JosephF3. Our server is in the same boat. But only the latest build we did, running on Server 2016. I have sourced a new server and building a sole Server 2016 RDS box, deciding to get rid of 2X/Parallels as they have dropped the ball in the past 12 months. ie. I have a post on this forum "Where's the update?" as there is some 'fixes' in the more recent releases of RAS that 'may' address the issue. But I cannot get a response from anyone, so good luck to you too! I shall be keeping an eye on your post also in a vain hope someone helps you/us out! We were averaging a reboot a day ... alas today, we've done 3 already (4th pending, but I'm desperately snooping around hoping I've missed something the past 1000 times I've looked) - so getting over sending broadcasts "just for a change, the servers stopped accepting connections and is being rebooted"
Further to your comments above, I have also noticed the lack of logs as to whats happened - on either the server or the RAS. I believe this is because the program/system doesn't realize there's an issue, as RAS allows people to connect and passes them over to the RDS but doesn't 'realize' (lack of better words) that their connection was never completed. Same boat with the Windows logs, the user technically connects - but doesn't? Then there's the quirk of everyone who's already connected - stays connected. But anyone trying to get on, can not.
Our users are greeted with a variety of error messages, from:
RDP Host busy
Portable Device Service has not responded in a timely fashion
Process is already running, but has not responded. The old process must be closed to open a new one.
On the Server - if you're able to log in.
Task Manager shows, instead of a username appearing in the "USER" column, a handful of username "(4)" (referring to how many processes are running) with the status of disconnected.
A lot of research I've done points to "one problematic client, find them and you're set" ... but when you have 270 users connecting at any which time, it's hard/impossible to narrow down - especially when there's no helpful log entries and user's don't report it's fallen over for XX minutes!
Not a very helpful post, sorry ... just someone on your side!
Oh, one last thing I remembered as I hit 'post ... we have an application published from another host, a windows 10 client, so no RDS involved - this app does NOT work when the before mentioned issue rears it's ugly head - which to me, suggests this is a problem with RAS (not the server) - thus I'm chasing the updates to RAS so badly. Also - try publishing "Internet Explorer" from the problematic RDS, watch it fire without a problem while the desktop problem exists - *baffled!*
edit- "30minutes later", glad I put this post in ... obviously someone saw it and kicked them in to gear - i now have the updates offered to my RAS, applying now! shall keep you in the loop ;D
Last edited: Mar 29, 2017