OPII IP comm problems vs. Thermostat comm?

pct88

Member
I am having problems with intermittent connection timeouts "Comm Status Retrying (100)" when using Dealer PC Access over IP to connect to an OPII with v3.0 firmware for downloading and uploading.  The connection is local with the PCA software running on a recent PC with Win 7 Pro, 100 Mb to a switch that also feeds the OPII.  HaikuHelper also suffers with occasional panel disconnect/reconnect warnings.
 
I have done various tests to narrow down the source of the problem including removing all non-critical programming lines, fully power cycling the panel, removing all serial and X10 connections.  It appears that when I disconnect the four RC series thermostat communications link the problem is reduced, but not solved.
 
Next of course I will disconnect one thermostat at a time to see if I can further isolate the source.  Yes, I confirmed that their addresses are properly 1, 2, 3, 4.
 
Other suggestions for ethernet communications issues?  I have also replaced cables, applied software upgrades, tried an alternate laptop, etc...
 
Welcome to the forum pct88!
 
Disconnecting and reconnecting the thermostat will not help.  Try disconnecting the network cable.
 
This is what used to fix my comm to thermostat issues.  I was able to watch this via a legacy Omnitouch serial screen and looking at the thermostat screen.
 
Relating to just networking problems I updated the firmware on my panel to V3.11A a few days ago.
 
Wierd stuff happening.  Three times now this past week the network interface went down for no apparent reason.
 
I could not even ping the interface. 
 
Yesterday it went down twice.  Now I have been somewhat paying attention to my networked OP devices.
 
The first time I just disconnected the network interface at the switch.  Waited a minute and plugged it back in and it came up.
 
Yesterday the first time it did this I did the same.  It happened again though within 2 hours of the first time.
 
This time I did a cold boot of the panel yesteday and its still fine today.
 
When this occurred I did not see any comm errors with the thermostat nor did the time go off as previously mentioned when I had my comm errors.
 
For the time being leaving it a FW 3.11A and just watching it a bit (really never paid much attention to it before).  I have though added the networked IP (8 now)  cameras to the Omnitouch 5.7e's.  The Omnitouch just bounces and stays at a "connecting prompt".
 
I haven't seen the issue relating to the network disconnecting that I wrote above after one day.  I did notice a post relating to this comm issue here:
 
http://cocoontech.com/forums/topic/23039-snaplink-mobile-on-iphone-temps-problem/?p=190149
 
If you are using Omnistat thermostats, output 8 and zone 16 become the transmit and receive of the thermostat bus. In this case do not name Zone 16, and set its type to Auxiliary. I suspect you currently have zone 16 set as a temperature zone and have it named "Thermostat"? This causes the Omni to try to interpret the serial data coming in on the thermostat bus, zone 16, as some sort of temperature, which will probably result in "unusual" readings.
 
Looking this morning at my configuration and I did have zone 16 named "Thermostat" for the last 10 years or so.  I also have multiple temperature and temperature/humidity sensors.  I did remove the "Thermostat" name for Zone 16 this morning.  I still don't know though why removing the network cable would make the serial comm issues go away when I would have this issue.
 
The manual states:
When any Thermostat Type is changed from the default setting of Not Used, Zone 16 and Output 8 are automatically configured for thermostats on all HAI Controller models except the Omni LT. There is no need to change the Zone Type for Zone 16 or Output Type for Output 8.
 
It doesn't really mention here about using the thermostat verbiage nor whether it causes any issues.  I named it for my own persoanal reference,
 
I have a similar intermittent problem with keypads occasionally loosing connection. They just beep and nothing in the logs. I've replaced keypads, wires, etc., nothing seems to help.
 
On the PC access side, I would suggest looking at other devices on your network. They may have bursts of Internet access and that could trip up the Omni timing. I see you said you have 100 Mb switches, and I used to have that problem with 100 Mb switches. I recommend you upgrade to 1000 Mb switches. I do realize the HAI panel only supports 100 Mb but still 1000 Mb switches operate better, so upgrade everything to 1 Gb that you can especially a switch between your PC and the HAI. 
 
I'm running 3.10D which seems pretty solid. HAI FW releases can be pretty flaky, so I think its best to wait a bit before upgrading.
 
I have not seen the "issue" of network transport connectivity come back in the last two days.
 
My question though is why am I seeing a serial comm issue and why is it related to the network transport?
 
Early on when I saw the serial comm issue I would just disconnect the network interface (a bad thing) and the comm issue would go away.
 
