Indeed it did not solved anything … I tried I rollback let see if it is more stable.
I guess nothing else to do except waiting for a fix to be availble …
It is anoying that for every new version of the zwave plugin I have trouble, hard to find a stable version.
???
7.17.2 was the version I basically had to run from day 1 on the Z-Stick 7 Plus in order for it to work at all. It’s not new, and there are a variety of firmware versions available in 7.18.x form as well as a couple in 7.19.x. I have basically run them all with no discernible difference in how they work.
Sorry if my last message was not clear. So changing the version of the zwave sitck did not changed anything. I hence rollebacked my whole HA instance to a previous version. Seems to be working fine so far.
From everything I’ve seen it’s the ZWave JS 11.x → 12.x update that broke everything - but no one from HA (that I’ve seen) has commented on what really changed / why it changed / what broke…
Hubitat (knock on wood) has been stable for me for the past week since I moved away from HA USB radios…
I raised a ticket with Aeotec, and they’ve provided a newer version of their firmware (7.20.2), however I’ve not found it any more stable than the previous version I was using (7.19.3). They have offered to check log files to see which devices are generating the noise, but I know which 2 sensors it is (Aeotec multisensor 6s), so I’ve tried to reduce the amount of chatter they generate.
I can’t imagine 7.20.2 is going to correct anything. There has been little to no discussion about the SDK it would be based on receiving any actual fixes, and the ticket that -I- have open with Aeotec has not yielded any directive to update. I suspect you are being used as a guinea pig. I would tell Chris that the newer firmware is not operating any better than the previous if you are still seeing the jammed issues.
FWIW, they replaced my z-stick and I have been running 7.17.x for a while to avoid the 7.19.x issue. I have not gotten any relief from the problem which means that it isn’t firmware related and is likely tied to something within HA that isn’t being acknowledged.
Hi Mark,
if you’re already running an older version of the firmware, wouldn’t this imply that this would be down to either Z-Wave JS or Z-Wave JS UI, or the server underneath? Does the issue affect both addons? I’m using Z-Wave JS. Is it possible to downgrade the server to an older version?
I reported the lack of change to Chris, to which he said he would review log files, to try and identify any chatty devices.
That’s exactly my point… There has been much discussion about a possible issue in the 7.19 SDK which would have meant problems in the 7.19 firmware. I am running 7.17.x and the issue is present there which means it is NOT exclusive to the 7.19.x firmwares.
The issue cropped up with an update to HA in early-to-mid September where not only would the “Jammed” status appear but the system would essentially stop responding / controlling devices. Once the stick would be reset and put back to “Ready”, it would start working again. No one seems to be acknowledging that there is likely something wrong in the underlying code of HA (whether it be ZwaveJS or elsewhere).
1 Like
Yes, EXACTLY, @ember1205.
No one is acknowledging that HA Zwave JS 11 → 12 is the culprit.
I realize HA is a ‘free’ platform, but I do pay my nabu casa monthly dues and this issue has cost me about 200 bucks and countless hours in changes.
Do you know what version of the firmware Hubitat is using?
Seems to be conflicting identification of the cause, for example here, but I wonder if anyone that is involved in the z-wave js server is investigating, if hubitat is using the same firmware.
Honestly, I’m not smart enough to understand some of this.
@ember1205 posted an issue on GitHub (to which I also posted), but haven’t seen (personally) any traction / fwd movement there: Controller reporting "Jammed" and then "Ready", operations hang · Issue #104230 · home-assistant/core · GitHub
Long story short - in 3+ months I’m not sure I’ve seen a single person from HA or responsible for the ZWave piece(s)[?!?] of HA that are causing this respond to or work towards fixing it.
It’s why I finally threw in the towel and said F… HA and their USB / local / direct radio control (Zwave/Zigbee) and moved back to a hub. As my post in the above Git thread shows - I was on SmartThings for years prior with few-to-no issues (until Samsung purchased / screwed up) and recently decided to go w/ a Hubitat C8 to handle ZWave/Zigbee (bye-bye SkyConnect [which wasn’t 100% for me either]) too).
Anyhow, relatively happy again now (as is the wife), just sad it cost me $200+ and dozens of hair-pulling hours.
Yes, I know, I know, 10 yrs later (congrats!) HA is still in ‘beta’, but it’s ridiculous there are this many stupid issues in making your dumb home ‘smart’.
1 Like
Back when I was researching around Zwave JS UI before migrating from my SmartThings hubs there were already complaints of Gen7 Aeotec misbehaving as well as some of the latest fw for Gen5. That’s why I decided to go with a Gen5 and keep the same firmware it came with.
Have not touched it since (more than a year). So far have 30 nodes and no problems whatsoever.
The dongle is on a dedicated Pi in a strategically central location in the house and sends the zwave data to HA over websockets.
I’d been running an AEOTech gen 5 plus with whatever firmware it came with three or four years ago when I first installed it. Never had any issues with it until September with the 11 to 12 update.
Thinking it was just my stick that was outdated, I went and bought a new zooz 800 series, moved everything over and continued having the exact same issues…
As mentioned above, finally just said scr3w it and purchased a hubitat c8 and moved everything over (again)… Such a hassle.
FWIW, the zwavejs team is very responsive to triaging and responding to issues. I’ve filed several and I always get a response within 24 hours with helpful diagnosis. They currently have 198 open issues.
HA, on the other hand, seems to have no process to triage and respond to issues. They currently have 2,200 open issues - and most of the closed ones are autocloses by the bot due to lack of activity. Certainly triaging issues is a tough job, but perhaps one that Nabu Casa could fund. As a contributor to the code base, I struggle with sifting though and finding issues to fix and hence having them go through a triage would make it easier to pick one to work on.
Just to add a further update, I’ve restored a backup from Sept 23. I’ll advise on how successful this is in a day or two
The upgrade is showing for Z-Wave JS from 0.1.90 to 0.4.3, and checking the log files, Z-Wave JS is version 11.14.2.
Didn’t take long, already seeing the jammed message on a regular basis on the versions as above. I don’t appear to have any backups much earlier than this, unfortunately.
Appreciate you trying…
However - not sure that you would have been able to actually gain any ground as (IIRC) the developers have claimed that the upgrade that “broke it” didn’t actually do anything except expose the error in the UI. The claim is that it was happening prior but wasn’t visible.
Maybe this is a good time to look into the grandfather-father-son backup rotation strategy.
Is anyone who is having the problem using ZWaveJS on a remote host?
I don’t suppose they identified the cause of the errors, and how best to resolve?
Step 1 is acknowledge, Step 2 is identify. Let’s see if Step 1 happens.
2 Likes