Announcement

**mnsandler** · November 24, 2013, 04:56 PM

Claude.

#1 The "Reset and Reprogram Interface" function can only be executed in if the plugin can communicate with the hub. its intended to reset the database only.

#2 there is no other cmds avail. if the plugin can't restart the port, hs via an action won't help.

#3. what you need to do is to use the triggers (interface crashes) to restart hs. there are scripts by Jon00 that you can run to do this.

wishlist item:I could make the 10 failed counter a user ini setting

here is what i don't understand about your situation...

the log you posted indicates the plugin was able to reconnect with the hub. so what happens or doesn't happen after that?

Mark

**claude** · November 24, 2013, 06:09 PM

When the hub loses power and eventually comes back online, the plug-in restarts the hub interface (port) only after it realizes the last 10 transmits have failed, so I've essentially lost the actions of those last 10 transmits.

Example: a power failure in the afternoon, power comes back an hour later (doesn't matter, the link is broken), then at sunset my 9 scheduled events fire up to light up the house. I'll be coming home to a dark house.

The plug-in has the logic to restart the hub interface (as per log), so what I'm asking is, in order of preference:

1- have the plug-in restart the hub interface as soon as the connection is broken (not wait for 10 failed transmits as currently); this would entail the plug-in checks the connection periodically. What would make this even better would be a user configurable (.ini) interval in minutes between attempts to restart the port
2- if '1' is not possible, then 'externalize' the logic the plug-in uses to restart the hub (port) so users can trigger the hub restart as an Insteon action; this is not ideal because I have no way of knowing that the interface needs to be restarted, unless that's what you were referring to when you mentioned "use the triggers (interface crashes)"
3- Last resort, as you proposed, make the 'failed counter' a user ini setting, but I'll still lose the action of at least 1 failed transmit

I'm really gunning for #1. What do you think?

**mnsandler** · November 24, 2013, 07:51 PM

i think i need to check the code on this. i don't think this should even be an issue with the hub since I use async communications. Each transmission is a unique connection to the hub. I don't hold the port open.

**claude** · November 25, 2013, 03:51 PM

Further testing...

After turning on detailed debug, I now better (but not totally) understand what is going on. See attachment with my comments inserted.

What I understand: if the hub is without power long enough and scheduled Insteon events run in the mean time triggering at least 10 commands that will fail, then the plug-in will shut itself down. When power is restored to the hub, it won't matter, no scheduled Insteon events will run successfully because the plug-in is down.

What I don't understand: In the attached log, I disconnect power from the hub at 14:26:00, but at 14:26:49 the plug-in detects communication with the hub is broken. No Insteon command was issued.

Questions:
1- What action did the plug-in take at 1426:49 to discover the loss of communication?
2- What situations will invoke the Insteon trigger "Interface crashes" ? A user disabling the plug-in via setup? The plug-in shutting itself down? Any other situation?

Wish list, again: Could the plug-in restart the hub interface automatically after detecting it lost communication with it, ideally immediately after the break of communication, with a user defined interval between attempts ? BTW, I realize now the number of failed commands as a user configurable INI parameter would not be useful in my situation, but an interval INI parameter would work for me; I would likely set it at 1 minute.

Attached Files

HomeSeer.log.txt (26.5 KB, 5 views)

**mnsandler** · November 25, 2013, 06:55 PM

Originally posted by claude View Post

After turning on detailed debug, I now better (but not totally) understand what is going on. See attachment with my comments inserted.

What I understand: if the hub is without power long enough and scheduled Insteon events run in the mean time triggering at least 10 commands that will fail, then the plug-in will shut itself down. When power is restored to the hub, it won't matter, no scheduled Insteon events will run successfully because the plug-in is down.

What I don't understand: In the attached log, I disconnect power from the hub at 14:26:00, but at 14:26:49 the plug-in detects communication with the hub is broken. No Insteon command was issued.

Questions:
1- What action did the plug-in take at 1426:49 to discover the loss of communication?
2- What situations will invoke the Insteon trigger "Interface crashes" ? A user disabling the plug-in via setup? The plug-in shutting itself down? Any other situation?

Wish list, again: Could the plug-in restart the hub interface automatically after detecting it lost communication with it, ideally immediately after the break of communication, with a user defined interval between attempts ? BTW, I realize now the number of failed commands as a user configurable INI parameter would not be useful in my situation, but an interval INI parameter would work for me; I would likely set it at 1 minute.

