New Z Wave firmware improvements

Vera Controllers, Services & Devices Vera Controllers & Cameras
#1

Please correct me if I am wrong here:

I was under the impression that the latest Z-Wave firmware upgrade was meant to correct a problem where the routing tables were not being updated properly when a device was moved?

Example:
I have a 3-in-1 that was working intermittently - located in an extension of my “mesh” about 150ft from the Vera in a different building. There are other non-battery devices in between, working reliably and within 50ft of each other.

Last night I excluded the 3-in-1 and then included it to solve the Variables problem, and today I ran the ADD program to test the network. The output shows the remote 3-in-1 trying to talk to the devices closest to the Vera (I brought the 3-in-1 to the Vera rather than vice versa) and not to the hard-wired devices within a few feet of itself.

I know a solution is to dongle the Vera and take it to the 3-in-1 located where it should be, but shouldn’t the new Z-Wave firmware and a Heal fix the network?

Advice please?

#2

[quote=“FlyBoyBob, post:1, topic:168562”]Please correct me if I am wrong here:

I was under the impression that the latest Z-Wave firmware upgrade was meant to correct a problem where the routing tables were not being updated properly when a device was moved?

Example:
I have a 3-in-1 that was working intermittently - located in an extension of my “mesh” about 150ft from the Vera in a different building. There are other non-battery devices in between, working reliably and within 50ft of each other.

Last night I excluded the 3-in-1 and then included it to solve the Variables problem, and today I ran the ADD program to test the network. The output shows the remote 3-in-1 trying to talk to the devices closest to the Vera (I brought the 3-in-1 to the Vera rather than vice versa) and not to the hard-wired devices within a few feet of itself.

I know a solution is to dongle the Vera and take it to the 3-in-1 located where it should be, but shouldn’t the new Z-Wave firmware and a Heal fix the network?

Advice please?[/quote]

@Flyboybob

Yes routing should have improved. There are many different settings on the Z-wave routing with many different effects. i have posted mine (no issues so far) attached.

Can you post yours so we can try (with help from others with issues and/or more experienced users) to optimize routing settings so we can all benefit.

Henk

#3

[quote=“Henk, post:2, topic:168562”]Can you post yours so we can try (with help from others with issues and/or more experienced users) to optimize routing settings so we can all benefit.
Henk[/quote]

Sorry to be stupid Henk but what exactly to you mean by “post yours”? Where do I get my “routing” from (seen it in the ADD report and something in the Heal report) and where do I post it? And where is your’s posted?

Right now none of the 3-in-1’s have any entries in the ADD routing table. All are showing “cannot get any information on” in the dashboard after I foolishly did a Heal a few hours ago. The one I did an Exclude/Include last night on dropped the red flag after a while by itself, the others I will have to go around and push buttons and pull batteries on. Again!

The more I try to fix it the worser it gets!!

Frustrated,
bob

UPDATE: My Z Wave [Options] tab is identical to yours as posted
ALSO: When I do a Heal it says it may take hours, and the “Wait for battery operated devices to be discovered” box is 60 minutes, but the Heal finishes in about 10 minutes, and most battery devices disappear from the routing table. That sucks, quite honestly…

#4

[quote=“FlyBoyBob, post:3, topic:168562”][quote=“Henk, post:2, topic:168562”]Can you post yours so we can try (with help from others with issues and/or more experienced users) to optimize routing settings so we can all benefit.
Henk[/quote]

Sorry to be stupid Henk but what exactly to you mean by “post yours”? Where do I get my “routing” from (seen it in the ADD report and something in the Heal report) and where do I post it? And where is your’s posted?[/quote]

Bob,

not stupid… i was referring to your Z-wave device settings. I added a screenshot of mine to my earlier post.

Right now none of the 3-in-1's have any entries in the ADD routing table. All are showing "cannot get any information on" in the dashboard after I foolishly did a Heal a few hours ago. The one I did an Exclude/Include last night on dropped the red flag after a while by itself, the others I will have to go around and push buttons and pull batteries on. Again!

