Dear colleagues and support.
That is not something that was solved here, I found a couple of similar problems with 2008r2, but it cannot be applied to latest OS. And problem described in other topics are a little different
Our case:
We are using large infrastructure based on many Servers 2012r2 all in one RDS servers (Broker, Web, SH is installed at every server). Many servers published behind some load balancer.
As we have a lot of servers in farm - we have a redirected profiles configured (over GPO) and placed at DFS.
We started testing 2016 servers, and our test group start reporting problems with "Remote Desktop Services is currently busy".
Usually problem solved itself in 1-2 min, user have to try several times before he is able to connect. There was no hanged user sessions or something like this or some useful even log as i can see.
There was no solution for 2016 problem over the internet, so I was hoping to find it with 2019 server.
But we are able to reproduce problem even on "clean", powerful (8 cores, SSD, 32 GB RAM), freshly installed server with single role (RDS) on it for the minimal number of users (1-2 users online)
So that is not something like server load or not enough resources.
I assume it is related somehow to redirected profiles or some service really "busy" with what? But why it is working stable for 2012R2 and has such problem with 2016/2019? May be we can adjust some parameters, timeouts or something like this?
Do we have any workaround for this except disabling redirected profiles or living forever based on 2012 r2 (that is not an option when you have a farm for 1000 users online and many member servers behind balancer)
Many thanks for your ideas!
Regards,
Sergii V