Closed stan23 closed 10 months ago
Bitte trace log.
I don't think there's anything to do. We have a valid refresh token. Refresh fails with unspecified server timeout. Afterwards refresh totally fails with previously valid refresh token. My guess: PSA had a infrastructure hick-up and invalidated all tokens. Since we're not seeing this regularly wontfix.
Hmm, I'm seeing it every few weeks and one morning I was stuck with an empty battery.
Any idea how I can catch it externally? messaging
does not trigger on this kind of error.
I guess I need to check the status reported by evcc API when it appears the next time and see if anything indicates this constellation.
Could you try finding a log that does not have a timeout error before? Or does it always happen together with a timeout error?
I have only the last 3 occurrences due to log rotate, and it is always after that same timeout error.
But I'll keep an eye on it in the future once it happens again.
Does it ever refresh successfully? Think it did in the original log. So not sure what is going on at PSA...
It did refresh a number of times until it broke after the timeout at 18:48.
164:Dec 31 16:09:25 evccCT evcc[581451]: [peugeot] TRACE 2023/12/31 16:09:25 grant_type=refresh_token&refresh_token=e33aab16-9fa8-4ce3-a541-66d2347c94f2
513:Jan 2 08:03:24 evccCT evcc[581451]: [peugeot] TRACE 2024/01/02 08:03:24 grant_type=refresh_token&refresh_token=75b2ed6d-c7bb-4cf2-abaa-e672408ad944
569:Jan 2 12:25:34 evccCT evcc[581451]: [peugeot] TRACE 2024/01/02 12:25:34 grant_type=refresh_token&refresh_token=938db208-5f60-4e5e-9a0b-d759c92ec5d5
589:Jan 2 13:26:14 evccCT evcc[581451]: [peugeot] TRACE 2024/01/02 13:26:14 grant_type=refresh_token&refresh_token=a768f870-794d-48b9-b72e-28b222e97c7a
611:Jan 2 14:26:54 evccCT evcc[581451]: [peugeot] TRACE 2024/01/02 14:26:54 grant_type=refresh_token&refresh_token=f8f4509d-c08a-44e8-bad1-f7ab3cd686fd
630:Jan 2 15:27:34 evccCT evcc[581451]: [peugeot] TRACE 2024/01/02 15:27:34 grant_type=refresh_token&refresh_token=ede69ae1-d561-4bc2-b812-196458426492
707:Jan 2 21:08:54 evccCT evcc[581451]: [peugeot] TRACE 2024/01/02 21:08:54 grant_type=refresh_token&refresh_token=114ddf61-dfa2-48fb-9388-44cdcbc22ae1
823:Jan 3 10:08:25 evccCT evcc[581451]: [peugeot] TRACE 2024/01/03 10:08:25 grant_type=refresh_token&refresh_token=614db5e4-f136-4bd3-8dba-89cb69c4e4a7
1171:Jan 4 07:36:26 evccCT evcc[593399]: [peugeot] TRACE 2024/01/04 07:36:26 grant_type=refresh_token&refresh_token=f913cd59-3f27-4b6e-b8b2-7a29cf6a57d9
1190:Jan 4 08:44:56 evccCT evcc[593399]: [peugeot] TRACE 2024/01/04 08:44:56 grant_type=refresh_token&refresh_token=f52be6f8-df9f-4b79-ac63-8704e90f76fe
1214:Jan 4 09:45:26 evccCT evcc[593399]: [peugeot] TRACE 2024/01/04 09:45:26 grant_type=refresh_token&refresh_token=fc524313-c258-40fc-a2e5-717145083d6f
1294:Jan 4 17:45:36 evccCT evcc[593399]: [peugeot] TRACE 2024/01/04 17:45:36 grant_type=refresh_token&refresh_token=5f825017-74d1-4fce-b414-6b8fccb15a0f
1318:Jan 4 18:48:25 evccCT evcc[593399]: [peugeot] TRACE 2024/01/04 18:48:25 grant_type=refresh_token&refresh_token=f1a7f686-f46a-4e4a-832c-a18d805b9f92
1324:Jan 4 18:48:35 evccCT evcc[593399]: [peugeot] TRACE 2024/01/04 18:48:35 grant_type=refresh_token&refresh_token=f1a7f686-f46a-4e4a-832c-a18d805b9f92
1332:Jan 4 18:48:56 evccCT evcc[593399]: [peugeot] TRACE 2024/01/04 18:48:56 grant_type=refresh_token&refresh_token=f1a7f686-f46a-4e4a-832c-a18d805b9f92
1342:Jan 4 18:49:25 evccCT evcc[593399]: [peugeot] TRACE 2024/01/04 18:49:25 grant_type=refresh_token&refresh_token=f1a7f686-f46a-4e4a-832c-a18d805b9f92
This seems to indicate that refresh is working until PSA has a hickup. If this is a new problem let's monitor it for some time. If you're running on docker you could for example restart the container each night to trigger new login. If it persist we'll need to think about logging in again when refresh fails.
Good idea but this is a Proxmox LXC. I will wait until it happens again and then look at the evcc API to see if anything indicates the problem. If so, I'll send a push message to myself to restart evcc.
If I figure out anything I'll update this ticket. Thanks for your time looking into it π
Describe the bug
After some days or weeks the connection to PSA API breaks. As a result, the SoC is unknown and planned charging does not work anymore. Restart of cvcc solves the problem.
Is there a way to catch this and force a kind of re-authentication like a cvcc restart does?
Can I add even more details to the log beyond "Peugeot: trace"?
Steps to reproduce
No special steps, just wait and after few days it happens.
All logs are in attached file. logs.txt
Configuration details
Log details
What type of operating system are you running?
Linux
Version
0.123.5