1. an exception in the processing of a prev cmd which timed out or didn't complete successfully
2. when the plugin can't reconnect with the interface after several attempts. that's it.

there is another trigger the Insteon Interface Stops Responding trigger which fires immediately after the 10 failed attempts but before the restart process begins, and or fails.

I'm not going to mess with the recovery logic right now, but i can put in an ini setting that would disable this logic, and just let the plugin continue to run with errors until you do something (ie restart hs) or power is restored.

**claude** · December 2, 2013, 04:01 PM

Originally posted by mnsandler View Post

... I'm not going to mess with the recovery logic right now, but i can put in an ini setting that would disable this logic, and just let the plugin continue to run with errors until you do something (ie restart hs) or power is restored.

So if I got this right, with the INI switch, the plug-in would still count the number of failed commands (10), issue the trigger "Interface stops responding", but would NOT attempt to restart the hub, nor would it issue the "Interface crashes" trigger.

This would certainly get me out of trouble short term. Is it a GO then?

Would you consider in some future timeframe, when things are not so hectic, to have the plug-in manage on its own the restart of the hub, along these lines:

1- have the plug-in restart the hub interface as soon as the connection is broken (not wait for 10 failed transmits as currently); this would entail the plug-in checks the connection periodically. What would make this even better would be a user configurable (.ini) interval in minutes between attempts to restart the port

**mnsandler** · December 3, 2013, 07:31 PM

Originally posted by claude View Post

So if I got this right, with the INI switch, the plug-in would still count the number of failed commands (10), issue the trigger "Interface stops responding", but would NOT attempt to restart the hub, nor would it issue the "Interface crashes" trigger.

This would certainly get me out of trouble short term. Is it a GO then?

Would you consider in some future timeframe, when things are not so hectic, to have the plug-in manage on its own the restart of the hub, along these lines:

1- have the plug-in restart the hub interface as soon as the connection is broken (not wait for 10 failed transmits as currently); this would entail the plug-in checks the connection periodically. What would make this even better would be a user configurable (.ini) interval in minutes between attempts to restart the port

I'm not ready to make any changes to the plugin on this quite yet. Based on my previous comments; this shouldn't even be an issue assuming the network connection is good and the hub has power.

have you tried the following ini settings

[Hub]
ResponseTime=3
GetXMLBufferInterval=1000

ResponseTime (seconds) - how long the plugin will wait for a response from the hub
GetXMLBufferInterval (msecs) - how frequently the plugin will poll the hub for new inbound messages.

please add the following to the end of your insteon.ini and restart hs

[Hub]
GetXMLBufferInterval=1500

**claude** · December 4, 2013, 12:25 AM

Originally posted by mnsandler View Post

... this shouldn't even be an issue assuming the network connection is good and the hub has power ...

In my initial post, I said "I'm experiencing frequent power failures; all of my computer system and peripherals are on UPS, except of course for the Insteon hub".

The issue is the hub may be without power for hours. Meanwhile the plug-in is on UPS, keeps running scheduled events and eventually shuts down because the last 10 commands failed, obviously because the hub is without power.

That's why at a minimum I need the plug-in to keep working regardless of COM errors with the hub. Eventually power will be restored to the hub and subsequent events would be processed.

**joegr** · December 4, 2013, 09:15 AM

Originally posted by claude View Post

In my initial post, I said "I'm experiencing frequent power failures; all of my computer system and peripherals are on UPS, except of course for the Insteon hub".

The issue is the hub may be without power for hours. Meanwhile the plug-in is on UPS, keeps running scheduled events and eventually shuts down because the last 10 commands failed, obviously because the hub is without power.

That's why at a minimum I need the plug-in to keep working regardless of COM errors with the hub. Eventually power will be restored to the hub and subsequent events would be processed.

I have the PLM, not the hub, but my system in similar in that the PC is backed up and the PLM is not. My work-around to the problem is that I have an event that sets a flag when the plug-in shuts down. After the UPS indicates that AC power is back, I have an event that reboots the PC if the flag saying the Plug-in crashed is set. So far, that has been working for me.

Here's an alternate thought on it. Would it be possible to have a special Insteon action that disables/enables the Insteon interface? Then I could have an event triggered by the UPS being on battery power that would disable the sending of Insteon commands while the power was out. That should prevent the plug-in from ever shutting down in the first place.

