An amusing C7/ZWave mesh failure

I have a Zooz ZEN 16 that I use to switch water valves. One channel is for an Outdoor Pool Shower and the second channel is for a Pool Filler.

A couple days ago the pool was looking a bit low and instead of checking it I just pushed a HomeKit button to add 12 minutes of water. (About 1.5" in 12 min.) I didn't check that either because it always works. :slight_smile: At least turning on always works, but I'm pretty nervous about it turning off.

Yesterday, I noticed the pool was still low and pushed the Home App button again... expecting another 12 minutes. Nothing. I checked via the device page and it wasn't working. ZWave repair failed. I haven't had a device "fall off the mesh" in years but I'd been reading all of these topics where that's what people were describing!!

I rebooted the Hub, something I pretty much never do.. outside of code upgrades and it didn't help. I did a few more ZWave repairs and 4-5 device were out. GRRRRR.. those darn C-7's. :smiley:

Today, I found I had 20 mins I could dedicate to re-adding the ZEN-16. So I went to where it's mounted and looked at it and I didn't notice anything amiss. I found the power wire and followed it down and around to power cycle it and there it was...

The power brick had fallen out of the wall. Plugged it back in and the blue led on the ZEN-16 lit up... which I had forgotten was on that surface, so didn't realize it was off a few moments before.

So yea.. my C-7 completely was unable to plug the device's power brick back in the wall. :smiley:

12 Likes

I thought you were going to say when you tried to fill the pool, the shower was turning on instead! :grinning:

1 Like

Maybe you should setup Device Activity Check, available in HPM.

I think the point was: I started debugging the hub because of the rumors I'd read here... My normal debugging would have started at the device end but I was swayed by the reports of mesh problems. It's almost like suddenly finding Flat Earth theories to be credible. :smiley: So I wrote this up to caution others that mesh issues can be due to the mesh being shattered by a single device failure.

And now I have to add: the earth is round. :smiley:

7 Likes

:rofl: :rofl: :rofl:

3 Likes

:joy: :joy: :joy:

Thoroughly enjoyed this!

2 Likes

Very true... I had my mesh go nutz when a mains powered device failed in a weird but not completely way.. Battery powered devices don't usually have the same destructive capability when they fail..

1 Like

Good story. I always have to remind myself this is why the OSI model existsโ€ฆ

Rats! I have my pool filter and solar heater panels on the C7/WebCore and NOW you are making me go and set up a valve so I don't have to top my pool manually! :crazy_face:

1 Like

That is the ultimate HA goal to remove all the manual processes :wink:

1 Like

True but there is no water source so I would have to trench and bury. Another summer fun project. :wink:

I had to do that for my Pool Shower project. 240vac at 40 amps for the heater. Water was closer but at right angles, so two trenches.

https://www.tractorsupply.com/tsc/product/atmor-65-kw-electric-tankless-water-heater-at900-06-1462472

And I responded without looking at who the commenter was. I should know better. :embarrassed:

What? No embarrassed emoji?

2 Likes

I guess that would count as "physical layer" :slight_smile:

1 Like

It's me that should be embarrassed by this story. I think at the time, I was, but it's been two days now.. I'm over it :smiley:

It's not like I have no experience with these Hubs. I've had one since they first came to market.. not early enough to get a Signed C-3 but darn close. :smiley:

I guess that was motivation also.. to overcome the embarrassment to assist our new community members, advising that each installation is unique and to just carefully work through the problem, not the rumors.

It wasn't horrible or wrong to start at the hub end, just took longer to find the astonishingly simple answer. :smiley:

4 Likes

yup. definitely layer 1.

the number of times I'd gone down the rabbit hole of "network drivers", servers, etc only to find a cabling issue...

OSI is your friend. =)

3 Likes

Back when I worked in networking, had a device quit. Tried new in package ethernet patch cable. Device still down. Tried second new cable, and voila! both original and first replacement cables were bad. Only took most of the day troubleshooting.

2 Likes

Worked for 3com back in the day. Had a customer call in one time saying they got a box of cards, and none of them would get IPs. turns out, they all got programmed with identical MAC addresses through some mishap. Obviously swapped them out, but that day was the day I realized that indeed, MAC addresses should not be treated as globally unique. Obviously they should be LOCALLY unique, and 3com F'd up, but...

The cabling one also reminds me of what we used to ask customers to do without (directly) implying they're stupid. Rather than saying "Are you SURE the cable is plugged in fully on both ends?" we'd ask them to unplug both ends, flip the cable over, and plug it back in. Trying to listen for the "clicks". But also, usually got them to fix partially seated cables. =)

2 Likes

Hah - same thing happened to a couple of HP printer cards with same MAC address. That one took quite awhile to figure out.

1 Like

Sold a laptop to a client. She called about 3 weeks later and said she couldnโ€™t get on the Internet. After an hour of uninstalling TCP and going as far as wiping the Windows install, a very very basic thing was assumed.

She forgot to pay her Internet bill.