It looks like the container can see the reader, but the OS is blocking access because of permission rules on the device.
On ZimaOS the FTDI device appears as
crw-rw---- root dialout
which means only root and the dialout group can open it.
Inside your container you are running as PUID 1000 and PGID 1000. That user does not belong to the dialout group, so even with privileged mode enabled the kernel still applies the host permissions. This is why Oscam shows
errno 13 permission denied.
I suggest trying one of these approaches
Run the container as root inside the container.
Oscam does not require root, but this is the simplest way to match the device permissions.
Add the container user to the dialout group.
ZimaOS does not automatically map supplementary groups into containers, so you may need to override the user mapping or switch to a root user. Many LinuxServer images allow
PUID 0 and PGID 0
which makes access immediate.
Confirm the host sees the reader correctly.
If lsusb shows the FTDI chip and ttyUSB0 is created, then the problem is only permission mapping and not hardware.
I think option 1 is the fastest fix on ZimaOS.
Change your compose to run Oscam as root (PUID 0 PGID 0) and keep privileged mode on. That should allow Oscam to open ttyUSB0 without error.
If it still fails, then the issue is not permissions but the driver binding, and we can explore forcing the ftdi_sio driver or checking udev rules.
At this point the FTDI / USB issue is resolved.
What you’re seeing now is a pure network connectivity problem, not permissions.
The key line is
connect failed: Connection timed out
That means Oscam never reaches the remote CCcam server at all.
A few important points I believe apply here:
Bridge mode + CCcam can be problematic
You are running the container in network_mode: bridge.
CCcam is very sensitive to NAT and port translation.
I strongly suggest switching Oscam to network_mode: host
and removing the explicit port mappings. This avoids Docker NAT entirely and mirrors how it worked on TrueNAS.
Confirm the remote side is actually reachable
From the ZimaOS host (not the container), test:
ping 192.168.2.200
telnet 192.168.2.200 10023
If this times out, the issue is upstream (firewall, routing, or the CCcam server itself).
Check CCcam is listening on the server
On the remote Oscam server, confirm port 10023 is bound to the correct interface (0.0.0.0, not localhost only).
No permission issue anymore
Running with PUID 0 PGID 0 and privileged mode rules out local access problems. The timeout confirms this is not an Oscam or USB issue.
In short, I think the fix is:
switch to host networking
verify basic TCP reachability to 192.168.2.200:10023
If it still times out after that, the problem is definitely on the CCcam server side or the network between them, not ZimaOS or Docker.
The error “No route to host” means there is no network path to that IP at all.
That points to one of these, in order of likelihood:
Different subnets with no routing
The dm920 (and likely ZimaOS) is not on the 192.168.2.0/24 network, and there is no router or gateway allowing traffic between subnets.
Wrong IP or interface on the server
The CCcam server may no longer be on 192.168.2.200, or it is bound to a different interface or VLAN.
Firewall blocking ICMP and TCP
A firewall on the server or router is blocking both ping and port 10023. Blocking ping alone is common, but blocking both usually indicates routing or firewall rules.
Network isolation / VLAN
If the server is on a VLAN or isolated LAN, devices outside that segment will not reach it without explicit routing.
Important takeaway:
If ping fails, CCcam will never work, regardless of Oscam config, Docker mode, or ZimaOS.
I suggest:
Verify the actual IP of the CCcam server
Confirm both devices are on the same subnet, or that routing exists between them
Temporarily disable firewall rules on the server to test connectivity
Once basic ping works, CCcam will connect immediately.
I’m back after a while, after struggling with everything. The latest Docker update helped. So the culprit wasn’t ZimaOS or Dreambox, but the OSCAM binary itself or Docker (linuxserver/oscam - Docker Image). Now everything works as it should, and I’m happy.
That’s actually really valuable info for others:
it confirms the issue wasn’t ZimaOS or the Dreambox network at all, but something inside the linuxserver/oscam container (OSCAM binary and or Docker layer), and the newer Docker update resolved it.
Really glad it’s now working properly and stable
If you notice anything repeatable (specific version that broke vs version that fixed), sharing that would help the community a lot.