I am investigating an issue with different versions of CWA for Windows 10 (1912 - 2212) where in double-hop scenario with hop one VDI running on a VM with Nvidia T4 GRID, starting Citrix delivered applications from this VDI causing all local applications to slow down. More applications I start, slower local applications are responding. I am talking about switching between applications, navigating inside applications, or even explorer. Starting single application has no impact, starting 2-3, becomes noticeably slow, while 3+ downright unusable. All applications are hosted of non-GPU 1912 XenApp site (including multi-session VDAs). I can also reproduce same issue if I RDP into the same 1st hop VDI. I can even uninstall all VDA components from the same VDI. Issue always the same until I fully disable Nvidia GPU at VM level. Once GPU is disabled, local application lag goes away. I already tried disabling CWA hardware acceleration, as well as list of other settings I found on-line. As of now, I can not identify root cause for this. What is even more strange, I do not see where this lag is coming from. VDI CPU, Memory, IOPS, GPU utilization, all look pretty minimal and nothing is pegged. ICA RTT and latency are set to update every 1s and are in single numbers. VDI is 8CPU and 64Gb of memory, running on Vmware 7 and we have other non-GPU VDIs in the environment, as well as physical remote PCs with Nvidia cards that do not exhibit this behaviour. Have support case open with Citrix and Nvidia and Nvidia is pointing finger at Citrix and Citrix is still investigating collected traces. Figured, let me bounce this against the community, as I am sure this is not first ever implementation with Nvidia GRID backed VDIs.
Question
Dmitry Parmit1709162581
Hello all.
I am investigating an issue with different versions of CWA for Windows 10 (1912 - 2212) where in double-hop scenario with hop one VDI running on a VM with Nvidia T4 GRID, starting Citrix delivered applications from this VDI causing all local applications to slow down. More applications I start, slower local applications are responding. I am talking about switching between applications, navigating inside applications, or even explorer. Starting single application has no impact, starting 2-3, becomes noticeably slow, while 3+ downright unusable. All applications are hosted of non-GPU 1912 XenApp site (including multi-session VDAs). I can also reproduce same issue if I RDP into the same 1st hop VDI. I can even uninstall all VDA components from the same VDI. Issue always the same until I fully disable Nvidia GPU at VM level. Once GPU is disabled, local application lag goes away. I already tried disabling CWA hardware acceleration, as well as list of other settings I found on-line. As of now, I can not identify root cause for this. What is even more strange, I do not see where this lag is coming from. VDI CPU, Memory, IOPS, GPU utilization, all look pretty minimal and nothing is pegged. ICA RTT and latency are set to update every 1s and are in single numbers. VDI is 8CPU and 64Gb of memory, running on Vmware 7 and we have other non-GPU VDIs in the environment, as well as physical remote PCs with Nvidia cards that do not exhibit this behaviour. Have support case open with Citrix and Nvidia and Nvidia is pointing finger at Citrix and Citrix is still investigating collected traces. Figured, let me bounce this against the community, as I am sure this is not first ever implementation with Nvidia GRID backed VDIs.
Thank you in advance.
Link to comment
2 answers to this question
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now