This message was deleted.

Blame it on the ISP? 😆

break it down.. draw out the whole communication flow and look into each small portion (if possible).. and look at connectivity, speed etc is doing there.. treat it as micro services to find where the issue is. but draw it on a whiteboard, make blocks of pieces you can look into, test and work from the back to the client… and as Steve says, it probably is the ISP but in case it is not you want to see where it is..

Yes they are on the internet. I was hoping for a service or something. The users are mainly in India

This is the hardest part of our job. But like Steve said, ISP. Are the all from the same region? What is it normally?

Yes mainly India. I was thinking either Citrix cloud, or the ITM service. They just say Citrix is slow

I really dislike the two words. "Citrix Slow"

They do have a POP in India but most US based enterprises won't allow users to use it. They have to control in point of egress/ingress and put policies around it

2000ms is really bad. Using EDT?

I amusing edt. 2000 is actually the low end, checkout the peak!

I dont know ur environment. But edt would be set in citrix studio and on the netscaler gateway via dtls. On workspace you can check the connections and it will say dtls. In controlup insights it will show u what protocol is being used. Such as upd aka elighted Data transport

A good RTT from India would be around 250ms in my experience. If you’re going through a NetScaler then keep in mind that the default TCP profiles assume much better connectivity than that, so you would likely want to look into some tweaking. That said, I wouldn’t start fiddling with that until you involve the ISP/local network team as 2000ms is unacceptable no matter where you’re connecting from. HDX insight can at least show you more in depth data like zero window events and retransmits that can help you point the finger at the client network rather than yours.

I have to say our community is awesome, everyone sooo helpful :->

2000ms or worse is not normal.. it is India not the mars.. I can imagine Citrix is slow over 2000ms.. that they can even work is amazing. my question would also be the POP that is used… but the breakdown would show that. let us know what pop is used.

2000ms doesn’t even qualify as Internet in my opinion. I have users in India using the Citrix Cloud/Azure PoP released earlier this year accessing VDAs in other regions and it’s very fast for them. There’s something going on at the last mile for your users Cory (aka their office network with some kind of outbound packet inspection that hasn’t whitelisted remoting traffic, something up with their ISP, etc)

@Jason Samuel yea they are all going to our Data center on prem. I was going to test Citrix cloud, but all resources are still here.

@Cory Zaner It’s worth testing. I had India users coming straight over the Internet to onprem NetScalers in Texas and it was the typical up to 300 ms at times. When getting them to try Citrix Workspace they started hitting the Singapore POP (the Chennai POP was not available yet). Their experience was much faster and latency was a lot lower on ControlUp and Director. Lots of positive comments from the users. It’s usually that last mile that causes issues and if you optimize that, it makes a big difference.

@Jason Samuel is they enter the pop does there traffic go over the azure back haul or right from the client to the cloud connectors on prem?

@Cory Zaner if you have expressroute between your datacenter and azure it will go over that backbone (which is really nice). If you don’t then it will go from that POP over the regular Internet and down to your cloud connectors on prem.