NICE DCV on EC2 RHEL 9 fails

When following the install instructions for NICE dcv for RHEL 8 on my RHEL 9 instance, when trying to start the dcv service with:

systemctl start dcvserver

fails and a "systemctl status dcvserver" gives: $ sudo systemctl status dcvserver [sudo] password for builder: × dcvserver.service - NICE DCV server daemon Loaded: loaded (/usr/lib/systemd/system/dcvserver.service; enabled; vendor preset: disabled) Active: failed (Result: exit-code) since Fri 2023-01-06 23:05:04 UTC; 19min ago Process: 1092 ExecStartPre=/sbin/modprobe -a eveusb v4l2loopback (code=exited, status=1/FAILURE) Process: 1093 ExecStart=/usr/bin/dcvserver -d --service (code=exited, status=133) Main PID: 1093 (code=exited, status=133) CPU: 33ms

*** systemd[1]: dcvserver.service: Scheduled restart job, restart counter is at 5. *** systemd[1]: Stopped NICE DCV server daemon. *** systemd[1]: dcvserver.service: Start request repeated too quickly. *** systemd[1]: dcvserver.service: Failed with result 'exit-code'. *** systemd[1]: Failed to start NICE DCV server daemon.

also: $ dcv list-sessions

(process:2041): libsoup-ERROR **: 23:27:52.145: libsoup2 symbols detected. Using libsoup2 and libsoup3 in the same process is not supported. /usr/bin/dcv: line 26: 2041 Trace/breakpoint trap (core dumped) "${dcv_bin}" "$@"

Any ideas? I would really prefer to use dcv over xrdp to connect to these instances.

AMI is on us-gov-east-1 but is probably the same as: "RHEL-9.0.0_HVM-20220513-x86_64-0-Hourly2-GP2" in other regions I added the server GUI packages and the usual Nice dcv stuff. No "GL" stuff involved.