Can't view PLM links table

[MarkJames]

I'm doing some troubleshooting and am trying to check my PLM links.

 

I've stopped all programs, disabled my Elk, and turned off the PC that integrates my ISY to Homeseer so the ISY and PLM should be completely isolated.

 

I start the Tools->Diagnostics->Show PLM Links and it counts to maybe 4 at which point the event viewer shows 'Writing 0 bytes to devices' and the count stops.

 

Is there something I can do to clear this behavior?

 

mark

 

[edit] I've disabled the write to device options and it goes farther now - but it still won't make it more than 30-40 links in.  I should be pushing 750-1000 links

[MarkJames]

Here's a video screencap of what I'm seeing

 

https://www.dropbox.com/s/q8j492hm4iwnndx/Rec_008.avi?dl=0

[paulbates]

Mark-

 

I've never run it with event viewer, but I don't think that should be a problem.

 

Are all insteon devices working, to the best of your knowledge, if you issue a command to a sampling of devices from the admin console, they all respond as expected? If you go to those devices, and turn them on and off, the admin console is updated?

 

If yes, then, I would try unplugging the ISY and and PLM. Plugin the plm, wait 10 seconds, then the ISY, and try the 'Show PLM Links Table' again

 

If no, then its potentially more difficult.. let's discuss

 

Paul

[MarkJames]

Everything works quite well - I'm just trying to figure out some issues with devices that are quite a distance from the main house.  We have a pool in the backyard and I have some communication issues with devices there so I've been swapping out some of the dual-band devices in the house with the older ones down there attempting to improve matters.

 

I wanted to check my PLM links just to ensure that I'm not over the 1k link limit.  I have a LOT of links and a LOT of devices so it's a constant fear.  I'm wondering if perhaps this PLM behavior isn't a manifestation of too many links but I've not found any articles/posts on the topic.

 

The view you're seeing is actually from exactly what you suggested - I powered down the PLM for a minute and the ISY as well.  I then powered up the PLM for a minute or two then powered up the ISY.  I've put a condition on the programs main folder (if 0=1) so make sure that nothing in the folder runs.  I've turned off my Elk and disabled the M1XEP connection through the ISY configuration and I've turned off my Homeseer PC.  That's why I was running the event viewer - to make sure that I didn't have some unwanted traffic messing up the count.

 

I'll reboot again and see....

 

mark

[paulbates]

Mark

 

What's the report from the 'show ISY links' look like? 

 

Paul

[MarkJames]

Mark

 

What's the report from the 'show ISY links' look like? 

 

Paul

 

I posted a link of the session above so you can see it happen.  It just counts up a few and then stops with no message.  Weird.

 

mark

[paulbates]

Here's a video screencap of what I'm seeing

 

https://www.dropbox.com/s/q8j492hm4iwnndx/Rec_008.avi?dl=0

 

I posted a link of the session above so you can see it happen.  It just counts up a few and then stops with no message.  Weird.

 

mark

 

Mark

The video shows the PLM links table, and is queried from the PLM. I wanted to see what the contents of the ISY Links table, which is the db of links the ISY keeps based on the devices you add to it. Its used to build the PLM links table that the insteon network uses.  One of your options is doing a restore PLM from that db, but before asking you to try it, given the size of your network, I wondered what the master db in the ISY looks like.

 

It doesn't make sense that everything seems to be working both ISY to Insteon device, and Insteon device to ISY, with the table that small. Seeing what the ISY has would be a helpful next step

 

PAul

[LeeG]

The Show PLM Links Table will be incorrect if any Insteon activity reaches the PLM during the Show.  Motion Sensor activity, thermostat humidity, Insteon switches, anything that matches a PLM link record.   The next PLM record read will be the one following the unexpected Insteon message.

 

The 02 6A 15 indicates the last active record has been read but many PLM link records can be skipped if an Insteon activity reaches the PLM. 

[MarkJames]

Mark

The video shows the PLM links table, and is queried from the PLM. I wanted to see what the contents of the ISY Links table, which is the db of links the ISY keeps based on the devices you add to it. Its used to build the PLM links table that the insteon network uses.  One of your options is doing a restore PLM from that db, but before asking you to try it, given the size of your network, I wondered what the master db in the ISY looks like.

 

It doesn't make sense that everything seems to be working both ISY to Insteon device, and Insteon device to ISY, with the table that small. Seeing what the ISY has would be a helpful next step

 

PAul

 

Hi Paul,

 

Is there an option to show the ISY links table for the entire database?  All I see is an option to show it device by device...