**mnsandler** · December 4, 2013, 10:54 AM

Originally posted by joegr View Post

I have the PLM, not the hub, but my system in similar in that the PC is backed up and the PLM is not. My work-around to the problem is that I have an event that sets a flag when the plug-in shuts down. After the UPS indicates that AC power is back, I have an event that reboots the PC if the flag saying the Plug-in crashed is set. So far, that has been working for me.

Here's an alternate thought on it. Would it be possible to have a special Insteon action that disables/enables the Insteon interface? Then I could have an event triggered by the UPS being on battery power that would disable the sending of Insteon commands while the power was out. That should prevent the plug-in from ever shutting down in the first place.

Joe, that's an interesting idea...instead of shutting down the plugin, just shut down the communications to the interface so one could restart communications when power is restored.

I'll put this on the list, but I can't mess with it immediately. I'm working on my other plugin, and taking care of hs3 bug right now

**brmeeke** · December 17, 2013, 08:58 AM

PLM Restarts

My system has been very stable for the last several months. Recently, I have noticed that Events are trying to run but nothing happens. I have done some digging in my Log files and found some lines that will hopefully be helpful.

12/13/2013 6:25:07 AM Insteon Interface successfully restarted.
12/13/2013 6:25:07 AM Insteon RF PLM contacted on COM3 with an Insteon address of 19.79.A6 and firmware version 98
12/13/2013 6:25:01 AM Insteon Attempting to restart the port and Interface...
12/13/2013 6:25:00 AM Insteon Warning! Communications lost. Attempting to restart the port...
12/13/2013 6:25:00 AM Insteon Transmit failed to device Kitchen Kitchen Counter (07.B8.12) for cmd 00 00 00 07 B8 12 0F 11 FF after 5 tries.
12/13/2013 6:25:00 AM Insteon The RF PLM has failed to transmit the last 10 commands.
12/13/2013 6:25:00 AM Device Control Device: Kitchen Kitchen Counter (\1) ON
12/13/2013 6:25:00 AM Event Event Trigger "Wake Up"

Does anyone have any ideas about why this is happening?

I am running HS2 Version 2.5.0.52
Insteon Plug-In Version 1.8.2.8 but I see that there is a newer version available.

**mnsandler** · December 17, 2013, 09:09 AM

Please do some searching on 'usb disconnect' within this forum and following some of the suggestions. This is also documented in the consolidated documentation thread at the top of the forum.

**claude** · January 16, 2014, 01:58 AM

Sorry to be a pain with this again, just trying to understand. Well maybe a bit also to try to influence you

I got another communication failure with the Hub, this time without a power failure.

<TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime1" style="WHITE-SPACE: nowrap" align=left>Jan-15 16:04:25 </TD> <TD class=LogPri1 align=left></TD> <TD class=LogType1 colSpan=3 align=left>Insteon </TD> <TD class=LogEntry1 colSpan=8 align=left>Warning: Unable to connect to the remote server </TD></TR></TBODY></TABLE> <TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime0" style="WHITE-SPACE: nowrap" align=left>Jan-15 16:04:25 </TD> <TD class=LogPri0 align=left></TD> <TD class=LogType0 colSpan=3 align=left>Insteon </TD> <TD class=LogEntry0 colSpan=8 align=left>You need to set the Interface Port on the Config page, and then restart HS. </TD></TR></TBODY></TABLE> <TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime1" style="WHITE-SPACE: nowrap" align=left>Jan-15 16:15:00 </TD> <TD class=LogPri1 align=left></TD> <TD class=LogType1 colSpan=3 align=left>Log Info </TD> <TD class=LogEntry1 colSpan=8 align=left>The log database is currently 6.39MB in size. </TD></TR></TBODY></TABLE> <TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime0" style="WHITE-SPACE: nowrap" align=left>Jan-15 16:41:36 </TD> <TD class=LogPri0 align=left></TD> <TD class=LogType0 colSpan=3 align=left>Event </TD> <TD class=LogEntry0 colSpan=8 align=left></TD></TR></TBODY></TABLE> <TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime1" style="WHITE-SPACE: nowrap" align=left>Jan-15 16:41:36 </TD> <TD class=LogPri1 align=left></TD> <TD class=LogType1 colSpan=3 align=left>Device Control </TD> <TD class=LogEntry1 colSpan=8 align=left></TD></TR></TBODY></TABLE> <TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime0" style="WHITE-SPACE: nowrap" align=left>Jan-15 16:41:36 </TD> <TD class=LogPri0 align=left></TD> <TD class=LogType0 colSpan=3 align=left>Device Control </TD> <TD class=LogEntry0 colSpan=8 align=left></TD></TR></TBODY></TABLE> <TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime1" style="WHITE-SPACE: nowrap" align=left>Jan-15 16:41:36 </TD> <TD class=LogPri1 align=left></TD> <TD class=LogType1 colSpan=3 align=left>Insteon </TD> <TD class=LogEntry1 colSpan=8 align=left></TD></TR></TBODY></TABLE> <TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime0" style="WHITE-SPACE: nowrap" align=left>Jan-15 16:41:36 </TD> <TD class=LogPri0 align=left></TD> <TD class=LogType0 colSpan=3 align=left>Insteon </TD> <TD class=LogEntry0 colSpan=8 align=left>The PLM has failed to transmit the last 10 commands. </TD></TR></TBODY></TABLE> <TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime1" style="WHITE-SPACE: nowrap" align=left>Jan-15 16:41:36 </TD> <TD class=LogPri1 align=left></TD> <TD class=LogType1 colSpan=3 align=left>Insteon </TD> <TD class=LogEntry1 colSpan=8 align=left>Warning! Communications lost. Attempting to restart the port... </TD></TR></TBODY></TABLE> <TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime0" style="WHITE-SPACE: nowrap" align=left>Jan-15 16:41:36 </TD> <TD class=LogPri0 align=left></TD> <TD class=LogType0 colSpan=3 align=left>Insteon </TD> <TD class=LogEntry0 colSpan=8 align=left></TD></TR></TBODY></TABLE> <TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime1" style="WHITE-SPACE: nowrap" align=left>Jan-15 16:41:37 </TD> <TD class=LogPri1 align=left></TD> <TD class=LogType1 colSpan=3 align=left>Insteon </TD> <TD class=LogEntry1 colSpan=8 align=left>Attempting to restart the port and Interface... </TD></TR></TBODY></TABLE> <TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime0" style="WHITE-SPACE: nowrap" align=left>Jan-15 16:41:44 </TD> <TD class=LogPri0 align=left></TD> <TD class=LogType0 colSpan=3 align=left>Insteon </TD> <TD class=LogEntry0 colSpan=8 align=left>PLM contacted on 192.168.1.15:25105 with an Insteon address of 1E.9E.CF and firmware version 9C </TD></TR></TBODY></TABLE> <TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime1" style="WHITE-SPACE: nowrap" align=left>Jan-15 16:41:45 </TD> <TD class=LogPri1 align=left></TD> <TD class=LogType1 colSpan=3 align=left>Insteon </TD> <TD class=LogEntry1 colSpan=8 align=left>Interface successfully restarted. </TD></TR></TBODY></TABLE>
As far as I can tell, all occurences start with that warning about being "Unable to connect to the remote server ", which I assume is the Hub, then the plug-in continues until 10 transmit failures, then the plug-in restarts the port and interface and then all is well.

When the plug-in first senses it can't connect to the remote server, would it not be possible for the plug-in to immediately restart the port and interface ? If it senses it can't connect to the remote server, what's the point of continuing anyway?

**mnsandler** · January 16, 2014, 07:17 AM

Originally posted by claude View Post

Sorry to be a pain with this again, just trying to understand. Well maybe a bit also to try to influence you

I got another communication failure with the Hub, this time without a power failure.

<TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime1" style="WHITE-SPACE: nowrap" align=left>Jan-15 16:04:25 </TD> <TD class=LogPri1 align=left></TD> <TD class=LogType1 colSpan=3 align=left>Insteon </TD> <TD class=LogEntry1 colSpan=8 align=left>Warning: Unable to connect to the remote server </TD></TR></TBODY></TABLE> <TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime0" style="WHITE-SPACE: nowrap" align=left>Jan-15 16:04:25 </TD> <TD class=LogPri0 align=left></TD> <TD class=LogType0 colSpan=3 align=left>Insteon </TD> <TD class=LogEntry0 colSpan=8 align=left>You need to set the Interface Port on the Config page, and then restart HS. </TD></TR></TBODY></TABLE> <TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime1" style="WHITE-SPACE: nowrap" align=left>Jan-15 16:15:00 </TD> <TD class=LogPri1 align=left></TD> <TD class=LogType1 colSpan=3 align=left>Log Info </TD> <TD class=LogEntry1 colSpan=8 align=left>The log database is currently 6.39MB in size. </TD></TR></TBODY></TABLE> <TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime0" style="WHITE-SPACE: nowrap" align=left>Jan-15 16:41:36 </TD> <TD class=LogPri0 align=left></TD> <TD class=LogType0 colSpan=3 align=left>Event </TD> <TD class=LogEntry0 colSpan=8 align=left></TD></TR></TBODY></TABLE> <TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime1" style="WHITE-SPACE: nowrap" align=left>Jan-15 16:41:36 </TD> <TD class=LogPri1 align=left></TD> <TD class=LogType1 colSpan=3 align=left>Device Control </TD> <TD class=LogEntry1 colSpan=8 align=left></TD></TR></TBODY></TABLE> <TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime0" style="WHITE-SPACE: nowrap" align=left>Jan-15 16:41:36 </TD> <TD class=LogPri0 align=left></TD> <TD class=LogType0 colSpan=3 align=left>Device Control </TD> <TD class=LogEntry0 colSpan=8 align=left></TD></TR></TBODY></TABLE> <TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime1" style="WHITE-SPACE: nowrap" align=left>Jan-15 16:41:36 </TD> <TD class=LogPri1 align=left></TD> <TD class=LogType1 colSpan=3 align=left>Insteon </TD> <TD class=LogEntry1 colSpan=8 align=left></TD></TR></TBODY></TABLE> <TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime0" style="WHITE-SPACE: nowrap" align=left>Jan-15 16:41:36 </TD> <TD class=LogPri0 align=left></TD> <TD class=LogType0 colSpan=3 align=left>Insteon </TD> <TD class=LogEntry0 colSpan=8 align=left>The PLM has failed to transmit the last 10 commands. </TD></TR></TBODY></TABLE> <TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime1" style="WHITE-SPACE: nowrap" align=left>Jan-15 16:41:36 </TD> <TD class=LogPri1 align=left></TD> <TD class=LogType1 colSpan=3 align=left>Insteon </TD> <TD class=LogEntry1 colSpan=8 align=left>Warning! Communications lost. Attempting to restart the port... </TD></TR></TBODY></TABLE> <TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime0" style="WHITE-SPACE: nowrap" align=left>Jan-15 16:41:36 </TD> <TD class=LogPri0 align=left></TD> <TD class=LogType0 colSpan=3 align=left>Insteon </TD> <TD class=LogEntry0 colSpan=8 align=left></TD></TR></TBODY></TABLE> <TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime1" style="WHITE-SPACE: nowrap" align=left>Jan-15 16:41:37 </TD> <TD class=LogPri1 align=left></TD> <TD class=LogType1 colSpan=3 align=left>Insteon </TD> <TD class=LogEntry1 colSpan=8 align=left>Attempting to restart the port and Interface... </TD></TR></TBODY></TABLE> <TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime0" style="WHITE-SPACE: nowrap" align=left>Jan-15 16:41:44 </TD> <TD class=LogPri0 align=left></TD> <TD class=LogType0 colSpan=3 align=left>Insteon </TD> <TD class=LogEntry0 colSpan=8 align=left>PLM contacted on 192.168.1.15:25105 with an Insteon address of 1E.9E.CF and firmware version 9C </TD></TR></TBODY></TABLE> <TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime1" style="WHITE-SPACE: nowrap" align=left>Jan-15 16:41:45 </TD> <TD class=LogPri1 align=left></TD> <TD class=LogType1 colSpan=3 align=left>Insteon </TD> <TD class=LogEntry1 colSpan=8 align=left>Interface successfully restarted. </TD></TR></TBODY></TABLE>
As far as I can tell, all occurences start with that warning about being "Unable to connect to the remote server ", which I assume is the Hub, then the plug-in continues until 10 transmit failures, then the plug-in restarts the port and interface and then all is well.

When the plug-in first senses it can't connect to the remote server, would it not be possible for the plug-in to immediately restart the port and interface ? If it senses it can't connect to the remote server, what's the point of continuing anyway?

I will make the 10 count a user ini setting just for you

you can set it to 1 if you like.

this recovery logic is also used for the usb plm which doesn't always need to be recovered/restarted after a single failure.

Announcement

Having the plug-in restart the hub interface after a communication break

Having the plug-in restart the hub interface after a communication break

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment