Zwave Aeotec Gen5

Posted on
Wed Jun 15, 2022 10:24 pm
dduff617 offline
Posts: 592
Joined: Jul 05, 2006
Location: Massachusetts, USA

Re: Zwave Aeotec Gen5

i am having a problem that sounds very similar.

i mostly don't use z-wave a lot "interactively" - the majority of my z-wave devices currently are things like thermostats, sensors plus door locks.

recently, i got some z-wave modules on amazon because a) they were cheap b) they said they were "700 series" z-wave. i wanted to build-out more a more robust and reliable connection to some devices i need to be reliable (door locks). these were "minoston" brand. i was paying more attention than normal to indigo and z-wave and trying to get these devices to sync and i observed that most attempts to communicate with these devices (e.g. by sending a series of on/off commands over a short span) would seemingly cause z-wave to go through a repeatable sequence: lock up and stop responding. all z-wave commands (including to/from other devices stop working for a few seconds), then z-wave interface shuts down and fires a trigger i have setup to warn of "interface connection failure". z-wave connect re-establishes itself and comes back.

at first, i thought this was something unique/specific to these new modules. then i did a bit more testing and it seemed like i could reproduce the same set of symptoms by just sending on/off commands to some of my other z-wave devices. at that point, i started to think possibly this could be a symptom of one or more of a) recent migration to OS 12.3 then 12.4. b) recent migration from old mac hardware to newer mac hardware a few months ago. c) recent migration to Indigo version 2022.1. i also briefly entertained other ideas like that my z-stick was going bad.

if it is useful, i can send logs from multiple iterations through lockup / restart process but honestly from a glance, there's didn't seem to be much interesting detail there. the good news is i think i have a fairly reproducible test case with my current system. let me know if there's something i can do or test to help shed light on this.

Posted on
Thu Jun 16, 2022 9:28 am
jay (support) offline
Site Admin
User avatar
Posts: 17487
Joined: Mar 19, 2008
Location: Austin, Texas

Re: Zwave Aeotec Gen5

What generation of z-wave are the other devices (besides the new Minoston)? The question is are they all 700 series or not?

Second, when the Z-Wave interface stops working, what (if any) errors show up in the Event Log (including anything right before, during, and after Indigo reconnects).

Jay (Indigo Support)
Twitter | Facebook | LinkedIn

Posted on
Thu Jun 16, 2022 10:27 am
dduff617 offline
Posts: 592
Joined: Jul 05, 2006
Location: Massachusetts, USA

Re: Zwave Aeotec Gen5

i experienced the z-wave interface dropping off when sending a series of commands to a Zooz Zen26 (multi-outlet relay - not 700 series, afaik).

i thought i had a "reproducible" test where i could take the interface down, but when i tried sending a series of commands to the Zen26 just now, i did see many "send <cmd> failed" messages but i did not see the interface go down. so perhaps i was incorrect about the reproducibility. OR its possible that my physically unplugging and re-plugging my z-stick from the computer yesterday somehow cleared up the problem?

here's an odd thing. i tried sending some commands to my Zen26 unit just now. tried some on, off, on, off sequences separated between maybe half a second apart to 2 seconds apart. it did not seem anywhere near capable of "keeping up" - even when i slowed down to like 5 seconds or more between.

but now as i type this, it's been a full 2 minutes or more since i stopped the test and i can hear the air cleaner attached to the Zen26 is toggling off and on. so it seems clear that the Indigo and/or the Z-stick has a buffer of backed up z-wave commands and was waiting to transmit. and the time it was willing to wait was extremely long! that's kind of a surprise to me. it seems like in clearing the queue of buffered stuff, the z-stick may be waiting for some long period with no other transmissions. ( my fan air cleaner toggled on/off again - at least 4 mins since i stopped sending commands.) so perhaps with the sensors (motion, temperature), thermostats, and other devices, my network is not meeting the "idle" condition often enough for the z-stick to clear the queue?? (air cleaner toggled AGAIN).

so that seems like a sign that there's something not "healthy" about my z-wave network. noise? a corrupt device spamming the network with junk signals? how is it, i wonder, that the z-stick could take 5 mins or more to transmit some basic on/off commands?

i went back and tried sending on/off commands to the new Minoston devices i added the other day to see if i could get back to a reproducible z-stick reset condition. looks like repeated on/off's to one of those units are working ok now. if i repeat fast enough, i can definitely see some transmit failures, but nothing brings down the z-stick. the other unit can't seem to communicate. i'll try optimize-network for that specific device. that seemed to work (took about 5mins). device was initially unreachable, but it dropped back and did neighbor discovery on other devices and eventually found a path and fixed itself. so hooray for z-wave network optimization. however now i can not send ANY z-wave commands to ANY device. nothing is working. ALL commands are failing (and taking a very long time after i attempted them to show up in the logs). so looks like z-stick is hosed, but for whatever reason has not reset itself (yet). it also seems i've stuffed a bunch of on/off commands into the queue and they're (very slowly) draining out (now several minutes later) and generating error messages in Indigo. so it seems z-stick is "dead" but Indigo just doesn't know it yet. i'm seeing occasional inbound z-wave signals (from sensors, thermostats) but no output commands are working - perhaps because they are queued behind a bunch of other commands? i don't know. it seems i don't have a good testing methodology to isolate issues and discern actual causes of problems.

tried turning on debugging on the z-wave interface, but i realize as a side-effect, this resets the interface, so..... it's reset and back to working state again now.

sorry i thought i was going to make a short post and provide simple data, but i've only added confusion. advice welcome - there are clearly things about workings of z-wave i'm learning.

Posted on
Thu Jun 16, 2022 3:40 pm
dduff617 offline
Posts: 592
Joined: Jul 05, 2006
Location: Massachusetts, USA

Re: Zwave Aeotec Gen5

now that i'm paying more attention to z-wave... here's another possible bug i saw during optimization attempt. i had done an optimization earlier and a few battery-powered devices were deferred for later. this shows a battery-remote unit (for Bali Autoview shades) that woke up and tried to optimize, resulting in crashing the interface:


Code: Select all
  Z-Wave                          Optimizing - starting for awake device "094 - APT Shades Master Remote"
   Z-Wave Error                    Optimizing - update failed: "094 - APT Shades Master Remote"
   Z-Wave Error                    Error in plugin execution:

Traceback (most recent call last):
  File "/Library/Application Support/Perceptive Automation/Indigo 2022.1/IndigoPluginHost3.app/Contents/Resources/PlugIns/serialinterface.py", line 547, in start_comm_thread
    self.process_comm_thread()
  File "/Library/Application Support/Perceptive Automation/Indigo 2022.1/IndigoPluginHost3.app/Contents/Resources/PlugIns/serialinterface.py", line 731, in process_comm_thread
    self.process_queue_in_packet(packet)        # Subclass is required to define.
  File "zwaveinterface.py", line 2882, in process_queue_in_packet
  File "zwaveinterface.py", line 2854, in _processFuncCall
  File "zwaveinterface.py", line 2920, in _processCommandReply
  File "zwaveinterface.py", line 2903, in _processFuncTriggersAndBroadcast
  File "zwaveinterface.py", line 2852, in _processFuncCall
  File "zwavecmd_wakeinterval.py", line 197, in _processNodeAwake
  File "zwavecmd_wakeinterval.py", line 143, in _reallyProcessNodeAwake
TypeError: '>' not supported between instances of 'NoneType' and 'int'

   Z-Wave Error                    exc type is <class 'TypeError'>
   Z-Wave Error                    restarting serial connection in 5.0 seconds
   Interface Failure               lost connection with interface
   Trigger                         Interface Connection Failure
   Pushover                        Pushover notification was sent successfully, title: Indigo Interface connection failure, body: Indigo has detected an interface connection failure on one or more interfaces.  This may mean that either Insteon and/or Z-Wave signals can not be sent and/or received.    Device may need to be power-cycled or physical connection re-established.
   Z-Wave                          connected to Z-Stick Gen5 (ZW090) interface on /dev/cu.usbmodem14301 (firmware 1.00, SDK unknown)

Posted on
Thu Jun 23, 2022 4:58 pm
matt (support) offline
Site Admin
User avatar
Posts: 20987
Joined: Jan 27, 2003
Location: Texas

Re: Zwave Aeotec Gen5

That Event Log snippet is definitely showing a bug. I'll have that fixed in the next release (hopefully released in the next couple of days). I'm not 100% confident that will fix the problem you are experiencing or not, but let's give that a try and see how it works.

Image

Who is online

Users browsing this forum: No registered users and 7 guests