Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

after updating visonicinterface to ESPHome 2023.4 - panel does not connect #77

Closed
EyalRotem opened this issue Apr 20, 2023 · 78 comments
Closed

Comments

@EyalRotem
Copy link

Hi,

ISSUE: panel not connecting after updating the visonicinterface to version 2023.4

THINGS I TRIED: this happened before, but a reset to HA usually fixed it (or a reset to the D1mini and reloading the integration). unfortunately - this time, nothing helped.

POSSIBLE CAUSE: a breaking change in the ESPHome - regarding UART devices. this is the changelog, and this is the issue that is linked to the breaking change.

FIX FOR NOW: restored an older version of HA with the previous ESPHome version - 2023.3.2. after a reboot, everything is working as it should.

LOGS: sorry, i restored and have nothing. but when i looked (after adding the logger debug stuff in configuration.yaml) there wasn't anything that "screams" an issue, just failure to connect- if needed - i can update again to replicate the issue, but i rather not, if possible haha.
what i can remember is this is the last line that was the same - and after it, other lines that to not match the working log:
[custom_components.visonic.pyvisonic] Setting TCP socket Options

Let me know if I can help.

@konradwalsh
Copy link

konradwalsh commented Apr 21, 2023

I can confirm the same behaviour

This error originated from a custom integration.

Logger: custom_components.visonic.client
Source: custom_components/visonic/client.py:1185
Integration: Visonic Intruder Alarm (documentation, issues)
First occurred: 11:04:32 (1 occurrences)
Last logged: 11:04:32

Failed to connect into Visonic Alarm. Check Settings.

@davesmeghead
Copy link
Owner

This line just tells me that it didn't connect: Source: custom_components/visonic/client.py:1185

If you have updated ESPHome to the new version, have you also updated your visonic.yaml file from the wiki. I try to keep the wiki visonic.yaml file up to date with the versions of ESPHome. You may need to also add to your secrets.yaml to set the necessary parameters.

If that doesn't help, and I keep saying this in the forum and troubleshooting section of the wiki, please capture and show me the Home Assistant log file.

@davesmeghead
Copy link
Owner

davesmeghead commented Apr 21, 2023

I should have added, I have updated to the latest ESPHome and my development connection my PowerMaster 30 works.
Also the uart0 ESPHome breaking change isn't the problem, I don't use that, I use "uart"

@EyalRotem
Copy link
Author

OK, so i have updated to the latest visonic.yaml when it was last changed.
only thing i changed in it are commenting the "service" and "switch" as i have no need in the triggers and siren. I did try a re-flash previously with no avail.

I will try to find the time to update again - and produce the logs with as much info as possible.

to add info on my setup - it's a powermax 10.

@davesmeghead
Copy link
Owner

When you create a log file, remember to set debug for this integration as per here

@EyalRotem
Copy link
Author

hi, just finished collecting logs:

here's a good one with esphome 2023.3.2.

here's a bad one after updating the d1 mini to esphome 2023.4.

@davesmeghead
Copy link
Owner

davesmeghead commented Apr 24, 2023

I can see a couple of problems in both the "good" and "bad" log files.

In both "good" and "bad" log files I find this

2023-04-23 19:55:43.388 DEBUG (MainThread) [custom_components.visonic.pyvisonic] [Controller] ****************************** Response Timer Expired ********************************
2023-04-23 19:55:43.389 DEBUG (MainThread) [custom_components.visonic.pyvisonic] [_clearList] Setting queue empty
2023-04-23 19:55:43.390 DEBUG (MainThread) [custom_components.visonic.pyvisonic] [sendPdu] Resetting expected response counter, it got to 0   Response list is now 2
2023-04-23 19:55:43.391 DEBUG (MainThread) [custom_components.visonic.pyvisonic] [sendPdu] Sending Command (Getting Status)    raw data 0d a2 00 00 00 00 00 00 00 00 00 00 43 1a 0a    waiting for message response ['0XA5', '0X2']
2023-04-23 19:55:43.578 DEBUG (MainThread) [custom_components.visonic.pyvisonic] [data receiver] msgType 0X2 got it so removed from list, list is now ['0XA5']
2023-04-23 19:55:43.578 DEBUG (MainThread) [custom_components.visonic.pyvisonic] [_processReceivedPacket] Had a response timeout PROBLEM but received a data packet and entering Standard Plus Mode

Many months ago I discovered a message acknowledge to the panel when in Powerlink state, it seems like this is the cause of your problems so let's go back to the old version.

If you are able to edit the pyvisonic.py file can you try this for me please

On line 1845, for the text MSG_ALIVE2 delete the "2" so it looks like this

                        log.debug("[Controller]   Sending new I'm alive message")
                        self._sendCommand("MSG_ALIVE")
                    else:

On line 3938, do the same, delete the "2" so it looks like this

            self._sendCommand("MSG_ALIVE")       # EXPERIMENTAL 29/8/2022.  The Powerlink module sends this when it gets an i'm alive from the panel.

Can you please give this a try for me.

I'm not sure why the ESPHome update has done this but it looks like it has.

@davesmeghead
Copy link
Owner

I'll just add this as well though, in both bad and good log files they both join to and interact with the panel so you should not have seen the Failed to connect into Visonic Alarm. Check Settings. message.

Can you please confirm whether the "bad" log file failed to bring up the alarm control and sensors in the HA frontend?

@konradwalsh
Copy link

konradwalsh commented Apr 24, 2023

I can confirm that making those changes worked AFTER a full reboot and not just a restart HA

@EyalRotem
Copy link
Author

I'll just add this as well though, in both bad and good log files they both join to and interact with the panel so you should not have seen the Failed to connect into Visonic Alarm. Check Settings. message.

Can you please confirm whether the "bad" log file failed to bring up the alarm control and sensors in the HA frontend?

I can confirm that during the “bad” log, no matter what combination of resetting I tried, it just never connected to the panel, no standard mode, no powerlink, nothing worked.

Than, with just installing the older esphome on the d1mini- it started to work as previously.

later today I’ll try the changes you mentioned.

@EyalRotem
Copy link
Author

I can confirm that making those changes worked AFTER a full reboot and not just a restart HA

You changed the py files and it worked?
It may be helpful if you can produce logs as well! Thanks for pitching in 🤙🏼

@konradwalsh
Copy link

konradwalsh commented Apr 24, 2023

Here is my log...
home-assistant_2023-04-24T08-49-43.095Z.log

I can confirm that making those changes worked AFTER a full reboot and not just a restart HA

You changed the py files and it worked? It may be helpful if you can produce logs as well! Thanks for pitching in 🤙🏼

Yes ... it connects and sees the sensors now and I can activate the alarm..

I do see strange errors in my log but everything works

@EyalRotem
Copy link
Author

Ok, got a chance to change the MSG_ALIVE2 as suggested.

First of all- it seemed that after my trials yesterday the d1 mini was actually on 2023.4 and working… not sure what changed really (maybe tried a full reboot that fixed It?).

this is on the esphome 2023.4:
https://pastebin.com/r7v7P77U

this is after updating to 2023.4.1:
https://pastebin.com/Bm1gn5QM

both seems similar in my eyes.

and now even with the MSG_ALIVE2, it works as it previously did.
https://pastebin.com/XEv556Tt

Basically, everything works as expected on newest esphome, without any changes.
Not sure what happens 😅.

but i do want to pursue the issues you found - should I open a different issue?

sorry for the mess- I hope somehow we understand with the logs what may have happened.

@konradwalsh
Copy link

konradwalsh commented Apr 24, 2023

hmm.. based on your info.. I went back and edited the .py file and restored the MSG_ALIVE2

and everything still works.

The key difference is ESPHOME was updated to a newer version since all of this started.

@EyalRotem
Copy link
Author

hmm.. based on yourinfo.. I went back and edited the .py file and restored the MSG_ALIVE2

and now everything works.

The key difference is ESPHOME was updated to a newer version since all of this started.

Yes, that did happen, but I haven’t updated it until the last trial, so I figure it is irrelevant.

that’s freakin weird😅

@davesmeghead
Copy link
Owner

@konradwalsh
Looking through your log file, it disconnects and reconnects back to the panel many times.
In lines 886 to 887 there is no response from the panel when the "Get Bypass Status" message (line 886) is sent to the panel (although there is no guarantee that the message got to the panel).
In fact after line 886 to 951 the integration sends many messages to the panel with no response.
On line 952 your operating system (on whatever hardware you have) declares that the Ethernet connection has been terminated. There is no evidence that your automation did this. This means I think that you have a problem with your network or the ESPHome device. Having said that, because of the sequencing of the messages to the log file by HA, it can be difficult to determine.
On line 980 the integration attempts to reconnect to the panel and succeeds.
That is until line 1021 where the reconnect service is called (do you have an automation that does this), just as it started to download the EPROM.
At line 1049 where the reconnect service is called again.
Can you please disable all the HA restart automation that you have and let it do its own thing, then recapture a log file please.

@EyalRotem
All 3 log files look OK. Each gets to Standard Plus mode but the log files aren't long enough to see if they get to PowerLink mode.
You're using panel number 1 instead of the default 0, that's all I noticed. If your front end GUI devices (widgets / icons etc) expect panel 0 sensor names then it may look like the panel hasn't connected when it really has.
Also, I did notice this in your last log file (I think)

2023-04-24 14:19:12.147 ERROR (MainThread) [homeassistant.util.logging] Exception in setup_update_entity when dispatching 'esphome_1034ce64b157d501b983a18e43c89fde_on_device_update': ()
Traceback (most recent call last):
File "/usr/src/homeassistant/homeassistant/components/esphome/update.py", line 56, in setup_update_entity
async_add_entities([ESPHomeUpdateEntity(entry_data, dashboard)])
File "/usr/src/homeassistant/homeassistant/components/esphome/update.py", line 90, in __init__
if coordinator.supports_update and not self._device_info.has_deep_sleep:
File "/usr/src/homeassistant/homeassistant/components/esphome/dashboard.py", line 91, in supports_update
raise RuntimeError("Data needs to be loaded first")
RuntimeError: Data needs to be loaded first

This seems to be an ESPHome error but I don't know what it means.

@EyalRotem
Copy link
Author

@EyalRotem All 3 log files look OK. Each gets to Standard Plus mode but the log files aren't long enough to see if they get to PowerLink mode. You're using panel number 1 instead of the default 0, that's all I noticed. If your front end GUI devices (widgets / icons etc) expect panel 0 sensor names then it may look like the panel hasn't connected when it really has. Also, I did notice this in your last log file (I think)

2023-04-24 14:19:12.147 ERROR (MainThread) [homeassistant.util.logging] Exception in setup_update_entity when dispatching 'esphome_1034ce64b157d501b983a18e43c89fde_on_device_update': ()
Traceback (most recent call last):
File "/usr/src/homeassistant/homeassistant/components/esphome/update.py", line 56, in setup_update_entity
async_add_entities([ESPHomeUpdateEntity(entry_data, dashboard)])
File "/usr/src/homeassistant/homeassistant/components/esphome/update.py", line 90, in __init__
if coordinator.supports_update and not self._device_info.has_deep_sleep:
File "/usr/src/homeassistant/homeassistant/components/esphome/dashboard.py", line 91, in supports_update
raise RuntimeError("Data needs to be loaded first")
RuntimeError: Data needs to be loaded first

This seems to be an ESPHome error but I don't know what it means.

i captured the logs up until i noticed that the lovelace card changed to the correct panel status.
Panel does get to powerlink mode.
I'm not sure where and if i ever changed the panel number-i only have 1 (and a template for another for homekit bridge (as i do not use the night/vacation modes - this template removes them from homekit).
where do i change it if needed?

thanks for all the help, btw..

@davesmeghead
Copy link
Owner

To change the panel number you'd need to delete the integration out of HA, and then add it again, with the panel number set to 0. If that doesn't work then dleete it, remove all traces of visonic from the HA config files and then add it again. However, if you're happy with it set to panel 1 then just leave it as is :)

@EyalRotem
Copy link
Author

To change the panel number you'd need to delete the integration out of HA, and then add it again, with the panel number set to 0. If that doesn't work then delete it, remove all traces of visonic from the HA config files and then add it again. However, if you're happy with it set to panel 1 then just leave it as is :)

ok thanks, it doesn't really matter to me if it doesn't effect the usage =] .
if we're on the topic of the logs here, might I slip in another question? how can I remove from the UI the "night mode"? is it something from the integration, or from lovelace and UI from HA?
image

@davesmeghead
Copy link
Owner

davesmeghead commented Apr 30, 2023

It's that good of an idea that I've uploaded version 0.8.4.2 to github. I have not made it a HACS release yet though so to try it you'll need to download the zip file from Github and unzip it to replace the files if you know how to do this.

There are 2 new Configuration settings to enable/disable Home and Night. If you give it a try and let me know if you think it's OK.

It wasn't really necessary until HA have implemented that panel as the GUI panel you need to manually select which buttons are shown by editing the dashboard, like this
image

@EyalRotem
Copy link
Author

Works fine for me!

but I did find a mistyped 'f' I think in the client.py file:
image

@davesmeghead
Copy link
Owner

OK thanks, I'll tidy it up and release it as it's a simple change that doesn't affect any other functionality

For the "f" string formatting, it's part of python, look here

@EyalRotem
Copy link
Author

Oh cool, it looked odd since the last version didn’t had it and the first word after it coincidently started with an ‘f’ as well 😅

it looks similar to what ‘$’ does in c# which I’m more familiar with.

@vettronics
Copy link

Hi guys,
Today i updated ESPHome to 2023.4 and had the connection problem.
I have rolled back to previous ESPHome version and its working again.
My question is: will new version 0.8.4.2 fix this problem or do we still have to wait for ESPHome changes?

@EyalRotem
Copy link
Author

As of now-I’m on 2023.4.3 even, and don’t have the issues.
If you can provide logs as Dave requested it might help.
I was unable to reproduce it.

something I and @konradwalsh did that might helped- is a full restart of the home assistant machine/operating system.

0.8.4.2 will only add a way to disable alarm type in the GUI AS presented in a previous reply (or maybe some more other stuff- that’s up to dave🤗)

@davesmeghead
Copy link
Owner

If the integration worked with an old version of ESPHome and doesn't wotk with a new version of ESPHome then the first place to look is ESPHome. So it could either be the ESPHome yaml file that I have on the wiki or something within ESPHome itself. As I haven't changed the yaml file (and I assume neither have you) then is must be ESPHome itself. I would assume that new versions of ESPHome should be backwards compatible but sometimes they are not.

So the first place to check is ESPHome itself. After you update, as @EyalRotem suggested, try restarting everything! Including the Wemos (or other gadget) on which you have installed ESPHome. After you updated ESPHome to the new version, did you reinstall the yaml config to the Wemos?

Although 0.8.4.0 made many changes to the HA Event mechanisms, release 0.8.4.0 up to and including 0.8.4.3 would not have prevented ESPHome from working.

@vettronics
Copy link

After the update, I rebooted both HA machine and the ESP8266 that is connected to the panel with no success...
The log I get from selecting "enable debug logging" on the visonic integration comes with all kind of gibberish from other integrations. Is this the one needed?

@EyalRotem
Copy link
Author

After the update, I rebooted both HA machine and the ESP8266 that is connected to the panel with no success...
The log I get from selecting "enable debug logging" on the visonic integration comes with all kind of gibberish from other integrations. Is this the one needed?

You might need to change this temporarily: https://github.com/davesmeghead/visonic/wiki#logger-settings

@isablend
Copy link

isablend commented May 5, 2023

This week I updated ESPHome on the D1Mini to 2023.4.4 as well as HA to 2023.5.0 and experienced the same loss of connectivity. Based on notes from the issue raised I tried a combination of various restores to earlier versions and I could only get it working again with (HA) 2023.4.5 and (ESpHome) 2023.3.2. I did try just downgrading ESPHome, but that wasn't sufficient and I did try full reboots between restores. So for me at least it seems to be a combination of HA and ESPHome causing the issue.

