Hello Nick,
Thanks for your reply! We have kept an eye on resource usage across the guac
VM depending on which use case we tried. In order to try and minimize that
possibility we gave 8vCPU and 8GB ram to the VM and tried with one concurrent
user. Using top/iotop to keep tabs on things, we have not seen any obvious
bottlenecks with this approach. Thanks again!
Best regards,
Ryan
From: Nick Couchman
Sent: Thursday, April 16, 2020 2:09 PM
To: user@guacamole.apache.org
Subject: EXT: Re: Graphical Performance Question
On Thu, Apr 16, 2020 at 3:04 PM Nanney, Ryan (GE Healthcare, consultant)
mailto:ryan.nan...@ge.com>> wrote:
Hello all,
I am a technical product owner with GE Healthcare and we are using Guacamole to
provide medical professionals an ability to remotely access applications while
reviewing diagnostic imaging such as MRI scans. Overall our experience with
Guacamole has been quite good, though we have an area of concern and are
curious what types of options exist. The symptom we are facing is during
certain instances of cinematic playback there are blocky (64x64 squares) areas
which become noticeable at certain times. I was able to watch the websocket
traffic to identify this scenario.
I have attached a small MKV video to illustrate the playback issue folks are
having issue with, for your review.
We are operating off of a Gigabit LAN environment and Guacamole is exposed to
the LAN, the rest of the traffic happens inside of a virtualized environment.
Our endpoints are all using RDP, either from Windows Server 2012R2 or Windows
Server 2019.
It has been observed that direct RDP provides a smoother playback experience.
Also, it seems that server 2019 is better than 2012R2. We have tried both the
docker container and building from source, with version 1.0.0 and 1.1.0 which
produce similar experiences. An nginx server is also in play, though we have
tested with that removed from the stack.
I certainly welcome any feedback or ideas anyone may have. We would like to
understand the cause and what the potential fix(es) may be. It seems to be
something within guacd or tomcat?
Ryan,
Have you examined resource utilization on both the system running guacd and the
system running Tomcat (may be the same system?) to make sure there isn't any
resource shortage, there?
-Nick