The more I try to fix it the worser it gets!!

Frustrated,
bob

UPDATE: My Z Wave [Options] tab is identical to yours as posted
ALSO: When I do a Heal it says it may take hours, and the “Wait for battery operated devices to be discovered” box is 60 minutes, but the Heal finishes in about 10 minutes, and most battery devices disappear from the routing table. That sucks, quite honestly…

Ok good to hear you were able to confirm your settings. We would have to do some troubleshooting to find out your optimal settings.

What you have now is “default”

#5

Hi Henk. I compared my settings to yours and found some differences that I would like to understand. At the top mine indicates version 2.78 L:1
and below the 2nd and 4th boxes are unchecked.

I have never changed any of these settings since I installed it late last year. I also have issues with battery devices.

Thanks,
Ron

#6

–edit–
More info linked:
http://forum.mios.com/index.php?topic=6898.0

[quote=“Minnies, post:5, topic:168562”]Hi Henk. I compared my settings to yours and found some differences that I would like to understand. At the top mine indicates version 2.78 L:1
and below the 2nd and 4th boxes are unchecked.

I have never changed any of these settings since I installed it late last year. I also have issues with battery devices.

Thanks,
Ron[/quote]

Good! thats the way it should be. Some of use have not only upgraded to the latest firmware (1.1.1245 in which a bug was recently detected by the way and MCV is working on getting an update out there) but also did an update on the Z-wave chip.

2.78L is working ok for most ppl at the moment, so i would say, if it isnt broken, dont fix it right now.
More info is to be found on the wiki, here:
http://wiki.micasaverde.com/index.php/Migrate_To_Z-wave_version_452 (452 = 3.20)

Your ticked boxes are corresponding to your 2.78L Z-wave version, so leave them… especially if it works…

#7

So we’re running 1.1.1245, 2.78. No choice, Vera1 + dongle.

Before we upgraded, we were seeing the Z-wave network wedge itself about every five days, typically at dusk when a lot of lights get turned on. Power cycle always got us functional again, and we may yet add a cron job to this box to reboot it nightly, if this goes on.

Here’s my question: in the Z-wave options, “MIOS routing” is unchecked. I believe the word around here is that the MCV folks found a back-door way to clear and rewrite Z-wave route data to the devices’ Z-wave chips, even using the dongle. Is this so, and if so, do I need to have this “MIOS routing” option checked to get away from the dead-end routing problems due to initial routing data never being purged from devices?

And a side question: with a V1, if I check this option, do I suddenly start getting the nightly heals? Is there some way I can make these devices smart about actual routing without Vera coming along and braining the remote locks and battery-powered sensors every night?

Would really like to hear MCV chime in on this with a definitive answer.

–Richard

#8

Richard,

1.1.1245 was released with improved routing in mind.
The problem is that not all features are documented very well (yet)

I have run 1.1.1245 icm 3.20 for a while with all boxes ticked with no problems and no 2am automatic heals.
Other users have reported automatic heals at 2am, and reported it stopped when unticking the particular box you mentioned.

There is a method to combine the both by unchecking automatic node configuration, which will keep your routing matrix intact if heals would occur.

more on this subject here (which you probably read already)
http://forum.micasaverde.com/index.php?topic=6898.0

[quote=“rlmalisz, post:7, topic:168562”]So we’re running 1.1.1245, 2.78. No choice, Vera1 + dongle.

Before we upgraded, we were seeing the Z-wave network wedge itself about every five days, typically at dusk when a lot of lights get turned on. Power cycle always got us functional again, and we may yet add a cron job to this box to reboot it nightly, if this goes on.

Here’s my question: in the Z-wave options, “MIOS routing” is unchecked. I believe the word around here is that the MCV folks found a back-door way to clear and rewrite Z-wave route data to the devices’ Z-wave chips, even using the dongle. Is this so, and if so, do I need to have this “MIOS routing” option checked to get away from the dead-end routing problems due to initial routing data never being purged from devices?

