Ok, so I’ve added code to handle this sort of format coming through the config handler and added some more debug in case we find new and wonderful formats being passed in!
This is a bit of a kludge, but as I can’t easily work out what is causing this format to be passed in to the system, it seems the fastest route to satisfaction to just handle it.
After upgrading from 2.3 to 2.4, all wave devices was removed and added back, as suggested by the procedure.
PaperUI was used for this.
FGMS001 stopped sending motion events, but luminance temp etc was reported as expected.
Removed the FGMS001 and added it back several times. No success.
Finally added the FGMS001 using Habmin, and it has been working flawlessly since.
/Jan
After updating to openHAB 2.5.0 Build #1484, things seem to have improved, as in I’mm having more luck with my FGMS sensors working now, not 100% they’re working as expected - they seem a bit more fragile than normal.
I’m still seeing the associated devices being removed in the UI of both Habmin and PaperUI however.
Unfortunately I don’t have debug running atm so have no logs.
Not sure if this makes any sense, but in PaperUI, this is what I see. When I set the Association Groups to “Controller”, that sticks for a short while, then they either disappear, or remain but as “node_1”. You can see from the drop down, “node_1” doesn’t exist. I assume “Controller” has the ID “node_1”;
Thanks for the feedback.
When you get a chance, please provide a debug log that is long enough to work out where the associations are being removed - without this it will be hard to work out what is happening.
Turns out I do have DEBUG enabled for ZWAVE, but not log4j2.logger.openhab.name = org.openhab. Is this sufficient?
edit: Assuming sufficient, they’re available at oh_zw_db_logs.zip
In terms of a time frame, I updated OpenHAB on 31/12/2018 at around noon. I noticed shenanigans with NODE_19 around 14:00 on 01/01/2019
Unsure whats happening with this issue, but I’ve ended up ripping out all my FGMS and gone with the Hue Motion sensors. My household is entirely Hue-ified now, and as I was only really using the FGMS for lighting they’re somewhat surplus to requirements.
I took a quick look at your logs. FWIW, I saw only one instance where the Group 1 (Lifeline) association was removed. After removing it, it was immediately set to Node 1 (controller). This likely was done during initialization (after you updated the binding?) where the binding ensures that the Lifeline association is set to the controller node.
The only conclusion I can draw is that the issue you’re having with disappearing associations is a known display issue with HABmin and PaperUI.
This was on Dec 31.
I’ve only just caught up on this as been away from home for a few months. Came back during that time to find the FGMS001 Sensor not working after an update to OpenHAB. (now on 2.5 Snapshot).
Factory Reset did appear to work, then eventually it stopped reporting movement. Reading over the last couple of months it does seem to be connected to the config updates as I had been sending out an update to turn off the flash on movement while I was away. Config done via PaperUI.
This time back home full factory reset but this time only using HABMIN to update the config. 24hr later still working, have not turned off the LED yet. Have logs running in DEBUG mode so can post if it goes turnip.
One thing I have noticed is that HABMIN send the following to the event log
2019-02-02 14:56:34.485 [me.event.ThingUpdatedEvent] - Thing 'zwave:device:512:node14' has been updated.
2019-02-02 14:56:34.504 [vent.ConfigStatusInfoEvent] - ConfigStatusInfo [configStatusMessages=[ConfigStatusMessage [parameterName=config_1_2, type=PENDING, messageKey=null, arguments=null, message=null, statusCode=null]]]
Where as when I used PaperUI all the pararmeters were sent. Not sure if that helps. Will keep checking and post updates
That is the reason why it is recommended to use HABmin to change parameters: it does not clutter your network because it does only send the changed parameters.
Sensor has been working fine the last few days so today I renamed it to be Balcony Sensor using Paper UI. Log shows the full config going out, A while later the sensor reports its update (along with an Association update) and now it is not reporting back. Fun stuff starts about 8am.
Hope this helps.
I recently changed from another smart home provider to OpenHAB.
When I added my FIbaro Motion sensors they worked, for a couple of days.
Then they lost their association groups randomly it seemed.
After a couple of weeks of testing all the stuff I found a solution for me.
It seems like my Aeotec Z-wave stick, when I put it in “Inclusion mode” something happens with the communication with the sensors. Sometimes in a couple of hours it stops working or a couple of days.
Then I uninstalled my z-wave binding in Paper UI, and installed it again. After I did this I have had no problems at all. Everything works perfectly, but I need to do this every time I put my Aeotec Z-wave stick in inclusion mode.
What exactly do you mean when you talk about " I unbinded my z-wave binding in Paper UI" ?
Hi All
Im assuming no one is using 0801:3002 with V3.3 or, if you are, do you have this working??
Cheers
As we discussed via email yesterday, this is not currently in the database. I added it yesterday so it will be in the next database update.
Hi Chris
I know, I saw some posts back to December so I was surprised no one else was having the issue given we are almost March.
This seems like a rather tiring job to constantly have to update the database (for you) . How do other HA platforms accomplish this?
Thanks!
From the tracking down I did it seemed to only occur if I updated the device config via Paper UI, which is not something done very often. I was also on snapshot at the time and due to the transition back from Eclispe I rolled back to milestone (which solved myOpenHab issue). Logs are available from the testing I did.
I can’t find any reference to this from December (just from a search, so it might miss something), but it’s not very difficult to add the codes to the database and normally there are people to help…
If you wrote a how to (maybe you already did) I’d be more than happy to help with the workload.
Thanks @dastrix80.
There is a database guide here. There are a few people here who can help as well (@sihui, @5iver and @mhilbush are also admin on the database).
