This message was deleted.
# citrix-vad
s
This message was deleted.
s
same thing happens if you boot from ISO?
d
BDM is iso or am I mistaken, in my abbreviations? 🙂
s
BDM is actually a partition on the VM
d
I've always referred BDM to using ISO. We are only booting from ISO. I'll edit the post
s
and that screenshot is where it just gets stuck and sit there?
d
Yup
s
any FW b/t the VDA and PVS Server? Any relevant event logs on the PVS Server?
d
We moved one of the PVS servers, to the VDA vlan and changed the streaming IP. Created a new boot iso, with the single PVS server, just to eliminate that possibility of missing ports in the FW. Nothing out of the ordinary in the logs.
s
you see the logs on the PVS server to see that it's talking, or switching servers at all?
is it always stuck on that screen connecting to the same PVS server?
d
Its stuck on both of them. I’ll look through the logs tomorrow, and give a shout, if I find anything out of the ordinary. Thanks for the help so far
s
just confirming older versions or older vdisks don't work as well?
like, it's not something image related
if you goto the target device and the last tab and enable more logging, does any of that show up in event viewer on the server?
both pvs servers have been rebooted?
j
I had once a sneaky issue because the PVS subnet was misconfigured so depending of the client IP it couldn't reach the PVS server. Double check your network conf
💯 1
d
Hi guys. Sorry for the suuuuper late answer. So we ended up with, formatting D, from ReFS to NTFS. Copy vhdx's to the drive and now, it seems, everything is working fine and restart times are off the chart. We didn't find any, this-is-whats-wrong and may never find it. How is your experience with NTFS vs. ReFS in a PVS setup?
🙌 2
d
using refs on new PVS servers. no boot issues so far. But cant say boot is blazing fast....
j
ReFS is definately faster for larger files.
m
I keep hearing people say refs is not good if there’s a lot of connections to the server, so for PVS I guess it really depends how many machines you have talking to the server
for FSLogix for example it seems to fall apart over 500 users/connections
j
all version of ReFS or older OS ? we need some benchmarks
m
on server 2019
from what I was hearing
j
was there any major change to ReFS between 2019 and 2022?
m
I’v never tried it with more than 500 so I can’t say if that holds true
j
and was that with a Vmware PVSCSI dedicated controller?
m
don’t know
I don’t have first hand experience of refs not working well
d
500 per PVS server is a lot of users...
m
yeah, I’ve seen it though…
I would not recommend it
d
i bet. would be a good problem to have 🙂
d
There's something like 22 or 23 VDI's per PVS server. So it really shouldn't impact, whether it's NTFS or ReFS. We, the consulting firm I work for, is suspect VMware SRM(Site Recovery Manager) to cause the issue, but can't prove it. Unfortunately.
l
We updated all of our PVS data disks to ReFS about a year ago. Everything is faster. Vdisk merge is seconds which is nice. We targeted about 600 Vms per PVS server (all single session) and never had an issue. Sometimes it could get to 800ish and still no problem.
m
wow, that’s good data to know
l
We also moved from bare metal physical servers with 2x10GB LACP to virtual machines on Nutanix and saw a MUCH improved PVS VM boot times. We had a large number of PVS pods as well - 50K machines across 6 or 7 PVS environments. If you are doing PVS on Nutanix, definitely use the ISO boot - it gives you so much better granularity over PXE.