And a side question: with a V1, if I check this option, do I suddenly start getting the nightly heals? Is there some way I can make these devices smart about actual routing without Vera coming along and braining the remote locks and battery-powered sensors every night?

Would really like to hear MCV chime in on this with a definitive answer.

–Richard[/quote]

#9

UPDATE:
My Vera2 is still running the latest upgrades and still having the same problems. However MCV Support has been trying different things remotely, without success, and I have been assured that my problems have been escalated to “2nd level support” so I am optimistic these problems can be resolved and am anxious to help wherever I can. Here is what we have learned so far:

  1. Support initiated a remote Heal with all options turned on, on my network where all but 1 device were active on the network. After the Heal all 3-in-1’s were off the network and their Neighbor fields were blank.

  2. A nightly automated Heal recovered all devices but the one, which is very remote from the Vera but well within other hard wired devices. It seems these Heals only route devices the Vera can contact directly(?)

  3. MCV Support says “…you should know that the nightly heal runs only if during the day were detected route issues…” Excusing the slightly inverted grammar, this makes sense as to why some are seeing these automated heals are others are not. In plain English the algorithm appears to be “if it ain’t broke don’t fix it”. Unfortunately they’re still working on “if it is broke, fix it”.

  4. Unchecking “Only update MiOS routing” stopped the nightly heal, as Henk suggested.

My next step is to try to exclude/include devices in situ, hoping they will detect their real neighbors and use them, also fixing the corrupted device data with the exclude. I will also leave the auto nightly heals unchecked to remove that variable from the fix.

Wish me luck, hope this helps someone,
bob

#10

@Flyboybob

This is very important information, thanks for letting us know.
Every piece of documentation on MCV specifick routing leads to a dead end on both the wiki and the docs pages, so everything we can learn here is vital.

If you dont mind, i will “borrow” some of your lines to put in the intitial setup/routing explained thread.

As for your problem with the far node, you think manual routing would be a solution?
I gathered what info i could find on that here:
http://forum.mios.com/index.php?topic=6898.msg44194#msg44194

Best,

Henk

[quote=“FlyBoyBob, post:9, topic:168562”]UPDATE:
My Vera2 is still running the latest upgrades and still having the same problems. However MCV Support has been trying different things remotely, without success, and I have been assured that my problems have been escalated to “2nd level support” so I am optimistic these problems can be resolved and am anxious to help wherever I can. Here is what we have learned so far:

  1. Support initiated a remote Heal with all options turned on, on my network where all but 1 device were active on the network. After the Heal all 3-in-1’s were off the network and their Neighbor fields were blank.

  2. A nightly automated Heal recovered all devices but the one, which is very remote from the Vera but well within other hard wired devices. It seems these Heals only route devices the Vera can contact directly(?)

  3. MCV Support says “…you should know that the nightly heal runs only if during the day were detected route issues…” Excusing the slightly inverted grammar, this makes sense as to why some are seeing these automated heals are others are not. In plain English the algorithm appears to be “if it ain’t broke don’t fix it”. Unfortunately they’re still working on “if it is broke, fix it”.

  4. Unchecking “Only update MiOS routing” stopped the nightly heal, as Henk suggested.

My next step is to try to exclude/include devices in situ, hoping they will detect their real neighbors and use them, also fixing the corrupted device data with the exclude. I will also leave the auto nightly heals unchecked to remove that variable from the fix.

