Counting/estimating PLM links

[gfridland]

I know this subject has been discussed in the past...but no practical/reliable solution has been reached.

 

I am trying to determine a link count in my PLM (dual Band) to ensure that I have enough unused capacity to add an additional set of devices.

 

I have attempted to count PLM links via diagnostics (Tools/Diagnostics/Show PLM Links Table). I repeatedly get 450 links (ten times in a row), then, one time, the count was 759...but...the 450 count has been solid. I even watched the event viewer during the count and no other traffic was reported during the counting process.

 

Is there a reliable way to verify accuracy?

 

If I count the ISY links (Tools/Diagnostics/Show ISY Links Table) I get 584 total for just my 7 KPLs. In addition to the KPLs I have 70 other devices which average approx 13 links per device. If this was true, I would have already exceeded the Dual Band PLM capacity of 1024 links...

 

Is it a valid assumption that the number of links in the PLM should be the combined total of links in the ISY link Table for all devices? If this is true, I would already be greater than 1024.

 

Thanks for the continued support!

[LeeG]

gfridland

 

There is no direct relation between the number of ISY link records scattered across the Insteon device inventory and the number of link records in the PLM. So long as the Show PLM Links Table is not interrupted by other Insteon traffic the PLM link record count is accurate. The 450 number should be the correct number when it can be obtained several times in a row.

 

The problem with the PLM link count is inbound Insteon traffic to the PLM alters the Next Link Record pointer the PLM firmware is using during the Get Next command sequence used by the ISY. This results in the Next Link Record pointer being bumped ahead such that link records are skipped (link count low) or being reset to a link record that has already been retrieved such that link records are retrieved multiple times (link count high).

 

If you have a spare FilterLinc the PLM can be plugged into the Filtered outlet on the FilterLinc to keep inbound Insteon traffic from reaching the PLM. It is not perfect as close by Insteon device signals may still get through but it does isolate most of the devices. Of course be sure to remove the FilterLinc when done.

 

Lee

[gfridland]

Thanks LeeG,

 

I apreciate the clarification...what puzzled me was that 759 count that appeared during one count. I would have been ok if the number was below 450, but when I got greater than 450, it raised questions.

 

Also...I did use a filterlink to isolate the PLM from the "grid", but was unable to ensure no RF was present.

 

In your experience, is 450 a reasonable count for 70 devices (majority Togglelink, but 6 IOLinc, and 1 EZIO8) and 7 KPLs?

 

Thanks again!

[LeeG]

If the number of PLM link records counted has reached say 350 and an inbound Insteon message is received that resolves to a PLM link record that is near the beginning of the link database the Get Next command counting link records will return the next link record following the one that resolved to the inbound Insteon message. This results in double counting many of the link records. If more than one inbound message is received the count can be very high, much higher than can be physically held in the PLM memory because many link records can be counted multiple times.

 

The number seems reasonable but it very much depends on the number of ISY Scenes and how many devices are responders in the Scene. Links between devices (not between PLM and device) can quickly increase the total ISY Link count but has nothing to do with the number of link records in the PLM itself. Only those link records needed for the PLM to be aware of button/paddle presses and those needed for the PLM to control Scene devices when the Scene name is used in Programs or the Admin Console are actually in the PLM itself. The ISY Links information is maintained in the ISY memory which is quite separate from the PLM link database memory.

[gfridland]

Thanks for the clarification LeeG!!!