Log In
Home
04/02/2020
Offline
Hi,
we are currently in the process of implementing Discovery in our servers and clients. My question is about why it is done like this.
As far is I understood, I have to have at least one LDS-ME in my network since the RegisterServer needs to be called on a known discovery server. Also all clients have to run an instance of LDS-ME in order to retrieve discovery information. Can someone please explain why this makes sense over just using mDNS between servers and clients? I understand that embedded systems are allowed to share their information via mDNS but this would still require LDS-ME on all clients, which in my case, are also embedded systems. What is recommended to do in such a case?
Thanks,
Markus
05/30/2017
Offline
The LDS-ME is only needed for systems where there are multiple applications (potentially from different vendors) running such as a workstation. For embedded systems with a single server then there is no LDS and the server should use the mDNS directly to announce itself on the network).
The same logic would apply to embedded devices with a single client.
IOW, the LDS-ME is an implementation convenience for vendors that find it convenient. Any vendor is free to replicate the capabilities if it makes sense for them.
1 Guest(s)
Usage Policy