

process.env.NODE_ENV === 'development' ? 'localhost' : 'mgmt.srs.local'

Tencent Cloud Lighthouse has a WordPress application image, which makes it very convenient to set up WordPress. It comes with a Baota, which cannot install Cloud SRS, possibly due to some compatibility issues.

Because this image sets WordPress as the default site, and Cloud SRS also needs to be set as the default site, it cannot be installed. I can modify it, but if you want to install it, you need to set a subdomain for Cloud SRS to use.

Workaround:

1. After installing the WordPress image, set the default site to empty in the site settings.
2. Use Baota to install the Cloud SRS application plugin, and the default site will be set to Cloud SRS.
3. Change the default site to WordPress and add a subdomain for Cloud SRS at the same time.

Of course, you can also keep the default site as Cloud SRS and add a subdomain for WordPress. The point is that there can only be one default site or none, but both WordPress and Cloud SRS can be accessed through subdomains:

1. You can set the default site as Cloud SRS, in which case you must set a subdomain for WordPress to access.
2. You can set the default site as WordPress, in which case you must set a subdomain for Cloud SRS to access.
3. You can set the default site to empty, in which case both WordPress and Cloud SRS must have subdomains.

As long as you have a domain, you can have an unlimited number of subdomains, and the number of subdomains is not a problem.

The APIs and console control panel of SRS are all exposed to the public network, allowing others to arbitrarily kick the stream. This issue also exists in SRS4, SRS5, and SRS6. I don't understand why there isn't a login authentication added.

Run cloud SRS on the company's intranet server, record to the local disk, and the local disk is large enough.

There are some known issues with cloud SRS local recording:

1. If there are many streams, the disk will be very busy, especially when using shared storage. It is necessary to monitor disk IO and load. Solution: Use high-performance disks and shared storage.
2. Although the local disk is large enough, cloud storage is truly unlimited, while the local disk is still limited. It is necessary to monitor disk space. Solution: Expand disk space.
3. Temporary support for local file management, such as deletion and cleanup, is not available. Solution: You can use server management tools like BaoTa.

Is it possible to support API access to local file playback addresses?

The live streaming transcoding function is excellent, thank you to the developers. I have encountered two minor issues while using the Docker version:

1. It only supports one-way transcoding.
2. If two or more live streams are pushed to the server simultaneously after transcoding is enabled, the Docker displays red error messages.

If srs-cloud supports aaPanel, we're able to install on any Linux OS or cloud platform, or even private virtual server.

For english user, please install aaPanel first, then follow this post

Supporting Baota means supporting any Linux and cloud platform, including private cloud.

Baota users, please first install Baota, and then follow the guide for operation.

If you need to deploy on overseas servers, please use aaPanel instead of Baota.

TRANS_BY_GPT3

The panel is quite easy to use, can it be deployed in your own environment?

Currently, the stream is fixed as livestream. If you want to support multiple streams, you need to manually change the stream name. One-click stream name change can be supported.

Supports video padding. If the stream stops pushing, it switches to the video file, and automatically switches back when the stream resumes.

Instead of using MGMT PASSWORD, use UUID tokens to keep the API and login state separate.

How to fill in the vMix push stream address and playback address? It would be better to have some guidance.

Failed to pull the image, image not found.

root@instance-20221210-1433:~# docker pull registry.cn-hangzhou.aliyuncs.com/ossrs/srs-cloud:1 Error response from daemon: manifest for registry.cn-hangzhou.aliyuncs.com/ossrs/srs-cloud:1 not found: manifest unknown: manifest unknown

TRANS_BY_GPT3

Docker deployment on Ubuntu, virtual live broadcast is very useful, but encountered several problems:

I don't know how to delete the files uploaded for virtual live broadcast.
A file needs to be virtually live broadcasted to several platforms, and it needs to be uploaded repeatedly, which takes up a lot of space. Can you also choose the server's local file method besides uploading?

Currently, the cloud recording and cloud on-demand features of Cloud SRS can only be used on Tencent Cloud's virtual machines, whether it's Lighthouse, CVM, or Baota-managed virtual machines.

The main reason is to support intranet transmission, which means that writing to cloud recording and cloud on-demand by Cloud SRS is free of traffic charges, with high-quality intranet transmission and high bandwidth.

If you use other clouds or your own intranet machines, then recording the stream to Tencent Cloud will incur traffic charges and use the downstream bandwidth of Cloud SRS.

However, this feature is necessary, as it can directly record to the cloud. Some adaptations need to be made, and there may be additional charges involved.

That is: you can select one or more files from a specific directory on the server's local machine.

Supports full system initialization, obtaining tokens, and calling APIs through API integration.