RTC Forums

Subscription => Support => Topic started by: mkurnia on April 15, 2016, 04:29:09 PM



Title: how to use RtcDualDataClientLink
Post by: mkurnia on April 15, 2016, 04:29:09 PM
hi danijel,


I have two addresses connected to the server with the address 192.168.100.10 and 192.168.100.20. I've tried using 2 rtchttpclient and 1 RtcDualDataClientLink. My goal, when the server with the address 192.168.100.10 out for connections. then RtcDualDataClientLink will use the server address 192.168.100.20. but it did not happen. Please you advise

thankyou


Title: Re: how to use RtcDualDataClientLink
Post by: D.Tkalcec (RTC) on April 15, 2016, 05:01:10 PM
Here is a copy/paste of the TRtcDualDataClientLink component description:

You can use TRtcDualDataClientLink components to create a small pool of client connections. When posting requests through this component, you will never really know which connection the request will be posted from, since the component itself will determine that, depending on the number of currently active requests on each connection component.

In other words ... when posting a request through a TRtcDualDataClientLink component, the request will be placed into the request queue of the Client component with the smallest number of currently pending requests. This simplifies working with multiple connections going to the same Server, but is NOT really suitable for working with multiple Servers, unless ALL of your Servers are ALWAYS accessible and interchangeable, so it does NOT matter to which Server which request would go.

If you want your Client to work with multiple Servers and switch from one Servers to the next if the previous one became inaccessible, you have to implement this manually. There are NO components in the RTC SDK which would work with multiple Servers and automatically switch from one Server to the next, based on Server accessibility.

PS. When working with multiple interchangeable Servers, the most common scenario is to place a Load Balancer between your Clients and your Servers. Your Clients would be connecting to your Load Balancer (one Address) and the Load Balancer would manage connections to your Servers (any number of connections to any number of Servers), keep track of each Servers status (accessibility and load) and make decisions about which request should be forwarded to which Server.

Best Regards,
Danijel Tkalcec