Wish me luck, hope this helps someone,
bob[/quote]

#11

Manual routing sounded promising Henk, but…

Here’s what I did:

  1. Battery-dongled my Vera 2 and took it to the 3-in-1 in the garage.
  2. Exclude / Include the device in place with all valid Neighbors.
  3. Re-start the Vera in my house, old device is gone, new device appears.
  4. Used Configure Node Now in Dashboard, Vera gets info and creates 3 more devices (temperature, light, motion), so obviously can talk to the device at the working distance.
  5. Checked the main device Variables = no Neighbors, no routing info, manufacturer is “Ryhert Industries”?? instead of “Express 3-in-1”
  6. Did a Heal = all 3-in-1’s gone, as usual.
  7. Set checkbox for “Let MiOS configure…” whatever to force a nightly heal. No report in the morning, some 3-in-1’s back online, and the checkbox has dissappeared from the Z-Wave Repair tab(??). MCV must be working overtime… :smiley:

This morning decided to try your Manual Route (note that AutoRoute variable not there) and entered “hop” nodes for best routing as suggested.

Tried “Configure Node Now” and get “Node not Configured” every time (very helpful, …not)

Tried “refresh local data” icon on top of dashboard and now all but one 3-in-1 have eternal “Waiting for wakeup to configure” messages attached.

(will update this post after longest wakeup time (1/2 hour) has passed)
UPDATE: more than the 1/2 hours has passed - all waiting-for-wakeup devices still waiting…

bob

#12

@flyboybob

Sounds as if Vera has a broken database entry for this specifick device.

Are you able to exclude it through Vera, bring the sensor back to factory defaults (i think the manual helps there)

Delete the device from Vera (if needed) and then go through it one more time?

Let me know when you loose Hope :wink: :wink:

Henk

#13

Seems to load the wrong set of schemas too

Did the full Exclude/Include dance several times, as far as Vera knows it’s a new device each time. Done it in place, “full power” assuming the Vera can’t tell the difference between the 12v from the battery and the 12v from the wall? No network plugged in of course, if that matters?

Every day and twice on Sunday Henk, makes no never-mind.

Done that too - a LONG time ago! :-\

bob

#14

Did the full Exclude/Include dance several times, as far as Vera knows it’s a new device each time. Done it in place, “full power” assuming the Vera can’t tell the difference between the 12v from the battery and the 12v from the wall? No network plugged in of course, if that matters?[/quote]

Well what i was suggesting was the full power inclusion with also Vera in its original position and see how that works. Also you would be able to see the device appear and not have to reinstall Vera with probably a powercycle needed.

Sent you a PM on that!

#15

That’s the way I did it many many times before. Bringing the battery-dongled Vera to the device “in situ” was something new that I tried. In all situations when the 3-in-1 started reporting values they were wrong, and after a Heal the device(s) were no longer in contact with the Vera no matter what I try.

That goes for a 3-in-1 sitting within about 6ft of the Vera as well. Looking at that one right now - it is showing a “0” light level in a bright room and a temperature of “75” which has not changed since the last time I Excluded / Included it. In fact here’s an ADD report on the near one:

Here’s one I pulled the battery on to re-configure last week, shortly after the upgrade:

bob

#16

@FlyBoyBob,

Mmm, odd. Are you saying the HSM100 that’s close to Vera isn’t working properly either. If so, I’d focus on that and try to get that sensor going first. Exclude, then include. Let it configure. Then see if it starts reporting temperature and light level every 30 minutes. Check the ‘LastWakeup’ field.

This hopefully gets around anything related to routing as the sensor is in direct reach of Vera. If a (nightly) heal then breaks this sensor, it would give MCV something to look at.

#17

@flyboybob

Would this thread help maybe?
Let us know as i think there might be some usefull tips in there!

–edit–
correct link added
http://forum.mios.com/index.php?topic=4450

That’s the way I did it many many times before. Bringing the battery-dongled Vera to the device “in situ” was something new that I tried. In all situations when the 3-in-1 started reporting values they were wrong, and after a Heal the device(s) were no longer in contact with the Vera no matter what I try.

