Wowza Streaming Engine (Linux PAID)
Streaming multiple indoor cameras has improved remote viewing but still needs built-in AI analytics
What is our primary use case?
Wowza Streaming Cloud is used for indoor cameras, and the live stream is fed to Wowza Streaming Cloud. Then on the client side, a mobile app or multiple mobile apps can connect to Wowza Streaming Cloud and can see the streaming or can see the recorded videos. Wowza Streaming Cloud is basically the intermediate point where everything is stored or streamed.
As an example, there is a client, and the indoor camera which we have developed is installed in my home. When I want to see the streaming on my mobile app when I am away from home, I send the RTSP request over Wowza Streaming Cloud. Then Wowza Streaming Cloud is the one that sends the streaming request to the IoT camera, and so the streaming is started. Then, with two-way communication, the streaming is sent to Wowza Streaming Cloud, and then Wowza Streaming Cloud forwards the same live stream to the mobile app so I can see the live streaming. This is one of the use cases.
One of the normal use cases where we see users using Wowza Streaming Cloud is live streaming. There are use cases where an event happens and the user gets a notification, so the user starts the stream. We can ask Wowza Streaming Cloud to start streaming on the IoT camera, and it starts the camera, and we have the live stream. This is one of the most important use cases for Wowza Streaming Cloud because one of the good things about Wowza Streaming Cloud is that multiple clients can be connected and see the streaming. This is the normal use case we are using it for.
What is most valuable?
One thing is the security offered by Wowza Streaming Cloud. We can have TCP streaming with security, so one thing we can achieve with Wowza Streaming Cloud is that there is a lesser chance of spoofing. Another thing is that we have used TCP in the streaming, and basically, Wowza Streaming Cloud is performing well with the TCP packet. These are some of the good things. Obviously, the other thing is that in Wowza Streaming Cloud, we have the capacity to store the streams, so someone can see it when they want to see the recorded video. These are some of the good facilities.
Wowza Streaming Cloud is performing well and has positively impacted our organization. We have developed the IoT camera and the user response is good because with Wowza Streaming Cloud, when multiple clients are connected on the same camera, we are seeing good performance, and Wowza Streaming Cloud is sufficiently capable to handle the multiple clients. We are able to achieve what we want with the help of Wowza Streaming Cloud.
What needs improvement?
Wowza Streaming Cloud can be improved if it includes some latest use cases where we can integrate AI. As an example, Wowza Streaming Cloud can integrate the analytics facility in the stream so that someone does not need to develop its own AI analytics, but Wowza Streaming Cloud can provide this facility, and this will enhance the user experience.
Security is one of the aspects of Wowza Streaming Cloud that everybody wants. If the AI analytics can be run on the captured stream, there is no chance that we can breach some security protocol. The analytics result should be run with the LLM which is within the premises of Wowza Streaming Cloud, so there is no concern about the security.
For how long have I used the solution?
We have been using Wowza Streaming Cloud for two years.
What other advice do I have?
The end-users are happy with the performance of Wowza Streaming Cloud. They are seeing sufficient streaming performance when they connect with the same stream. They watch with multiple clients, such as multiple mobile devices or one mobile device and one laptop. The performance is good.
I would rate Wowza Streaming Cloud a seven on a scale of one to ten. Seven is the rating because it is sufficient as per the requirement, but I would suggest that since we have the satisfactory result, I give the seven because there is still a chance that Wowza Streaming Cloud can have more features which help to enhance the user experience.
Reliability and accuracy is one of the concerns regarding Wowza Streaming Cloud. In terms of the accuracy or the precision in terms of AI, if it provides a fifty to sixty percent result, then it is still good because none of the algorithms are perfect. If I use OpenCV or anything, the result is not always one hundred percent achievable, but with the help of AI, we can improve it gradually, starting with sixty percent accuracy.
I did purchase Wowza Streaming Cloud through the AWS Marketplace.
I would suggest that whenever you want to use Wowza Streaming Cloud, always start with the use case where the free tier version is available and then do some stress testing. When you are happy with Wowza Streaming Cloud, you can convert it to the real production scenario and use it. I would rate this product a seven out of ten.
Which deployment model are you using for this solution?
If public cloud, private cloud, or hybrid cloud, which cloud provider do you use?
Highly Reliable and Easy to Use
Reliable Streaming, But HA Docs and Troubleshooting Need Work
Troubleshooting is another pain point. Wowza typically runs on a VM, and when something goes wrong it's genuinely hard to investigate quickly. If you open a support ticket, by the time someone responds the issue has often already gone away on its own, so you're left waiting for it to happen again just to grab logs and dig into what's going on.
We also had a couple of issues that never got a real resolution. Support offered workarounds, try this, try that, but we never actually landed on a root cause. There are still a fair number of bugs in the product, though it does feel like it's maturing over time.
Powerful Automation, But Bugs and Cumbersome Multi-Platform Streaming
Also, streaming to alternate sources, such as social media platforms and/or YouTube, is cumbersome.
Awful Upgrades and Out-of-Memory Crashes Under Heavy Load
It provides reliable streams at medium and low usage levels.
Under heavy load the entire server can crash with a out of memory error. There is no obvious attempt to mitigate this problem; all your streams die.