ID 4096: Reconfiguration process failed

[tinman 1,070]

while full reconfiguration of a device i got "ID 4096: Reconfiguration process failed" message, now out of curiosity, is 4096 the current devices IDs limit? 

[petergebruers 1,268]

The "4096" is a temporary ID assigned during reconfiguration. No device should have that ID during normal operation... I think you'll see no trace of it after you refresh the web page or query /api/devices. The max ID was not disclosed afaik but the ID field in the database is either 16 or 32 bit.

[tinman 1,070]

this is exactly why i asked, knowing the 0x1000 was thinking what if HCx have reached that number

[petergebruers 1,268]

As far as I remember, Fibaro told me that this limit is very hard to reach, due to the limited number of possible devices (232 IDs), multiplied by the theoretical number of slave IDs per device + al the scenes etcetera. So it may or may not be an issue when a user reaches that.  I was at ID 1200 after several years and several upgrades, and that included the 3->4 series of firmware upgrade that created a whole new range of IDs. After I reInstalled my HC2 I'm now below ID 500. But OK, let's assume you keep reconfiguring, adding, deleting, upgrading and you reach ID 4095... What will happen? I think you will get assigned a special support case, get a free lunch and the development team changes the temporary ID from 4096 to 8192 in the next firmware.  

 

Seriously... I don't want to ridicule the question. I know you from this forum and you always want to dig deeper and know more. Me too, so I asked this question a long time ago. I may have the details somewhere, but they may no longer apply to more recent firmware releases. But I honestly think that it *may* fail at ID 4096, mr Socha neither confirmed nor denied this, only the fact that 4096 is a temporary ID and that reaching 4096 as an ID on a production system sounds... unlikely.

[tinman 1,070]

oh well, 4101 and still ok ...

 

Please login or register to see this image.

/monthly_2016_10/over4096.JPG.3cb57eb69b81d3ff97e50ec1a2d3065b.JPG" alt="over4096.JPG" />

 

however it would be nice to hear something from Mr. Socha what is the real max number.

[A.Socha 133]

Hi

don't worry for DeviceID  you still got a lot of free 4 - 4'294'967'295

You can have problem with NodeID (only z-wave device) 2-233 (only 1 per device) but that you can resolve with slave Gataway, controller going in loop and take first free NodeID in adding device so when you got for example 2,3,5,6 then when he will go to 233 he will add next device with id4 every removing device release NodeID

 

[petergebruers 1,268]

On 24-10-2016 at 1:13 PM, A.Socha said:

Hi

don't worry for DeviceID  you still got a lot of free 4 - 4'294'967'295

You can have problem with NodeID (only z-wave device) 2-233 (only 1 per device) but that you can resolve with slave Gataway, controller going in loop and take first free NodeID in adding device so when you got for example 2,3,5,6 then when he will go to 233 he will add next device with id4 every removing device release NodeID

 

 

Thank you for explaining this in detail! Understood! But this is tinman's topic, so it's not up to me to say if this answers his question(s).

Anyway, seems to me that a user won't hit the DeviceID limit, but possibly (let's say > 120 physical devices and as many full exclude/include cycles) he could hit a NodeID exhaustion. Unless he has a gateway.

[A.Socha 133]

12 minut temu, petergebruers napisał:

Anyway, seems to me that a user won't hit the DeviceID limit, but possibly (let's say > 120 physical devices and as many full exclude/include cycles) he could hit a NodeID exhaustion. Unless he has a gateway.

No, it will not; If you got 120 physical devices and you will remove and include them again then you will still have 120+1(controller) NodeID only used. (Not talking about situation when you will leave in HC orphans from manual reset devices and you will add them again) If you gonna have 233 physical devices then you will exhaust NodeID (on master gateway) and you gonna need slave gateway

[petergebruers 1,268]

13 minutes ago, A.Socha said:

No, it will not; If you got 120 physical devices and you will remove and include them again then you will still have 120+1(controller) NodeID only used. (Not talking about situation when you will leave in HC orphans from manual reset devices and you will add them again) If you gonna have 233 physical devices then you will exhaust NodeID (on master gateway) and you gonna need slave gateway

 

Ah! I see! I thought it worked that way. The Z-Wave network assigns incremental NodeID, isn't it? But as you remove a device in the middle of the range, and do it properly, that node ID will become reused if necessary, but not right away. I don't own a gateway... but I've always thought that it joins two complete separate Z-Wave networks through a software solution. Is that correct?

[A.Socha 133]

Cytuj

The Z-Wave network assigns incremental NodeID, isn't it?

Yes

Cytuj

But as you remove a device in the middle of the range, and do it properly, that node ID will become reused if necessary

Yes, when controller go to NodeID 232 he will start from begin using "released" NodeID

Cytuj

I don't own a gateway... but I've always thought that it joins two complete separate Z-Wave networks through a software solution. Is that correct?

Yes that is software solution, then you got two separate  independent z-wave networks . You will get  2x 232 NodeID but device from network 1 can't mesh by network 2 or be routed or controlled (associations) so the best solution is to have 1 HC per floor(first floor, second floor etc) or part of house (east wing).

[petergebruers 1,268]

2 minutes ago, A.Socha said:

Yes

Yes, when controller go to NodeID 232 he will start from begin using "released" NodeID

Yes that is software solution, then you got two separate  independent z-wave networks . You will get  2x 232 NodeID but device from network 1 can't mesh by network 2 or be routed or controlled (associations) so the best solution is to have 1 HC per floor(first floor, second floor etc) or part of house (east wing).

 

Got it! Thanks again! And sorry for the thread hijack, mr tinman!

[tinman 1,070]

oh well, let' hijack even more

 

Please login or register to see this link.

,

when talking about Fibaro gateways max. values, is there any limitation for number of slave gateways? (except of course resources problem, but that's different story) 

 

[A.Socha 133]

 

19 godzin temu, tinman napisał:

oh well, let' hijack even more

 

Please login or register to see this link.

,

when talking about Fibaro gateways max. values, is there any limitation for number of slave gateways? (except of course resources problem, but that's different story) 

 

we not added any

 

how much you can have device on HC?

It depend on what PC you got(and from type of device), because some operation we are doing in javascripts so if you got better CPU and more RAM then you can have more device on HC2

[tinman 1,070]

sure, available devices, resources and the setup itself is always making limitation, but good to know that there is "no limit" as such.