@EyalRotem
Copy link
Author

This week I updated ESPHome on the D1Mini to 2023.4.4 as well as HA to 2023.5.0 and experienced the same loss of connectivity. Based on notes from the issue raised I tried a combination of various restores to earlier versions and I could only get it working again with (HA) 2023.4.5 and (ESpHome) 2023.3.2. I did try just downgrading ESPHome, but that wasn't sufficient and I did try full reboots between restores. So for me at least it seems to be a combination of HA and ESPHome causing the issue.

since i cannot reproduce the issue as it "fixed itself", you might wanna try updating again and producing logs - this might help dave find the issue.
when i tried moving between versions, and restarting and all - it just fixed itself.
I'm on HA 2023.5.1 + ESPHome 2023.4.3 now and everything is working as it should

@isablend
Copy link

ok, one further update. I've restored back to my base environment (HA - 2023.5.2, ESPHome - 2023.4.4 and Visonic - 0.8.3.0). I then upgraded to HA to 2023.5.3 and ESPHome to 2023.5.0. All working. I then installed the ESPHome upgrade to the D1mini and the integration stops working, without upgrading the integration itself so still on 0.8.3.0. So for what its worth it looks like the issue is with ESPHome. Not sure how to solve it, but...

I have the log file from the point at which it stopped where can i post it?

@EyalRotem
Copy link
Author

Pastebin should be fine

@isablend
Copy link

sorry going to appear a bit thick here... where do i find pastebin?

@EyalRotem
Copy link
Author

Just paste the log within this site - and copy the url ..
https://pastebin.com/

@isablend
Copy link

doesn't work, got a size exceeded message and no URL

@isablend
Copy link

One further small development. I've restored the baseline backup (HA - 2023.5.2, ESPHome - 2023.4.4 and Visonic - 0.8.3.0), however I updated the D1mini before this to 2023.5.0 and I've confirmed that it is still reporting as such. So I wondered if I update HA and ESHome (2023.5.3 and 2023.5.0 respectively) whether it would continue working? and it does.
So the only thing not yet updated is the Visonic integration (it's still at 0.8.3.0).

@davesmeghead
Copy link
Owner

davesmeghead commented May 17, 2023

I have ESPHome v2023.5.0 and HA 2023.5.3 i.e. the same as @isablend
I have updated the D1 mini

I have tried it with my current Visonic Integration version (which I haven't released to you all yet) and it works
I have restored the Visonic Integration version back to 0.8.3.0 and it works

The only thing that is different in my ESPHome system is that I'm trying it with encryption, I've attached the visonic.yaml file that I am currently using (as a zip file).
visonic.zip

The important change I made for encryption is here in the visonic.yaml file

# Enable Home Assistant API
api:
#  password: !secret device_password
  encryption:
    key: !secret esphome_encrypt
  services:

You will need to add this to your secrets.yaml

# go here to create a random key https://esphome.io/components/api.html
esphome_encrypt: <Put Random Key Here>

As it says, create a random key and put it in the secrets.yaml file, without the <>. It's under the "Configuration Variables", "encryption" and "key". Then you'll need the same key in the ESPHome Integration so you will need to delete the integration from HA and create it again but using the key this time.

The only reason for me doing this is that it said the ESPHome API Password is deprecated and to start using encryption so I decided to try it in my setup. But deprecated does not mean removed altogether so the existing api password should still work.

Anyway, let me know if this helps please

@davesmeghead
Copy link
Owner

As an afterthought...
It's interesting that you're using 0.8.3.0 as that's the first release to use the updated pyserial_asyncio library at version 0.6.
So that rules out that as a potential problem.

@isablend
Copy link

@davesmeghead I also have the encryption line in my esphome file for the D1min. The only other difference I could see was the header, so you had;

esphome:
name: ${plug_name}
platform: ESP8266
board: d1_mini

I had;
esphome:
name: ${plug_name}
esp8266:
board: d1_mini

I've updated to reflect your position, re-compiled from clean and re-installed the D1mini, but no impact. I've taken log files pre and post integration update only (after it dosn't work), but they only seem to show that there is no attempt to connect; ie. I'm missing the lines
2023-05-18 12:16:21.852 DEBUG (MainThread) [custom_components.visonic.client] ........... attempting connection
2023-05-18 12:16:21.852 DEBUG (MainThread) [custom_components.visonic.client] Visonic Connection Device Type is ethernet {<PyConfiguration.DownloadCode: 0>: 'AAAA', <PyConfiguration.ForceStandard: 9>: False, <PyConfiguration.ForceAutoEnroll: 1>: True, <PyConfiguration.AutoSyncTime: 2>: True, <PyConfiguration.PluginLanguage: 3>: 'EN', <PyConfiguration.MotionOffDelay: 4>: 120.0, <PyConfiguration.SirenTriggerList: 5>: ['intruder'], <PyConfiguration.B0_Enable: 6>: False, <PyConfiguration.B0_Min_Interval_Time: 7>: 5, <PyConfiguration.B0_Max_Wait_Time: 8>: 30}
2023-05-18 12:16:21.852 DEBUG (MainThread) [custom_components.visonic.pyvisonic] Setting TCP socket Options

@isablend
Copy link

Pre- update: https://pastebin.com/VsRU4Xb9
Post-update: https://pastebin.com/hx8hdbnm

@EyalRotem
Copy link
Author

Not sure, but I think k the post update doesn’t have the logger set to :
https://github.com/davesmeghead/visonic/wiki#logger-settings

Which might make it harder to decipher.z

@isablend
Copy link

once more unto the breach...I've re-done the restore and then the update, underneath is the logger output...
https://pastebin.com/wXAaF50F

@davesmeghead
Copy link
Owner

Hi @isablend
With the Post Update and latest release can you try:

  • restart HA and give it time to fail, a few minutes at least
  • go to the Integrations page, click the 3 vertical dots and then Reload.

Can you show me the log extract for this time period please.

@isablend
Copy link

isablend commented May 19, 2023

@davesmeghead
currently on 0.8.3.0, then per instruction, upgraded integration, restart, reload integration. Attached logs
https://pastebin.com/5tGQ9xNa
Will retain in this state until you come back, in case you need other activity.

@HumanSkunk
Copy link

So an update from me. Had some free time this morning so decided to go down the upgrade path once again to try and complete reinstall again but this time capture logs. However upon redoing everything it connected without error and I can control the alarm from within HA.

Steps I took:

  1. Updated everything in HA before changing anything with the integration. HA Core at 2023.5.3, ESPHome at 2023.5.1 and the build on the D1 also at that version.
  2. I removed the integration
  3. I deleted the integration from within HACS
  4. I used File Editor add-on to confirm everything had been removed
  5. Restarted HA (not a reboot)
  6. Reinstalled the Integration through HACS
  7. Restarted HA (not a reboot)
  8. Configured the integration through HA
  9. All connected fine.

The only difference between this as the last time I ran through was that ESPHome was then over version 2023.5.0. Perhaps they changed something with how the boards communicated and then reverted. I cant see anything in the release notes indicating that. If something didnt change in ESPHome then I really cant explain what was different. I can confirm I followed the exact same steps last week and I got nothing back from the board. Could argue I perhaps fat fingered something... but either way it's now fully working. @isablend perhaps try the latest ESPHome version and try again?

@isablend
Copy link

isablend commented May 19, 2023

@HumanSkunk @davesmeghead
I followed the steps laid out by @HumanSkunk and still no difference for me.
I am getting the Notification
image

The logs are here;
https://pastebin.com/VsRuX2QZ

Interestingly it looks as though the integration is, despite being loaded being ignored (from what I can read).

@isablend
Copy link

One last thing. I had a look at the 'Download Diagnostics' option on the integration and get the following;
"visonic": {
"version": "0.8.3.0",
"requirements": [
"pyserial_asyncio==0.6"
]
},

and that is despite removing the integration and from HACS and re-installing. That doesn't strike me as right?

