This message was deleted.
# citrix-vad
s
This message was deleted.
d
ive had it disconnect on me several times when I was testing it. Didn't spend much time considering migrating to it due to another bug i faced. https://support.citrix.com/article/CTX581101/after-upgrade-to-vda-2308-users-getting-message-youll-need-a-new-app-to-open-this-ctxgeoloc-link
r
Full DaaS with GaaS?
c
What is GaaS?
r
Gateway as a service comparable to Gateway on-prem.
c
No, we have Netscalers and Storefront on-prem.
r
Okay, on the problematic users. What is the total available bandwidth for them?
d
full DaaS and i guess Gaas lol. we just have on prem VDA's and citrix cloud handles the rest after we migrated earlier this year
c
122Mbps, 96Mbps, 256Mbps, 256Mbps. According to control Up the 4 users I just looked at had those values as Bandwidth Limit.
k
Check MTU, I’ve seen issues like that before (there has been topics on this slack instance around that)
r
Give this a read. Most of this should not be an issue anymore with the MTU discovery parts Citrix has now and had for the MTU discovery. But I would read it to see if any fits for you in your environment. You should not have to adjust MTU anymore. But the key word is Should not. https://blog.sachathomet.ch/2020/06/04/citrix-cvad-und-mtu-discovery/#:~:text=The%20solution%20was%20to%20s[…]2Fpmtud-support-on-citrix-gateway.html
Oh and this too. Again, as a reference. I am not saying go out and change your settings based on these links lol. Just material to help you understand if you are seeing any of this. https://support.citrix.com/article/CTX231821/how-to-configure-mss-when-using-edt-on-networks-with-nonstandard-mtu
j
Agree with Ray and Kees that it's more than likely MTU. I found that setting the MSS values in StoreFront config to 1270 gave our users a reliable connection from anywhere through crappy Internet and 4G services. 1270 MSS + 8 UDP header + 20 IP header = 1298 MTU
c
Thank you for all the replies. I will definitely give these a read. However, these users are connecting to the same image, through the same storefront, using UDP, with VDA 1912CU6 without any issues. Most of these users have either Verizon Fios, or Spectrum for Internet and they are in the same city as the infrastructure. I believe they added lossy audio over UDP, so not sure if they made changes to the discovery algorithm. We are going to downgrade the VDA on some test machines, and see if it makes a difference before we go too far into this.
Would packet fragmentation cause latency to be really high? Same user connected to the same desktop shows 51ms latency with NoEDT, 64000m with EDT.
r
So that 64000ms deal I have seen in many environments. It's false as you may already know. But yea, packet frags can cause issues in latency.
Oh, I meant to add. I only seen that when NetScalers are in the mix. about 3 years ago, I worked with CItrix on that. We had to disable some part on the NS side. But for the life of me I can't find it, which is odd because I keep 95% of things. However, if you want to call into them. DM me and I can share my information with you so that you can give them to look the case up.
j
@Craig Weiss yes, I recently updated of of my customer running Server 2022 from VDA 2212 to VDA 2308 and now some users are experiencing constant disconnects. I will turn off UDP temporarily to see if that helps.
I still have random user disconnection even with UDP disabled
c
Interesting. Disabling UDP solved it for all of our users. We downgraded our VDA to 2203cu3, and now everyone seems to be connecting with no issues.
r
Glad you have a fix/workaround. But stinks for the UDP part and the VDA verison
j
I'm rolling back to 2303 tonight
b
@Jonathan Pitre any news on this? Interested what the roll-back has done for you.
j
It seems to have fixed the issue, no more complaints from the customer
I do have others customers on 2308 with Win11 multi sessions without any issues. Maybe this is specific to Windows Server 2022.
👍 1