I've joined the "locked up" club

Community Playground General Discussions
1

Last night Vera stopped functioning and the dongle light went to solid blue. I unplugged and replugged the power cord and she started blinking again. Seems like an incredibly inelegant solution. The scenes didn’t activate, but maybe I need to tell them to “Go” again. Argh.

As a newcomer, I avoided the dongle recall by having the foresight to be shipped a Vera with a .78 dongle. :wink: I didn’t feel too left out of all the fun though, because I have had my fair share of challenges every step of the way as I’ve set up my locks, lights, and garden fountain controls. I’ve been messing around with my Trendnet cameras, but they haven’t worked since 616. Everything seems to take three or ten tries before it works, and there are plenty of opportunities to start over from scratch. As an avowed “early adopter” I can handle this, but Vera is really just something to play around with at this point. I think I’d say the same about HA in general. I run a software company, but I’m no programmer. I’m glad some of you guys are.

2

Welcome to the club!

A little hint: It might be that the solid light stays solid even when you unplug the dongle. Then you can use a paperclip to push the tiny reset button in the hole on the dongle.

I share with you the not-beeing-a-programmer-problem too. :-\

3

We’re trying to get to the bottom of this. It’s not an issue with our software or Vera–it’s the Z-Wave module in the dongle locking up. We’ve had a few users report this now. Unfortunately it’s very tough because (a) the dongle’s microprocessor has no logging are any way of debugging it after the fact. You have to be able to reproduce it on demand, and nobody can figure out how to do it, and (b) it’s not in our code–it’s in the Z-Wave module. So we can’t debug it or fix it, we can only report it to the folks at Sigma, and for them to do anything, they always need us to find a way to reproduce it on demand. It seems to happen to only a few customers so far, but the ones it does happen to, it seems to happen fairly often. We’re trying to find a pattern.

4

I’m not an electrical expert by any means, but I do recall a similar (somewhat infamous) “lockup” scenario with (of all things) the old CM11A (X10) interface, which would sometimes run hot and/or hang (stop responding).

The cause turned out to be a chip-level logic gate going ‘hi’ and remaining high due to INACTIVITY. The lockup/hang/freeze situation could be avoided simply by having the user initiate any command on the CM11A whatsoever, or cycle its power.

So, I have to ask, somewhat paradoxically … Do folks whose dongle has gone ‘solid blue’ recall NOT using Vera for an extended period? Did the dongle ever get hot?

[I mention this anecdote also for the benefit of MCV, now tasked with reproducing the problem – that it might pay to let a test Vera just sit idle for several hours, as a possible ‘reproduction’ state, perhaps even checking the dongle’s temperature throughout.]

5

LibraSun, how do you define an “extended period” of inactivity?

As I mentioned in other threads I have 25 devices plus controllers. They are managed by 20 scenes. I guess the longest period of not sending a command other than “poll device” is less than four hours.

@MCV: (For me) it’s good to hear that I’m not alone with the problem and you recognized it!

6

@huwu: 12 hours was about the right length of time in the case of the infamous CM11A Lockup issue, but I’m not suggesting a similar idle time would need to pass in order for Vera’s dongle to be affected … my conjecture is more focused on the possibility of a “hot emitter” within the dongle’s circuitry, or similar.

DISCLAIMER: My theory is likely invalid, considering my lack of electronics expertise!

7

I love that line…As an electrical engineer I would almost figure this is a firmware problem not with Vera but the dongle itself. Since the MCU used with ZWave radio is an 8051 I would almost guarantee that they are over running the stack and corrupting the program counter.

8

I’ve been talking to the folks at Zensys and Aeon (who makes the dongle) about this. Unfortunately this is one of the REALLY frustrating time wasting bugs because it’s random, intermittent, and we can’t reproduce it in our office. I’ve never seen one of our dongles lockup like this, and it seems to only affect 10% or so of the users.

So there is something about certain homes that is causing the dongle to lock up. Either (a) a defective dongle, (b) some combination of Z-Wave devices that’s causing the Z-Wave module to go haywire, (c) environmental factors like ambient temperature or electrical interference, or (d) ???

In the case of the Z-Wave dongle problems we suffered until the new dongle recall, the problem was (b). There was a bug in the Z-Wave 5.02 serial api that caused some sort of stack overflow particularly when there was a mixture of new 44 kbps Z-Wave modules with the old 9 kbps modules (ie Intermatic).

Since we can’t reproduce this, and neither Zensys nor Aeon has heard of this, what we need is the help of a user that has seen his dongle lockup frequently and regularly. If we can have a volunteer, I’d like someone to check ‘verbose logging’ and ‘lock log levels’ in Advanced Logs. Then everytime there’s a dongle lockup, submit a trouble ticket, go to Advanced, Tech Support, click ‘enable it’ to turn on the back door, and send me an email to aaron [at] micasaverde and put DONGLE LOCKUP in the subject line so it gets my attention, and has the back door code. Don’t reset the dongle until I’ve had a chance to log in and see what was the last thing the dongle did so I can find a pattern, and keep a log of the lockups, the frequency, time of day, etc.

Once we confirm you have a dongle that locks up regularly, I’d like to send you (a) another dongle with the identical firmware, (b) another Aeon dongle that does not have the Aeon battery and inclusion button active, and (c) a stock Zen-sys dev board running the same version of the Z-Wave module. We can use the clone dongle function to make all 3 new dongles exact mirrors of the existing dongle. And then we should see which of those 3 new dongles still locks up. That will isolate if it’s just a defective unit, a defect in the Aeon design or Aeon battery/inclusion stuff, or a general problem with the Z-Wave firmware itself.

