This message was deleted.

Currently there is something called direct workload connection which handles this in the future there will be a built in option from HDX. https://docs.citrix.com/en-us/citrix-workspace/optimize-cvad/workspace-network-location.html

Tech preview, seemingly. HDX Direct (Preview) | Citrix DaaS

Imp sorry, so the first one direct connect is not in preview and you can set it now. But the second one has to do just with HDX and is in tech preview? So in essence you can kind of have what I am looking for but not the HDX component?

Not sure what your asking but both options will provide direct connections fron endpoints to vdas. The first is ga now. The other is not quite ga yet.

Basically trying to understand the difference. I would assume that the Direct workload connection tech would encompass everything in the stream including HDX. But I guess not and we have to wait for HDX for the tech preview

No they are the same. The HDX Direct dors not require registering your networks in citrix cloud. It will just try a direct connection first and only use the gateway service if it cant connect.

Isn't this what the rendezvous protocol does? https://docs.citrix.com/en-us/citrix-daas/hdx-transport/rendezvous-protocol.html

Oh, I think i get it. At least for the HDX direct preview and the Direct workload. They are both the same but the HDX preview will work without defining specific IPs. (Thats the only differecnce)

No rendezvouse makes the connection. But it doesnt define whether or not you use the gateway service.

Thanks all on the heads up on the direct connect info. Saved me future pain

Service continuity should* provide redundancy for the enumerated published apps/desktops by caching the connection leases.

Yeah, that's if your users are using the locally installed Workspace app correct? People who use their web browsers instead of workspace wont get that correct?

Web users can. If they are running Edge/Chrome and use the provided plugin. I wouldn't stake my job on it though - plus we're trying to migrate away from the web only access due to other limitations. service continuity in browser

Mind noting some of those limitations? We are having a back and forth battle internally about this topic as i type 🙂

FIDO2

HDX Direct works differently than Direct Workload ( Network Locations) The big advantages of HDX direct: • No need to configure IPs • Direct connection works both internal & external (STUN) • Initial connection is always via Gateway, if Direct connection is possible it will silently switch over. If not, it will keep using Gateway • Traffic is always encrypted via HDX Direct, both internal as external So, I would definitely recommend looking into HDX Direct

Still having trouble wrapping my head around HDX direct, Direct workload, and rendezvous. Anyone have a tip to crack my thick skull. All are ways to direct traffic to the most direct route to your VDAs correct? Direct workload and HDX are similar by but HDX direct has more features. How is rendezvouse different and can you use rendezvouse with HDX direct or direct workload.

Rendezvous is a technology to bypass the Cloud Connector, so without Rendezvous the traffic flow is: VDA -> Cloud Connector -> Gateway Service -> Client (and the other way around) With Rendezvouz: VDA -> Gateway Service -> Client (and vice versa) So, with Rendezvous you remove load from the Cloud Connector and remove an additional hop. This also allows EDT, as that is only supported in the Rendezvous use case See: https://docs.citrix.com/en-us/citrix-daas/hdx-transport/rendezvous-protocol/rendezvous-v2.html#rendezvous-traffic-flow

What would the benefit be not having the cloud connector be? I would assume for most you are going to have a cloud connector at every site that has VDAs. So its not like you are saving on traffic going somewhere you don't want to go. Reasons I can see not wanting a cloud connector are: 1. You don't have or dont want to pay for a server to install a cloud connector on. 2. You have a vda location that you cant install a cloud connector on and the only way to get to it is through a cloud connector at another site thus causing unneeded traffic over the wire. 3. Rendezvous can take the load of your cloud connectors if you didn't plan correctly or you have to scale and are starting to have to many cloud connectors. To me pretty much if you have a VDA site that has a spare machine to install and less then a couple thousand users CC may be the way to go instaed of Rendezvous

Benefit is not passing all session traffic through it, as Rody mentioned. Don't underestimate that benefit.

Cloud Connector sizing is quite different if you do not use Rendezvous compared to if you would use it (and bypass it for ICA traffic)

Yeah. sizing to me is the only benifit. If you are under what Citrix defines as medium in this article it dont see a great benefit. You are really not gaining much unless you cant find two spare machines to install the CC on. https://docs.citrix.com/en-us/citrix-daas/install-configure/resource-location/cc-scale-and-size.html Now if you are in the medum and above its a no brainer.

You're looking at this wrong. Its not a case of "finding spare machines". its a removing an unneeded failure path in the traffic flow, reducing the overall load (and, hopefully) increasing the reliability of the cloud connectors. Plus you can leverage EDT.

I agree with the failure path. The EDT is another topic lol https://www.reddit.com/r/Citrix/comments/15yk1ub/disabling_udp_gives_better_performance/

ymmv as they say.