SSL errors all of a sudden, not sure of its SSL Lib, or Certificate #25
Comments
|
How are you running your HA ? Is it a docker container ? supervised ? core pip install ? |
|
Hi Im running in |
|
Just updated to Latest Core HA with same result. |
|
The AC uses older certificates, try editing |
|
I tried your suggestion in on the host machine - however that line didn't exist, and when i added it and restarted it made no difference. I looked for an etc directory in the python virt environment and didn't see one. |
|
OK I think I have fixed it. There are 2 files that need to change. I changed this line "sslContext = ssl.SSLContext(ssl.PROTOCOL_TLSv1)" I also tried using The only one to work was "TLS" ( or you can put "SSLv23" as its the same thing according to here: The second file I needed to change was /etc/ssl/openssl.cnf |
|
Great, thanks @tinstep I will try to get this mentioned in the readme as a possible solution. |
|
I experience the same problem but with a different (newer?) model. @@ -102,7 +102,10 @@
with requests.sessions.Session() as session:
self.logger.info(self._params)
try:
+ original_ciphers = requests.packages.urllib3.util.ssl_.DEFAULT_CIPHERS
+ requests.packages.urllib3.util.ssl_.DEFAULT_CIPHERS = 'ALL:@SECLEVEL=0'
resp = session.request(**self._params)
+ requests.packages.urllib3.util.ssl_.DEFAULT_CIPHERS = original_ciphers
self.logger.info(
"Command executed with code: {}, text: {}".format(
resp.status_code, resp.textThis is a very quick and dirty hack, but it seems to work. I do notice the following in the logs, but I think that's the request itself, and not modifying the ciphers. My python (especially async) is a bit rusty, but could someone confirm this? And if so, I think the better solution is to "mount" a specific adapter for this as described here I will try to see if I can implement this, but a confirmation on my line of thought would be great. |
|
this work, for me: |
|
Should be fixed in https://github.com/atxbyea/samsungrac/releases/tag/3.5.3 Thanks @gpancot |
Hi
I run HA in ubuntu 20.04, climate_ip stopped working somewhere in the last month. I've made no changes and it was controlling two Samsung AC units perfectly. When it stopped working added debug clauses for climate_ip to HA and didn't really see any additional details, however after updating HA today to 20210303 I can now see all these errors related to climate_ip coming through the logs.
This is using 3.5.1 of Climate Ip installed via HACS - this has not been updated for a while and was working fine on this version.
As per title, not sure if this is a new SSL library that now enforces something stronger, or the cert is at fault.
I pasted all the certs into a decoder and they all expire in 2059 - but my knowledge is pretty limited with this so hopefully you can help.
(Note I replaced part of token with xxxxx)
The text was updated successfully, but these errors were encountered: