Hi OH community,
I’ve been a very happy user of the MIOS binding for a very long time. It’s worked well for many years and I really appreciate the work of everyone who contributed.
However, I’ve always wondered if there’s room for better performance. Has anyone ever tried to migrate from native zwave to MIOS binding or vice-versa? Can anyone share their experiences as to which is better?
Thanks!
Just looking at the description of the MIOS binding for OH there is a lot of transformation going on with that.
Is there anything in the native binding giving you issues? @chris , @sihui, & @5iver would be very interested in helping.
Using the MIOS binding means you have a Vera ZWave (or general HA) controller, right ? The binding is meant to be temporarily used while you’re migrating your devices over to openhab.
If you are willing to move your ZWave devices to a new controller (means you need to reinclude) powered by OH ZWave binding then you can get rid of the Vera so in the end you’ll have less HW in operations with less pieces that might fail.
Or what do you understand by “better performance” ?
Sorry, I should have been more clear. I’ve never used the native binding. I originally came from vera and stumbled across openHAB. I used the MIOS binding as a stopgap measure to migrate from vera to test openHAB. I found how awesome it was, and all my rules have been on OH ever since. It’s one of those situations where everything just worked, so I never made the jump to fully migrate off the vera controller.
Now, I’d like to see if using the native binding would yield better performance. However, the only way I can compare is to do it in “production,” Before making such a big change and re-pairing all my (50) zwave devices all over again, I’d like to see if anyone has been through a similar experience?
Yes, that was my intention as well. However, it went from a temporary to permanent (3 years+) situation.
The MIOS binding has been surprisingly stable, but sometimes there’s a slight delay in switches turning on/off, slight delays in motion sensor reporting, etc.
I’d like to know if the zwave network on the native binding might be faster?
I would first check to verify all of my devices (by mfg, type, id) are listed in the zwave device database here.
If any are missing, the developers are very helpful in getting them included. Sometimes they need the xml file generated by OpenHAB when a device is not recognized though.
It’s as fast as the network allows it to be. There’s of course less processing overhead but if you see delays it’s likely due to the Vera being forced into being a bridge rather than a controller so yes you can expect those delays to disappear.
I migrated from using a Vera Lite and Mios binding earlier this year…id been using it quite successfully and reliably for many years …i think since the first Mios binding was released.
However, it was sometimes slow… and if kids got onto a wall tablet and turned stuff on/off/on/off/on/off …it would choke up…
Then my Vera started hanging…luckily i had a second unit and put it into service, but i thought since im going through this pain, lets try using a Zwave USB stick ( i had a Aeon v2 stick in my box-o-bits) and was able to somewhat successfully move my Zwave network across and set all devices up in OH quite easily… a few hours work…the performance was better than with Vera but…the Zstick would occasionally stop working and need to be unplugged etc…
So i ended up buying a Aeon Zstive Gen5 and rebuilding my Zwave network from scratch - this meant visiting all the modules, but many can be reset by pressing the wall switch button multiple times etc…so it wasnt too painful about 2hrs to add 25 modules or so… but the end result is awesome! The performance is now instantaneos ( or close enough!) and its rock solid…i havent had to touch the dongle since.
Sorry for the long story… but i absolutely recommend getting a Aeon Zstick Gen5 and start a fresh network… anp put your Vera to rest.
Awesome, thanks for the feedback!
That’s exactly what I wanted to know before rebuilding my network.
I guess it certainly depends on your setup. But I use the Mios binding w/ a Vera 3 and have been very happy with the response time which for me is almost immediate. Devices are always in the database and I can put the vera 3 centrally in my home while I run Openhab from a server in the basement. I haven’t found a reason not to use the vera as my zwave controller.
Thanks for everyone who replied.
Just to report back: I wound up testing this out and at least for me, I’m afraid the results were disastrous.
The zwave binding was great when I tested on a small scale (under 10 devices), but after I started including my whole network (30+ devices), things started going sideways. The problems were persistent enough where I had to hard reset the controller several times. Some issues encountered include:
This is definitely not a knock against the OH zwave binding. I think it’s pretty amazing, and the amount of community work that’s gone into it is remarkable. I really wanted it to work since I like the notion of embedding everything cleanly in one device.
However, zwave is complicated, and unfortunately, in my specific case, it didn’t work out as well as I was hoping. After two weeks of fighting with it for stability and performance, I had to revert back to using the Vera for zwave control.
Did you ask for any help here with your OH zwave binding issues? @chris is the very helpful developer who is a Z-Wave expert and works with SiLabs documentation to understand the standard.
@sihui and @5iver are also a big help here.
If you did not ask for help with your issues you did not give the binding a fair test. Some of us have turned off the daily heal, for instance due to some battery devices hanging.
Yeah, I did. The support from everyone has been great, and I’m grateful.
In fact, you helped reply to one of my issues here: Zwave after heal: Node is not communicating with controller. Sadly, I couldn’t resolve the connectivity problems even after turning off auto heal.
I definitely wanted to give it a fair shake and tried to wait it out, but fixing the issues was burning up all my free time, and two weeks without automation in our household was getting to be too long.
Anyway, this isn’t intended to sound like a complaint or again, a knock against the binding. My situation is just one case, and I fully expect other people have probably had much more success than me. Just wanted to feedback to the community for anyone else who might have the same question.