问题描述
假设我们在A，B和B之间建立一个WebRTC会议. C.然后会有A<->. B，A<-> C，B<-> C个人来电?
Suppose we establish a WebRTC conference among A, B & C. Then will there be A <--> B, A <--> C, B <--> C individual calls?
Qn-1 :如果是的话，那效率不低吗?，因为每个对等方两次上传相同的数据.
Qn-1: If yes, then isn't that inefficient? Because each peer uploads the same data twice.
现在，Google对会议采用了不同的策略.在他们的正式文件中，内容为:
Now, Google has a different strategy for conferences. In their official document, it reads:
Qn-2 :现在，假设我们仍然采用p2p方法.现在，我们以某种方式将所有流量路由到一个端口.那么WebRTC是否对它可以同时与同一个IP和同一个端口建立多少连接有任何限制?
Qn-2: Now suppose if we go with p2p approach nevertheless. Now somehow we route all the traffic to one single port. Then is there any WebRTC limitation on how many connections can it make at a same time to a same IP and same port?
出现上述问题的原因是，如果我们通过一些黑客手段做到这一点，那么在4个传出连接之后，第5个连接将始终不稳定，并且大多会失败.
The reason for above question is that, if we do that with some hacks then after 4 outgoing connections, the 5th connection will always be unstable and mostly fail.
推荐答案
有足够的材料表明基于 P2P的网状"网络适合多达4个用户稳定的会议.
There is enough material to suggest that the P2P based "Mesh" network is suitable for up to 4 user stable conference.
-
此视频指出网络是实施会议的最幼稚方式
This video states that "Mesh" network is the most naive way to implement a conference
此链接提示随着参与者P2P的增加，带宽不堪重负，质量下降:
This link suggests that the bandwidth is overwhelmed and the quality reduces as the participants increase in P2P:
在此链接中，它表示网状"随着参与者数量的增加，网络开始出现故障.
In this link, it states that "Mesh" network starts failing as the number of participants increase.
这篇关于基于WebRTC的音频/视频会议是否有这么多的对等网络，或者所有上载内容混在一起然后发送回去?的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持！