 

mark

[MarkJames]

The Show PLM Links Table will be incorrect if any Insteon activity reaches the PLM during the Show.  Motion Sensor activity, thermostat humidity, Insteon switches, anything that matches a PLM link record.   The next PLM record read will be the one following the unexpected Insteon message.

 

The 02 6A 15 indicates the last active record has been read but many PLM link records can be skipped if an Insteon activity reaches the PLM. 

 

Thanks Lee,

 

I understand about the issue with  Insteon traffic.  That's why I shut down my Elk and homeseer and stopped all programs.  I'm the only one in the house so nobody is changing any devices.  I have no wireless devices in my network whatsoever.  

 

How would I determine what is reaching the PLM that's stopping the link enumeration?  Alternatively, is there a way to determine the number of links the ISY has recorded for me so that I can determine if I've maxed out my PLM?

 

mark

[MarkJames]

Ok..... so I've been going through my ISY links device by device and I think it's pretty clear that I have more links than I should.  I'm only a quarter of the way through and I'm well over 700 links already.  I'll go through them all and see how many I actually have but it will take some time.

 

My understanding is that 1024 or so is the max on a 2413S.

 

So... any thoughts on how to reduce the number of links without reducing functionality?

 

mark

 

[edit]

 

1764 links is what I count by totallng up all ISY links from all devices....  I think I have a problem 

[LeeG]

The 2413S holds 992 link records.  The initial 256 decimal bytes are control information.

 

I can see the unexpected Insteon traffic in the Event Viewer (running 4.4.6).  I triggered a motion sensor and lost more than 50 PLM link records displayed.

 

The same Insteon traffic can produce false high counts.   If the PLM link record that matches the Insteon traffic has already been displayed by Show, all the link records past the motion sensor link will be displayed again.   I've seen folks report 1500 PLM link records displayed even though that is physically impossible for a 2413S.  The Show PLM Links Table uses

 

"Thu 05/26/2016 05:59:55 PM : [GET-NEXT-LNK-RSP] 02 6A 06". 

 

Normally the last record accessed is from the Show.   However, Get-Next-LNK reads the next record, even if the last PLM link record accessed was due to an inbound Insteon traffic.   So, either the PLM link record count will be too low or too high if inbound traffic.

 

 

Thu 05/26/2016 05:59:54 PM : [GET-LNK-RSP ] 02 69 06 

Thu 05/26/2016 05:59:54 PM : [LNK-DATA    ] 02 57 A2 09 01 DB B6 07 07 FF 

Thu 05/26/2016 05:59:55 PM : [GET-NEXT-LNK-RSP] 02 6A 06 

Thu 05/26/2016 05:59:55 PM : [LNK-DATA    ] 02 57 A2 0A 01 DB B6 07 07 FF 

Thu 05/26/2016 05:59:55 PM : [GET-NEXT-LNK-RSP] 02 6A 06 

Thu 05/26/2016 05:59:55 PM : [LNK-DATA    ] 02 57 A2 0B 01 DB B6 07 07 FF 

Thu 05/26/2016 05:59:56 PM : [GET-NEXT-LNK-RSP] 02 6A 06 

Thu 05/26/2016 05:59:56 PM : [LNK-DATA    ] 02 57 A2 0C 01 DB B6 07 07 FF 

Thu 05/26/2016 05:59:56 PM : [GET-NEXT-LNK-RSP] 02 6A 06 

Thu 05/26/2016 05:59:56 PM : [LNK-DATA    ] 02 57 A2 0D 01 DB B6 07 07 FF 

Thu 05/26/2016 05:59:56 PM : [iNST-SRX    ] 02 50 13.F2.57 00.00.01 CB 11 01    LTONRR (01) - motion sensor

Thu 05/26/2016 05:59:56 PM : [std-Group   ] 13.F2.57-->Group=1, Max Hops=3, Hops Left=2

Thu 05/26/2016 05:59:56 PM : [D2D EVENT   ] Event [13 F2 57 1] [DON] [1] uom=0 prec=-1

Thu 05/26/2016 05:59:56 PM : [  13 F2 57 1]      DON   1

Thu 05/26/2016 05:59:56 PM : [D2D-CMP 006D] CTL [13 F2 57 1] DON op=1 Event(val=1 uom=0 prec=-1) is Condition(val=0 uom=0 prec=-1) --> true

Thu 05/26/2016 05:59:56 PM : [D2D-CMP 0068] CTL [13 F2 57 1] DON op=1 Event(val=1 uom=0 prec=-1) is Condition(val=0 uom=0 prec=-1) --> true