I know this is kind of a nuisance for a user, but it would be much appreciated if we could get a volunteer. This benefits not just Mi Casa, but everyone in the Z-Wave community. The bugs in the 5.02 Z-Wave modules which caused all that instability in the past, and which were fixed in July, were isolated and identified because of our testing like this. So when we were able to demonstrate it to the Zensys engineers so they could fix it, it benefited everybody in the Z-Wave community because that Z-Wave unreliability was plaguing every Z-Wave product on the market that supported security. I have a suspicion that this dongle lockup issue is also a universal one effecting every Z-Wave controller, and probably is not unique to the Aeon dongles.

9

Now I know what’s different about Intermatic! :o

“In the case of the Z-Wave dongle problems we suffered until the new dongle recall, the problem was (b). There was a bug in the Z-Wave 5.02 serial api that caused some sort of stack overflow particularly when there was a mixture of new 44 kbps Z-Wave modules with the old 9 kbps modules (ie Intermatic).”

I’ve been looking for information to differentiate between inexpensive (Intermatic) devices and expensive (Leviton) ones. Is this some info that would be valuable to post to the wiki “Supported Devices” page? Would someone with the information update, say, the “Notes” field?

10

[quote=“CMRancho, post:9, topic:164863”]Now I know what’s different about Intermatic! :o

“In the case of the Z-Wave dongle problems we suffered until the new dongle recall, the problem was (b). There was a bug in the Z-Wave 5.02 serial api that caused some sort of stack overflow particularly when there was a mixture of new 44 kbps Z-Wave modules with the old 9 kbps modules (ie Intermatic).”

I’ve been looking for information to differentiate between inexpensive (Intermatic) devices and expensive (Leviton) ones. Is this some info that would be valuable to post to the wiki “Supported Devices” page? Would someone with the information update, say, the “Notes” field?[/quote]

Hmm, interesting. Where this info comes from?

11

[quote=“micasaverde, post:8, topic:164863”]If we can have a volunteer, I’d like someone to check ‘verbose logging’ and ‘lock log levels’ in Advanced Logs. Then everytime there’s a dongle lockup, submit a trouble ticket, go to Advanced, Tech Support, click ‘enable it’ to turn on the back door, and send me an email to aaron [at] micasaverde and put DONGLE LOCKUP in the subject line so it gets my attention, and has the back door code. Don’t reset the dongle until I’ve had a chance to log in and see what was the last thing the dongle did so I can find a pattern, and keep a log of the lockups, the frequency, time of day, etc.

Once we confirm you have a dongle that locks up regularly, I’d like to send you (a) another dongle with the identical firmware, (b) another Aeon dongle that does not have the Aeon battery and inclusion button active, and (c) a stock Zen-sys dev board running the same version of the Z-Wave module. We can use the clone dongle function to make all 3 new dongles exact mirrors of the existing dongle. And then we should see which of those 3 new dongles still locks up. That will isolate if it’s just a defective unit, a defect in the Aeon design or Aeon battery/inclusion stuff, or a general problem with the Z-Wave firmware itself.[/quote]

I checked these two and will inform you as requested if it happens at your office hours. The special dongle program might not be working for me as an overseas customer.

btw it’s the first time that dongle runs more than 28 hours without lockup.

12

no wonder why my dongle locked up all of the time… i have a house full of leviton dimmers and 2 cheapo intermatic outlets. these outlets are the ones that control my feeders which have about 12 timer events a day… thats a lot of events to intiate a possible lockup in such a system

13

I have a mixture of Leviton/Intermatic/GE and performance is definately suffering.

14

Some of us have no clue about what the differences are among all the devices out there, but with the huge price differences there must be something. Does anyone know why a Cooper Aspire device costs three or four times as much as an Intermatic? Or what the difference is between Wayne Dalton and Leviton? I’d sure like to know before I spend another boatload of money on fitting out the rest of my house!

15

I beg to differ with MCV on the subject of Lock-up
Lock for me is dependant on the firmware.
With the luup version my system locks almost every 12 hours. ( re-boots)
I used to get logs in the 24,000 jobs. I am lucky to get a 1,000 jobs before a reboot.

Maybe MCV is asking tooo much from this Router.

It would be nice to have a simple product that works like what we thought Vera would be.
KISS KEEP it SIMPLE STUxxx

16

I’d run vera 24/7 on an intel atom if it meant stable/faster performance

17

Dammit. Put me down as a “member”. Vera just locked up on me. FACK!

18

We’ve identified another memory leak in the Luup beta which can cause Vera to run out of memory over long periods, especially with lots of nodes. We’re working on fixing that. This is an issue in our code. Vera has a ‘watchdog’ though, and should reset itself automatically when this happens, so you generally won’t notice it, other than that the job numbers stay small.

Separately, some users have a problem with the dongle itself locking up (ie light stops blinking). This got escalated to our topmost priority because when it happens we Vera can’t recover–the dongle itself needs to be reset. It’s also harder to solve because it’s not in our code and we don’t have any logs–it’s a problem in the Z-Wave module inside the dongle.

19

I’m pretty sure that this will change after I post this… :wink:

…but the dongle and Vera were running the last ten days without locking and without rebooting. Every scene worked, the whole system is reliable as it should be. That’s how I like it to be.

I’m still on 616…

20

I too am on 616… And I don’t experience lock-ups. My dongle is a version 0.64,… All my components are Intermatic/Wayne Dalton. I AM going to be getting the Schlage locks, and maybe, down the road, the Wayne Dalton Z-wave garage door openers (conversion of existing ones is just not feature-rich enough).

Am I to understand that this issue is primarily related to the mixing of Intermatic (or the re-branded Wayne Dalton/Intermatic units) and Levitron components?!?!? Is the new version of the Vera firmware able to re-flash the dongle with it own updated firmware???