Changing a switch from an 8n button to a six button

[LeeG]

The PLM Restore could take hours. If there are other Scene consolidations planned you might do them before the restore.

 

Being able to control a device individually does not mean that Scenes can be managed or executed successfully. Individual device control is a single command that is retried multiple times. Scene management can take dozens to hundreds of commands that must all work (with retries). Often a single command will work but the network is not reliable enough to support dozens of (or more) commands back to back.

 

You can get some sense of how the network is working by looking at the Event Viewer with Level 3-Device communications events selected.

 

Look for Insteon messages that show Max Hops and Hops Left. When the Hops Left count tracks down to 1 or 0 when Max Hops start at 3 the network is working at the extreme end of reliability. What you want to see is Max Hops=3 and Hops Left=2. See Red values below.

 

Tue 02/14/2012 05:38:12 PM : [iNST-ACK ] 02 62 15.52.C8 0F 11 FF 06 LTONRR (FF)

 

Tue 02/14/2012 05:38:12 PM : [iNST-SRX ] 02 50 15.52.C8 19.70.06 2B 11 FF LTONRR (FF)

 

Tue 02/14/2012 05:38:12 PM : [standard-Direct Ack][15.52.C8-->ISY/PLM Group=0] Max Hops=3, Hops Left=2

 

Tue 02/14/2012 05:38:12 PM : [ 15 52 C8 1] ST 255

 

Tue 02/14/2012 05:38:18 PM : [iNST-ACK ] 02 62 15.52.C8 0F 13 00 06 LTOFFRR(00)

 

Tue 02/14/2012 05:38:18 PM : [iNST-SRX ] 02 50 15.52.C8 19.70.06 2B 13 00 LTOFFRR(00)

 

Tue 02/14/2012 05:38:18 PM : [standard-Direct Ack][15.52.C8-->ISY/PLM Group=0] Max Hops=3, Hops Left=2

 

Tue 02/14/2012 05:38:18 PM : [ 15 52 C8 1] ST 0

[Schoolhouse]

so if I understand, you are saying that a switch responding to the ISY may have had 100 tries in getting it successfully to respond. A switch only sends the message once to the ISY?

[Schoolhouse]

You asked I restore Modem (PLM). I will do that tonight.

 

Will it help if turn off query for tonight?

 

Would it also be helpful for me to restore devices?

 

Thanks for your help!

[LeeG]

No. The command from the ISY to the Switch will be retried three times before it fails. This is done by the PLM automatically without ISY intervention. These are Insteon Direct commands. Scene management uses Insteon Direct commands to read and write Insteon link records at the device. With link records being 8 bytes long and it taking at least 17 commands to write one link record you can see that message flow associated with updating/creating a Scene is exponentially higher than a simple device On/Off command.

 

Pressing a paddle/button sends an Insteon Group message sequence which is more complex than simple Insteon Direct messages. Too much to go into here. The Group messages are retried. If Scenes can be reliably read/written to the device but button/paddle presses are not seen at the ISY there is a link record issue. Unlike Insteon Direct command which do not use link records at all, Insteon Group messages require valid link records in both the device and the ISY PLM for the ISY to see the button/paddle press.

[LeeG]

Turn Off the 3 AM Query. I would not Restore Device at this point. It may be necessary with the comm errors that have occurred but let’s see what we have after the PLM restore. The Show Device Links Table function has a Compare button that compares what a device has in terms of link records versus what the ISY thinks should be in the device. If a device does not show state changes after the PLM restore and the number of link records in the PLM is not a concern then a Show Device Links Table/Compare for the device not showing state change is the next step

 

Do not have Insteon traffic while the PLM restore is running.

[Schoolhouse]

Looking at the event viewer, I see a STANDARD CLEANUP switch that is Max Hops =1, Hops left =0

[Schoolhouse]

Okay Thanks, I will restore PLM tonight.

 

Then in the morning, should I open up the Event Viewer and put it on #3 -- Device communications events, and walk around and push each button? I think this will show me the communication of each switch to the ISY, if any at all.

 

Is that the next step?

 

Also, If I am at the upper limit of the capacity of my PLM, would going to a bigger one now be the smart thing to do? Is there a bigger one?

[LeeG]

This is a message from a device to the PLM, part of the Insteon Group protocol. They start with Max Hop=1 for performance reasons. Each subsequent retry increases the Max Hops count. Coming from a device with Max Hops 1 and Hops Left=0 is good communication from the device to the ISY PLM.

 

Tue 02/14/2012 06:13:18 PM : [iNST-SRX ] 02 50 19.21.DA 19.70.06 41 11 01 LTONRR (01)

 

Tue 02/14/2012 06:13:18 PM : [standard-Cleanup][19.21.DA-->ISY/PLM Group=1] Max Hops=1, Hops Left=0

[LeeG]

You cannot buy a bigger PLM. The 2412S supports more link records but is no longer sold. Reduce the number of Scenes.

 

That will show what communication if any is coming from each device.

[Schoolhouse]

I have 64 scenes. Does that mean the PLM thinks I have 928-64, or 864 devices? I assure you I do not have that many. My lighting has 64 items. Say 50 are 8 button items. Does that get me to a little more than 400 devices?

 

If so, Something is wrong at the PLM that thinks I have 928

[LeeG]

Without any Scenes defined each 8 button KeypadLinc has 8 Responder links in the PLM, each 6 button KeypadLinc has 5 responder links in the PLM, so the ISY will be aware of a button press. From memory each KeypadLinc also has a Controller link in the PLM to support the My Lighting Query.

 

Each SwitchLinc has two links in the PLM, one Controller link and one Responder link.

 

The link count associated with an ISY Scene depends in the number of devices in the Scene. There will be a Controller link for every Responder and Controller/Responder in the Scene.

 

That is the minimum. There may be others that do not come to mind immediately.

[Schoolhouse]

So the count after the PLM restore is now 612. That is significantly lower than the 928 before the restore.

 

Unfortunately, the switch status changes occurring are still not registering in the ISY Status

 

I disabled, and enabled the switch, to no avail.

 

I restored switch and regarding communication, Max Hops =3, Hops left =1 all during the restore.

 

No luck there, either.

 

I ran device Links Table and it showed a count of: 70

 

Using Compare, The ISY Links table showed a count of: 69

 

The closeness of the number seems odd because there have been many times when I executed the switch failing to reveal status change

 

What now?

[LeeG]

What did the Compare show for each link record, Identical, mismatch, extra, etc. The count is probably okay. The Show Device Links Table includes the End Of List record which the ISY Links information does not include.

 

Is the switch a SwitchLinc or KeypadLinc?

 

Lets reset. With this switch the load control button/paddle can be pressed On/Off but no state change occurs at the Admin Console. Is the Current State column blank for all devices or do some switches show current state? If some of the switches show a current state do these switches change state as the load control button/paddle is pressed On/Off?

[Schoolhouse]

We moved the PLM to a new spot and the switch in discussion is now working at Max Hops=3, Hops left =2 for the switch we have been discussing.

 

The switch in questions a KeypadLinc and it still does not show changes in Current Status in ISY. It always shows a current state of "Off" which never changes in the ISY Current state window when the actual current state is cycled on and off at that particular switch. This same switch does change current state in the window only when the status is changed in the ISY button on my computer. When we observe the scene this button controls, all the associated devices also show "Off," despite all being "On."

 

When we go to a switch working correctly in the ISY, one that is also a controller in that same scene, all devices show "On," in the ISY, including the switch in question.

 

I do not find an End of List record in the device Links table. In fact, I ran the compare again, hours later, and the count numbers are identical to this morning. 69 and 70. Seem like they should be different! I have been pushing that button a lot since then.

[LeeG]

The Show Device Links Table pulls link records until it finds an End Of List link record so it found one. It is identified by a 00 in the first byte.

 

The new PLM location is communicating with the object switch better. A Hops Left=2 is good.

 

Confirm that some SwitchLincs and/or KeypadLincs do show state changes when their Primary load control button is cycled On/Off.

 

Does the Compare of the KeypadLinc Show Device Links Table indicate the link records are Identical?

[Schoolhouse]

The Show Device Links Table pulls link records until it finds an End Of List link record so it found one. It is identified by a 00 in the first byte.

 

 

ANSWER: I EE 0000 : 00 04 1B.28.CB FF 1F 04

 

 

 

The new PLM location is communicating with the object switch better. A Hops Left=2 is good.

 

GREAT, BUT WE INSTALLED A LAMMPLINC DUALBAND MODULE WICH BEHAVED IDETCAL TO THE SWITCH IN QUESTION. THAT IS TO SAY, THE ISY COULD CONTROL IT, BUT THE ISY DID NOT SEE CHANGES MADE AT THE DEVICE.

 

 

 

Confirm that some SwitchLincs and/or KeypadLincs do show state changes when their Primary load control button is cycled On/Off.

 

ANSWER: HOW DO YOU CONFIRM THIS? I DID IN MY EARLIER POST DISCUSS HOW OTHER CONTROLLERS IN THE SCENE DID SHOW THEIR CURRENT STATUS CHANGE.

 

 

Does the Compare of the KeypadLinc Show Device Links Table indicate the link records are Identical?

 

 

ANSWER: NO THEY ARE DIFFERENT. THE DEVICE COUNTS 70, THE ISY COUNTS 69. IN MY JUST SENT NOTE, I REMARKED HOW THIS SEEMED HIGHLY ODD BECAUSE DEESPITE NUMEROUS CHANGES ORIGINATING AT BOTH THE ISY AND THE DEVICE, MY COUNT DO NOT CHAGE AT ALL. THIS SEEMS WRONG, DOESN'T IT?

[LeeG]

After the Show Device Links Table display completes populating the window with the device link records the Compare button along the bottom becomes active. Click the Compare button. A second window is displayed showing the ISY Links. Look at the original Show Device Links Table window. At the beginning of each link record line there is an annotation which indicates if the particular link record is Identical, mismattch, etc.

 

Click on My Lighting at the top of the list of nodes. This displays all devices in the right pane showing the Current State for every device. The Current State value will change if the ISY sees the messages generated by a button/paddle press.

 

Are there devices that show a state change?

 

The LampLinc test is somewhat problematic as the outlet the LampLinc is plugged into could be a different electrical circuit from the KeypadLinc.

 

With consistent Hops Left=2 when pulling the link database or changing link records in the KeypadLinc the lack of state change at the Admin Console is either a link record issue or something else which I won't go into in detail unless needed. It is VERY IMPORTANT to know if other devices do reflect a state change in the Admin Console. I no devices show a state change something in the router and/or AV is blocking the "push" of state change information to the Admin Console. This is not likely since the Event Viewer would not be working either. This leaves link records as the first choice. The Compare results (not the count) will indicate if the necessary link records are in the device. If they are present and correct then we will have to look at the PLM.

[Schoolhouse]

I ran Compare again and every item shows [identical] to the left.

 

The difference of 1 extra in Device Links table compared to icy table seems to be the last line I sent earlier which says [ignore]

 

Is there more info you desire from these two table?

 

If there is a summary in the device Links table, I do not see it.

[LeeG]

The Compare results indicate the KeypadLinc has all the link records the ISY thinks it should. If the load control button is not changing state in the Admin Console one of three conditions exist

 

PLM is missing link records for this KeypadLinc

KeypadLinc is missing link records despite the ISY thinking they are all there

Comm problems

 

There are a few things yet to be confirmed

 

When the My Lighting entry is clicked to display all devices in the right pane do some devices show state changes and some do not?

 

Approx how many SwitchLincs/KeypadLincs show state changes, how many do not show state changes? I am only interested in the load control button/paddle, regardless of whether there is a load connected to it.

 

I would prefer to analyze a SwitchLinc that is not working as there are far fewer link records involved.

[Schoolhouse]

There are a few things yet to be confirmed

 

When the My Lighting entry is clicked to display all devices in the right pane do some devices show state changes and some do not?

 

ANSWER: ALL DEVICES SHOW CURRENT STATE CHANGE WHEN THE ALL ON OR ALL OFF BUTTONS ARE CYCLED IN MY LIGHTING.

 

Approx how many SwitchLincs/KeypadLincs show state changes, how many do not show state changes? I am only interested in the load control button/paddle, regardless of whether there is a load connected to it.

 

ANSWER: NOT RELEVANT SINCE ALL ARE SHOWING CURRENT STATE CHANGE

 

I would prefer to analyze a SwitchLinc that is not working as there are far fewer link records involved.

 

ANSWER: THAT IS WHAT WE HAVE BEEN FOCUSED UPON. ONE SWITCHLINC 8 BUTTON.

[Schoolhouse]

There are a few things yet to be confirmed

 

When the My Lighting entry is clicked to display all devices in the right pane do some devices show state changes and some do not?

 

ANSWER: ALL DEVICES SHOW CURRENT STATE CHANGE WHEN THE ALL ON OR ALL OFF BUTTONS ARE CYCLED IN MY LIGHTING.

 

Approx how many SwitchLincs/KeypadLincs show state changes, how many do not show state changes? I am only interested in the load control button/paddle, regardless of whether there is a load connected to it.

 

ANSWER: NOT RELEVANT SINCE ALL ARE SHOWING CURRENT STATE CHANGE

 

I would prefer to analyze a SwitchLinc that is not working as there are far fewer link records involved.

 

ANSWER: THAT IS WHAT WE HAVE BEEN FOCUSED UPON. ONE SWITCHLINC 8 BUTTON.

[Brian H]

The Status of My Lighting after sending an All On or All Off. Shows what it thinks the modules should be after the command.

A module can miss a command and still show what My Lighting thinks it should be.

I rarely have that happen myself. Maybe one or two times a year. I have an ApplianceLinc that misses a command and My Lighting shows what it expected it to be.

What you need to do is check My Lighting for errors after locally toggling some of the switches in your setup.

[Schoolhouse]

What you need to do is check My Lighting for errors after locally toggling some of the switches in your setup.

 

ANSWER: WE DID EXACTLY THAT FOR EVERY SINGLE DEVICE IN THE HOUSE. EVERY SINGLE DEVICE IS FULLY FUNCTIONAL BY THE ISY.

 

WE ALSO WENT TO EVERY DEVICE, AND DISCOVERED THAT 6 (ALL OF THEM KEYPADS) OUT OF 70 DEVICES DO NOT MAKE ANY REPORT OF CURRENT STATUS TO ISY, NOTWITHSTANDING THAT ALL OF THESE 6 DEVICES DO IN FACT OPERATE THE SCENE AS EXPECTED

[LeeG]

Unfortunately nothing you initiated from the Admin Console means anything. The ISY knows what command it issued and sets the Current Status column accordingly. Plus the commands used are Insteon Direct commands. Insteon Direct commands do not use link records. As I have stated before.

 

What I needed to know is how many devices show a state change when you push the load control button/paddle on the device itself. This action uses Insteon Group commands which are totally dependent on link records both in the device and in the PLM. It really was relevant.

 

Oh, by the way, a SwitchLinc is a single paddle device which requires one link record in the device and one link record in the PLM. A KeypadLinc is a multi-button device which requires from 5 to 8 link records in the device and the same number in the PLM. It would have been easier to look for link records when only one needs to be found.

 

It will be important to provide the next person who tries to help with accurate information to the questions they ask. You do not have the Insteon knowledge or experience to substitute your knowledge for theirs.

 

Sorry I was not able to find the root cause.

 

Good luck.

[Schoolhouse]

Sounds like our frustration here has upset you. Please accept my apologies and know that I am grateful for all your efforts.