Thu 05/26/2016 05:59:56 PM : [D2D-CMP 0067] CTL [13 F2 57 1] DON op=1 Event(val=1 uom=0 prec=-1) is Condition(val=0 uom=0 prec=-1) --> true

Thu 05/26/2016 05:59:57 PM : [iNST-SRX    ] 02 50 13.F2.57 00.00.01 CB 11 01    LTONRR (01)

Thu 05/26/2016 05:59:57 PM : [std-Group   ] 13.F2.57-->Group=1, Max Hops=3, Hops Left=2

Thu 05/26/2016 05:59:57 PM : [iNST-DUP    ] Previous message ignored.

Thu 05/26/2016 05:59:57 PM : [GET-NEXT-LNK-RSP] 02 6A 06 

Thu 05/26/2016 05:59:57 PM : [LNK-DATA    ] 02 57 A2 02 13 F2 57 10 01 34 

Thu 05/26/2016 05:59:57 PM : [iNST-TX-I1  ] 02 62 00 00 91 CF 13 00

Thu 05/26/2016 05:59:57 PM : [GET-NEXT-LNK-RSP] 02 6A 06 

 

Usually PLM links go up because of testing.   Scenes are changed and rewritten.  It is common for PLM link counts to go down after doing a Restore Modem (PLM) because the now deleted Scene records are not written by a Restore Modem (PLM).   Once you determine what is causing the Insteon traffic so a good Show PLM Links Table is obtained (should be run 3-4 times to insure count is good), do a Restore Modem (PLM) followed by another Show PLM Links Table (show done a few times to insure new count is correct).

 

Now you know your starting point.  May not be necessary to do anything at this point.

 

If link records still need to be reduced, combining Scenes, eliminating Scenes,  eliminating devices in Scenes, are some of the actions that can be done.   It is very rare to reach the PLM max when the PLM has been cleaned up with a Restore Modem (PLM).  

[MarkJames]

Thanks.  It occurred to me that I do have one RF device - a leak sensor.  I'll go pull the battery from it to see if it's the problem.

 

Should the number of PLM links be roughly the same as the number of ISY links?

 

mark

[LeeG]

Don't know what "ISY links" means.

 

The PLM links start with all the devices defined to the ISY.  An 8 button KPL has 9 link records, for example.  One link for each button so a button press is sent to the PLM/ISY plus one additional needed for I2CS and other things.   

 

Then come the link records required for the Scenes you define.  The number depends, of course, on the number of devices.   Are the devices Controller only, Responder only, or both.

[MarkJames]

By ISY Links I mean if I go into the Tools->Diagnostics->Show ISY Links I can see the records that ISY holds for each device.  If I go to each device one by one and total up all the records that ISY holds I get 1764 links from the 89 different physical devices I have.  This would represent the links for each of the single paddle devices, i/o devices, and KPL buttons plus links for all the scene memberships.  It's this number that has me concerned.

 

I've just finished deleting a very large scene that held every device whether it was a light or a KPL button so that I could turn off every device in the house from one button press.  That should save me a bunch of PLM space.  I'm going to go pull the battery on that leak sensor then restore my PLM and see if I can get a PLM link count.

 

mark

[LeeG]

That is not the number of links required by the PLM.    The Show ISY Links reflects what is written in the device, not the PLM.

 

Here is a link displayed by Show ISY Links Table for an ICON switch.

 

1 : 0FF0 : E2 01 22.80.0B FF 1F 01

 

It is a Controller link pointing to the PLM.    The link record required in the PLM was created when the device was added to the ISY.  Nothing new is required in the PLM.

 

Here are the Show ISY Links for an I2CS Motion Sensor that is in a Scene,   No PLM link records are required for this device beyond what was created when Motion Sensor added to ISY, even though it is a Controller in an ISY Scene.

 

0 : 0FF8 : E2 01 22.80.0B 01 00 01

1 : 0FF0 : E2 02 22.80.0B 01 00 02

2 : 0FE8 : E2 03 22.80.0B 01 00 03

3 : 0FE0 : E2 01 13.08.1B 01 00 01

4 : 0FD8 : E2 01 14.41.95 01 00 01

5 : 0FD0 : 00 00 00.00.00 00 00 00

 

 

I think your concern over the 17xx count is not an issue.    The Show ISY Links do indicated links are required in other devices but not the PLM.

 

Of course there are ISY Scenes that add to the PLM count but Show ISY Links does itself not mean much.

[MarkJames]

Thanks for the explanation.

