08-05-2025, 06:43 PM
I made a small change to the Keepalive routine that might help with the Eth02. I am just busy testing for a few days then will release a next release.
Ingo
Ingo
|
Home Assistant addon 1.5.0 crashes soon after startup
|
|
08-05-2025, 06:43 PM
I made a small change to the Keepalive routine that might help with the Eth02. I am just busy testing for a few days then will release a next release.
Ingo
08-29-2025, 08:06 AM
Hi there,
Back from holiday and noticed the new release 1.5.1, which I installed to test. It doesn't crash anymore, but something causes the connection to Comfort to be dropped and a reconnection: Is this something I am likely to be able to fix by configuration? Thanks, Jon.
08-29-2025, 09:26 AM
Unknown. The keepalive failed to be sent due to the connection being closed. Something is closing or breaking the TCP connection to Comfort which was the initial problem but now it doesn't crash, it should retry. Are the failures happening with a predictable interval, say every 2 mins, or are they random?
I tested an Eth02 I had and I could not simulate he issue you have. See if you can find another UCM/Eth to test with to see if it resolves your issue.
08-29-2025, 12:12 PM
The sequence is:
1) Start add-on 2) Add-on connects to Comfort 3) Add-on runs through connection sequence, finishing with the r?00F0.... command 4) Add-on reports "Error sending command 'cc00', closing socket" (as the very next line after the r?00F0.... command, every time) 5) Goto 2) slychiu mentioned in this thread that the r?00F0.... command may cause a reset on a non-ARM UCM or CM9001. I have an old UCM, which I am guessing is non-ARM. Maybe we could try a version where we could configure that command not to be sent, to test if that's the problem? Unfortunately I don't have access to any other UCM/ETH modules. Thanks, Jon.
08-29-2025, 12:32 PM
The r?00F00F is a perfectly legal command as I test on both ARM and Non-ARM systems. The one he mentioned is a slight variation of this which is not implemented in the AddOn and should not cause an issue.
The only thing I can think of is that the Eth02 is running out of internal buffer space and then gives up. Make sure you have the latest Eth02 firmware. Even though it's not officially supported and End of Life, I did not get the same results as you. The latest firmware Eth02 is 3.13 at the link below. https://cytech.biz/download_files.php?item_id=341
I think I'm on that version already, but I'll check when I get home from work.
Thanks, Jon. (08-29-2025, 12:32 PM)Ingo Wrote: The r?00F00F is a perfectly legal command as I test on both ARM and Non-ARM systems. The one he mentioned is a slight variation of this which is not implemented in the AddOn and should not cause an issue. This is the r?00F00F command copied from Slychiu's post warning that it may cause a reset: r?00F00F000000000000000000000000000000000000000000000000000000000000 This is the r?00F00F command copied from the addon log: r?00F00F000000000000000000000000000000000000000000000000000000000000 As far as I can tell these are identical?
08-31-2025, 09:17 AM
Confirmed my ETH02 is running firmware 3.13.
Jon.
09-01-2025, 12:57 AM
perhaps the easiest solution is to get an ETH03 to replace the old ETH02 plug-in module
09-01-2025, 04:38 AM
Easiest for others maybe, but it means spending something like £100 on a 20-year-old system with no guarantee I won't have to change something else just to get ongoing support in Home Assistant.
The fact that Ingo's testing with his ETH02 doesn't behave the same as mine to me points at it being a different issue than that - I have a very simple Comfort system so it's not like the ETH02 should be being overwhelmed with traffic when there's only 12 inputs and 1 keypad in use! I wrote my own plugin when I used HomeSeer, perhaps that's a better option for me, or maybe it's time I looked at a more modern alarm system? Thanks, Jon. |
| Users browsing this thread: |
| 2 Guest(s) |