I think that's what I have, but I'll confirm.
1 Like
I also checked the Z-Wave details for my C5 hub and see no devices missing the in/out clusters.
Most of my switches are the Leviton DZ6HD dimmers, so I switched to the Leviton DZ6HD driver and I noticed that the issues of multiple events went away.
I was previously using the Generic Z-Wave Smart Dimmer driver instead of the specific Leviton driver because there is an issue with the driver that all events are displayed as "physical" no matter if I turn on the switch from a dashboard, GH, or physically.
So now at this point I have two broken drivers, I even brought up the issue about the Leviton driver but it was dismissed as not an issue and I doubt anything will be done about it, so basically its a pointless driver.
I also tested a couple of the Leviton DH15S-1BZ switches and also experience multiple events with the Generic Z-Wave Smart Switch driver as well.
Also all of the Leviton Dimmers/Switches that I have are on the latest v1.20 Firmware.
There's a scenario that hasn't been mentioned yet.. mesh latency/errors, and because of Retries.
If the mesh is just weak enough, most packets get through, but some don't and the retries gets out of sync... The hub, thinks the device isn't hearing the Commands and resends, the device keeps thinking each retry is a brand new command and sends the data again, and again, and again.
Part two of this is that many devices INTENTIONALLY send two answers for some commands. There's a Basic answer and often a more specific answer. (I corrected this in a recent version of the Aeotec Multisensor 6 driver I'm fiddling with, endlessly. 
) The Multisensor will, for Motion events send two messages, one is a Notification, specifically for Motion/Tamper. It also sends either a BasicSet message or a SensorBinary message, which one is configurable by the driver itself. I realized that the motion portion of the Notification message is redundant.. I altered the driver to ignore the motion element of that message and only process it as if it were a Tamper message.
When I see people talking about duplicates in sets of two, I automatically think: Driver's processing both messages from a device... When I see people talking about duplicates in sets far larger than two, I automatically think, weak mesh.
Explains why the proper driver has fixed it for me on multiple occasions as well as when helping others.
In previous firmware versions this issue was not present with the built in generic drivers, so something must have changed.
I am seeing sets of 7 events but what I noticed in the logs is that half of the events are shown as "physical" and the other are "digital." It still seems like a driver issue. All my devices are z-wave plus and I have around 35 mains powered switches/dimmers all over the house. So I doubt this is a mesh issue. I also live in a large lot where the closest neighbor is like 1000ft away, I can't see there being interference either.
I also just noticed that it's different when I turn off the light.
Thanks. Good to keep in mind that duplicates in sets of two may be perfectly normal for sensor devices - i.e., because of Basic Set versus Notification messages. What I"m seeing is different - I'm seeing burst of multiple messages from switches and dimmers. For example, in the image shown below, there was a burst of 20+ messages all saying that a dimmer (Bathroom Toilet) was turned off.
As for the weak mesh theory - you've given me an idea there, I assume I can check the logs for each device and see if there is a small subset of devices doing this spamming and, if so, how many hops they are away, or if this is random. I think its random, but I'll check that tonight.
Note in this post: Z-Wave storms - looks like HE Hub's fault? someone did a Zniffer trace and found the hub was spamming the network with Get messages.
And, interestingly, I just noticed this: Spurious discovery frames [C7] which may be unrelated or may suggest there's something odd going on low-level.
Yeah, @jvm33 your seeing much more events then I am. After looking closer at mine they are only in sets of 2, with the exception of the dim percentage with 3. There may be two different issues here even, or possibly still related.
I am not trying to say that a storm isn't a storm, there could be storms, but do I think they are the FIRST thing to imagine?? It's the old hoof beats question, why think of Zebras? I'm ok with thinking Storm after determining it's not a mesh effect... 
v2.2.3.145 is the latest and in the area of C-7 ZWave, the best, thus far. The Storm Topic was a factor in the last week of July and the first week of August.. a month ago. I am ready to give v2.2.3.145 the benefit of the doubt here, that those issues got quashed.
I've run Zniffer, almost continually for the past month... (a few times a day I'll stop then restart the capture to discard everything to that moment,) but I'm not seeing discovery frames in that quantity.. yes, they do occur, just not (for me) in any alarming pattern.
This has all been helpful - one other thought I had - I have a HomeSeer Hubitat Plugin that I'm using. I'm going to disable that tonight to see if, just maybe, that plugin is generating lots of "get" request - I think that's a long shot, but it still makes sense to eliminat possible third-party plugin interactions. I've already taken the Desktop app off my system, so
@lewis.heidrick. Thanks for all the suggestions yesterday. I've noticed there's been a firmware update about once a week for C7 issues - I'm going to wait until the next one, see what happens after that, and then resume trying to fix this problem.
Also, on another theory, I think the last few times I saw the problem, the problem device was in a lighting "Group" - maybe there is a problem with the Groups/Scenes app. I'll experiment with this theory a bit more (i.e., by getting rid of the use of Groups ).
There is still an issue with generic drivers causing sets of 2 events for every action, which seems to happen irrespective of the hub version. I also suspect this to be the culprit of the increased latency. I just don't want this to be overlooked because I know there is a lot of changes happening to the C7 version.
Also the issue of the driver displaying 2 sets of events for both digital and physical, for just 1 action. Hopefully the Leviton DZ6HD driver is also fixed, as this has been broken since it's release.
I know a lot is going on with the C7 but these latency /driver issues were not present in firmware that was released before the inclusion of the C7, so it really should be a higher priority.
1 Like
I have the same issue of this when I use a community driver instead of the generic driver for my Honeywell Z-Wave Plus Dimmers (Non Enbrighten version). Using the generic driver doesn't cause the issue.
I have mostly the Leviton DZ6HD dimmers, but I do have 1 of the GE 14294 dimmers which also exhibits the problem as well. It may be that this only affects certain dimmers/switches, but either way this is a bug. Personally I would prefer to use the Leviton driver that was written specific for my devices, so hopefully the issue is fixed for both the generic and the Leviton drivers.
I switched back to the generic driver for the honeywell switches and it mostly went away although I still have a zooz power strip that is very enthusiastically telling me it's on. At least it isn't locking up anymore. Hopefully they'll track it down.
The issue still exists with the generic driver for me. If I switch to the Leviton driver the multiple events disappear but then all events are reported as physical, so it breaks automations like motion lighting then may need to know if the event is digital/physical for certain overrides.
Although even at this point with the generic driver I see both physical and digital events, so IDK what is working.
Sometimes devices don't handle physical or virtual events and just send whatever they feel like. I saw this a lot on ST, not so much now on HE but occasionally. I haven't messed with Leviton switches much other than regular occupancy switches (non-zwave/zigbee)
There was a case I saw a long time ago about someone using the non smart generic dimmer driver and it working correctly for them. I'll see if I can track it down.
Also came across this article that discusses Leviton dimmers and physical presses but it's quite old also.
The Leviton dimmers are Z-Wave plus and with the generic driver they did work in the past with the generic driver. It used to separate the events into digital/physical correctly and did not have multiple events, so this was broken at some point.
The Leviton driver never worked from its inception when it came to separating the events by physical/digital.
The thing is the Leviton DZ6HD driver is specific to that particular device, if it does not have the same functionality/compatibility as the generic driver then what is even the point of having the driver? It rubs me the wrong way too because it seems that the HE staff has no intention of fixing this driver, and if this is the case then why even have the driver included if its broken?