 

I've exhausted everything I can think of to do a PLM link count.   I backed up my ISY and did a restore PLM followed by a Show PLM links but I still can't get past around 40.  

 

I know it's not 'necessary' to do a link count - it just *should* work and it's not. The PLM I'm using is a version 95 that's a few years old.  Since installing it a couple of weeks ago I've had a few spontaneous ALL ON events that are quite maddening.  Yesterday I'd posted that I thought it resolved but then last night it happened again - not ALL on but MANY on and my GDO open this morning much to my chagrine.  I'm going to change control of that over to my Elk once I get the rest of this sorted.

 

I purchased a new PLM to keep as a backup that is version 9E. I'm going to replace the v95 with the v9E and see if any of these problems resolve.

 

Mark

[LeeG]

Is the PLM link record count always 40?  How many Insteon devices are installed and what types.

[MarkJames]

No - it sometimes get 4, sometimes 2, sometimes 1 - it's random.  If you watch the video I posted you can see both the count in action and the event viewer.  I see no insteon traffic in the event viewer that should disrupt the PLM link count - it just stops with the 02 6A 15.  The highest I've seen it count is 41. 

 

I have 89 devices plus another 10-12 that are not controlled by ISY in my system.  Of the 89 there are at least 20 KPL8's and another 20 KPL 6's.  So at the very least I should have 160 for the 8's and 120 for the 6's. Plus a minimum of 1 for each of the other 50 devices so I should see a minimum of 350 or so links.

 

I should probably just leave things alone as it works.  The only thing is that since my PLM failure a couple of weeks ago I've been getting random ALL-ON events so I'm looking at all aspects to see what may be wrong.  Most of the reading I've done points at the PLM so that's where I've been focussing.  With my previous PLM's (and I've gone through at least 5-6 in the dozen or more years I've been involved) I've NEVER had an ALL-ON event so this is very disconcerting.

 

Here is a list of my devices.  I know it's not considered right to post the insteon addresses but I'm not bothered.

 

mark

[LeeG]

The fact that Insteon devices exist that are not known to the PLM likely means the Insteon messages are coming from one or more of those devices not defined to the ISY.  The Insteon address(es) are not in the PLM link list but the PLM sees the messages and searches the link table for a match.  Not shown because the Insteon address is not known to the PLM but it breaks the GET-NEXT process.

 

Suggest you leave things alone unless you can stop all Insteon activity.

 

I publish my Insteon addresses all the time.  Not hard to discover what they are if someone really wants to know.

[MarkJames]

Thanks for your help, Lee.

 

I'm going to make an effort to map all my devices and eliminate any non-linked ones.  I've got ones dating back to the first years of Insteon that I recycle for various projects and so there may be ones that I don't know about.  As I sit here watching links write I just watched a series of X10 addresses get sent a P7 P6 P5 P4 P4/On series.  That would indicate that it comes from a very old device as I migrated to Insteon from X10 controlled by JDS Stargate.

 

I'd appreciate your opinion on the value of changing some of my older devices out for newer dual bands.  My installation is quite extensive and some of the more remote devices can be 300+ ft from the PLM.  Though I have a number of 2443 access points around and a growing number of dual band devices I still find a lot of missed writes and reads - the 'max hops=3, hops left=0' or the Set-MSB(0) errors during  link writes.  My home has 2 meters - it's a bit unusual.  It was an initiative from our power provider 30 years ago to charge one rate for regular electricity usage and a lower rate for heating usage.  So the power comes into our home through a main meter then feeds a second panel through a second meter.  This means that some of the Insteon signals must pass through a meter no matter where I put the PLM.  I try to mitigate the problem by using phase couplers on the panels and access points to form a bridge past the meters but there's still issues.  

 

Do dual band devices help as much as access points and is an installation with a large number of dual band devices generally more robust?  While it would be a fairly costly upgrade I've already got more than $10k invested in Insteon.  Another $1500 or so to solidify it would be do-able.

 

Thanks,

 

mark

[LeeG]

That is an interesting configuration.  I have 6-7 Insteon devices in a detached garage that is more than 300' of cable from main house (I buried that cable) plus the interior wiring inside garage.  I have no problem controlling those devices from the house and none are Dual Band.   The devices were installed before there was Dual Band beyond the old SignaLincs.  I have no experience going through a meter but I suspect having Dual Band devices on both 120v legs and on both sides of the meter would help with comm reliability.

 

I have many Dual Band devices driven by my test ISY.

 

Perhaps there other Insteon folks that have multiple meters who can pass on their experience.