@HumanSkunk
Copy link

When you are removing it from HACS are you sure it has removed everything? I know it should have but sometimes HACS left some files that I also had to manually delete. After you have removed it I would just verify that all the old files have gone from within the custom_components folder that is in the same folder as your configuration.yaml.

Also as a complete clean when you reinstalled your ESPHome configuration to the board did you also remove those files that are often left behind? they are in the esphome folder in the same master folder as above. there were also some files within a hidden folder .esphome as well. Granted deleting the above in the ESPHome folder didnt solve the issue for me, but perhaps a combination of all the things above did. Does make it more annoying to work out what the actual problem was in the first place...

This may not be in anyway linked to the issue but could be worth a try, especially when its reporting to be at a different version.

@davesmeghead
Copy link
Owner

For this...

Hi @isablend
With the Post Update and latest release can you try:

    restart HA and give it time to fail, a few minutes at least
    go to the Integrations page, click the 3 vertical dots and then Reload.

Can you show me the log extract for this time period please.

You used the old integration version, I asked you to do this with the latest release
Can you do it again with the latest release please

@davesmeghead
Copy link
Owner

When you see this

One last thing. I had a look at the 'Download Diagnostics' option on the integration and get the following;
"visonic": {
"version": "0.8.3.0",

It's probably my fault, I didn't update the version code in the manifest.json file and it still says 0.8.3.0
I've just updated manifest.json on guthub

@isablend
Copy link

@davesmeghead vis the version of the integration. The logs I took were after I had upgraded the version to 0.8.4.1 hence my confusion as to why it seemed to be reporting 0.8.3.0. Each time I remove the integration from HACS I check that it has removed the files from custom_components. I've tried both removing manually and from within HACS.

Unfortunately I'm away now for a couple of days so won't be able to make more changes, but will apply any new findings when i return. In the mean time I've had to return to the backup working version.

@isablend
Copy link

@davesmeghead I'm back and decided to go back and re-check each activity as its done. From within HACS I applied the update it was advertising. A re-start is then required. However before I did that I decided to check to see what was in the custom_components directory. All the files are date and timestamped as you would expect with todays date and time. However based on your comment about me failing to update the integration before re-starting I thought I'd check the manifest file you mentioned. This does not include the update that your github file is showing. I.e. it still shows 0.8.3.0 which implies that whatever HACS is doing it may not be doing the update I thought it was doing. I'm going to manually download all the files and see what impact this has and will report back with findings.

@EyalRotem
Copy link
Author

Dave said 2 comments ago that the manifest file wasn’t updated at his fault, and if I understand correctly-it affects logging but not how the integration works - as most of us have no issues🤗

the file is only updated on GitHub since the latest iteration isn’t released yet

@EyalRotem EyalRotem reopened this May 24, 2023
@isablend
Copy link

That's why I mentioned it. I updated the files using HACS and yet the file stored where the updated HACS file should be had the old information in it, not the new. So not within HA, but the file which should then be used as the basis for HA logging. That tells me that there may be in issue in the delivery of files from Github with HACS not with the integration itself.

@isablend
Copy link

ok, one further update.
I've been able to consistently go back to a backup of the system including integration 0.8.3.0 throughout this issue. When ever I've got to a position of stalemate I have reverted back to that backup and all works.
I've manually forced the individual files updates by downloading each file from Github to ensure that that code is in place where it should be.
I've today actually achieved the upgrade from 0.8.3.0 to 0.8.4.1 but that it required I remove the integration and then re-install and then that a restart of the alarm panel is required before it will communicate. Bottom line it is now working, can't say I understand why.

@davesmeghead
Copy link
Owner

I'm really happy that you've got it working. It shows me that it's not the Integration at fault but has been a problem with either HA itself or with HACS doing the upgrade. Although you should not have needed to restart the alarm panel.
D

@davesmeghead
Copy link
Owner

I'll close this issue but any further problems then create a new issue and refer back to this one if appropriate.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

6 participants