That goes for a 3-in-1 sitting within about 6ft of the Vera as well. Looking at that one right now - it is showing a “0” light level in a bright room and a temperature of “75” which has not changed since the last time I Excluded / Included it. In fact here’s an ADD report on the near one:

Here’s one I pulled the battery on to re-configure last week, shortly after the upgrade:

bob[/quote]

#18

[quote=“oTi@, post:16, topic:168562”]@FlyBoyBob,

Mmm, odd. Are you saying the HSM100 that’s close to Vera isn’t working properly either. If so, I’d focus on that and try to get that sensor going first. Exclude, then include. Let it configure. Then see if it starts reporting temperature and light level every 30 minutes. Check the ‘LastWakeup’ field.

This hopefully gets around anything related to routing as the sensor is in direct reach of Vera. If a (nightly) heal then breaks this sensor, it would give MCV something to look at.[/quote]
@oTi -

I did that last evening with the nearest HSM100 = Exclude, Include, Configure. Today that one sensor is reporting seemingly valid data, and its LastWakeup field shows well within the 1/2 hour set wakeup period.

Other nearby battery operated devices are showing invalid or stale data, the ADD program shows battery updates in the dozens or hundreds of hours ago, yet the LastWakup field shows valid recent wakeup times within their wakeup period, so the devices are apparently active and talking to Vera.

As I said earlier the Variables and Data seeem to have been corrupted on all devices after the firmware update. I will try to Exclude / Include more devices tonight, that seems to read the proper schemas and default values and configures them into the device.

There were 2 strange Heal reports last evening after I had gone to bed - I have given access to my box to MCV so I assume they are experimenting. The remote 3-in-1 (which was Excluded / Included by myself several times recently) still has no Neighbors and no mesh route to get to it, so is still not working at all. I had also put a ManualRoute Variable and value for a path through hard-wired devices to it = no difference.

bob

#19

Good!

[...] battery updates in the dozens or hundreds of hours ago [...] the Variables and Data seeem to have been corrupted [...]
Right. At least for some devices, battery reporting is autonomous, on its own timer, and has its own association group. So if not configured properly, I guess you could have no battery reports, but other stuff working fine.
I will try to Exclude / Include more devices tonight, that seems to read the proper schemas and default values and configures them into the device.
That would be my suggestion as well. Try to re-include the 'funny' devices. And keep them close to Vera for now, if able, just to rule out routing issues.
#20

Good![/quote]
Good up until this morning when for no apparent reason the LastWakeup is now 3 hours ago, and increasing. Fresh batteries btw. All HSM100’s last for a while and then drop off, regardless of recent exclude / include or distance from Vera. Is there a “self-destruct” variable I’ve missed?? ::slight_smile:

[quote=“oTi@, post:19, topic:168562”]

[…] battery updates in the dozens or hundreds of hours ago […] the Variables and Data seeem to have been corrupted […]

Right. At least for some devices, battery reporting is autonomous, on its own timer, and has its own association group. So if not configured properly, I guess you could have no battery reports, but other stuff working fine.[/quote]
Of the HSM100’s that were configured pre-upgrade vs recently exclude/include devices (all of those devices are identical btw) they all seem to break in an identical way, except for the remote one which is always broken. I have not made any configuration changes and they are all at factory reset values. However the Variable lists vary hugely on two recently re-included devices(?) In fairness - I believe MCV is working on my machine remotely even though they have not said - I had opend my machine to Support last week and was told it would be “escalated”. So these unexplainable inconsistencies might be them, but my whole system remains broken.

[quote=“oTi@, post:19, topic:168562”]

I will try to Exclude / Include more devices tonight, that seems to read the proper schemas and default values and configures them into the device.

That would be my suggestion as well. Try to re-include the ‘funny’ devices. And keep them close to Vera for now, if able, just to rule out routing issues.[/quote]
As I said above all pre-upgrade HSM100’s and all re-configured ones are failing. The one remote one used to work pre-upgrade, now not at all. BTW two battery-operated Temp / Humidity sensors also fail in the same way, while my Schlage lock (also battery) seems ok (thank goodness!)

bob