It presented itself on my Omnitouch 5.7 to Thermostat screens where as I would see the serial comm errors.  I would also see these errors with PCA connected. 
 
I have recently (a month or so ago) replaced the network cable and switch connectivity of the HAI panel to a Gb switch.  That said today my network is still a mix of 100 and Gb; but the switches are all Gb except for 1-2 of them.
 
Pete - for as long as you've been having these problems, have you ever sent your hardware to HAI to be evaluated?
 
This just seems like a long and ongoing problem that you've had.
 
Funny you ask about that Jon.  (personal no jest meant).
 
This board is relatively new.
 
I originally had a first gen OPII installed.  I updated it with a new OPII board a few years ago.  I put the old OPII board in Florida.  Its been fine.
 
That said I did a bad thing to my OPII board (new as it was).  Right after I installed it and in the dark I shorted the power supply.  I saw smoke; killed it per say.  That said HAI tech replaced it for me.  Great bunch of guys that they are.  Actually that is documented on this very forum and subtopic as I was concerned on the upgrade.  It was easy though.  The comm issues came long after (a couple of years maybe).
 
Not really content and wanting something new decided to upgrade my RC-80 to a new Omnistat2.  I did though procrastinate some maybe 2 years after I purchased it and it sat on a shelf for a bit before I decided to install it.
 
I installed it and within a day started to see the comm errors on my Omnitouch 5.7 screens and PCA.   That said I called HAI customer support and they were great.  We did an online diagnostics and HAI tech issued an RMA.  They sent me a new Omnitouch2.  I still had the same issue though.  I then unplugged my network cable (someone had mentioned to me on the forum that would maybe fix my problem).  It did fix it.
 
In the last couple of years and very intermittently I have seen the "serial com" "thermostat" and "time" issue.  It appears that the clock gets totally out of sync and starts to lose time when the serial comm issue with the thermostat crops up.  I do not pay attention typically.  It came to my attention the last time because my schedules of lighting events were totally out of wack; off by maybe an hour.  I don't typically pay attention to this stuff; because it just works (which I really like)
 
The last time the problem happened I went to watching the time and saw that it was losing more that a few minutes in an hour.  The final fix though was removing the network cable for a period of time.  That and while I did that replaced the network cable and the network switch connection (gb now).  The cable was old but fine after I tested it.  The switch is old and still functioning just fine and is just servering as an intermediate switch while I move the rest of the "stuff" to the Gb switches.
 
I did fall back for a time to the RC80 (maybe a week) and didn't see any comm issues.  From many years ago I do have installed the additional HAI power board and 24VAC. 
 
Just a few days back upgraded the firmware to 3.11A.  The network "disconnect" issue is brand new.  I have never seen this issue before.
 
The above noted; everything seems to be fine today.
 
It would not be an issue to send the board off for diagnostics as long as I could keep the system going with another board.
 
I don't really think though the board is the issue; but who am I to say? 
 
 
I continue to work on tracking the source of the issue.  I tried one of the suggestions above:  I replaced my FiOS 10/100 Actiontec router with a 10/100/1000 model.  Upgraded it to the most current firmware.  Plugged the OPII into one of the ports.  Curious-  I could no longer connect to the OPII.  Checked the port settings-  it had auto configured to 10 Mb half duplex.  Tried full duplex, 100 mbit, etc...  No luck.  Show link, but no communications.  Checked the IP address at the keypad.  Power cycled the OPII.  Changed the IP address at the keypad.
 
Finally I moved the OPII ethernet cable over to a Netgear 10/100 switch hooked back to the Actiontec.  Now I have a working connection to the OPII again.  So it appears the Actiontec and OPII aren't on speaking terms.  
 
I then borrowed my office Fluke cabling tester and confirmed that the cable between the OPII and the PC is fully certified to Cat 5e.
 
Next I configured the Win 7 Pro Core 2 Duo processor PC to a manual IP address and hooked the PC to the OPII through a crossover cable.  Connect with Dealer PC Access as I had done previously, but this time with the thermostat communications bus still disconnected.  Darn-  I still see bursty communications.  It is still giving occasional Comm Status: Retrying 100 errors-  "Communications--> Read from Controller...", deselect all but "Events", "OK".  To download only the 254 log records just took 50 seconds including a long stretch at the Retrying error.
 
I am wondering if there is an underlying issue with the ethernet implementation on the panel.  This is one of the older panels (chip based firmware)-  I purchased an OmniTouch 5.7e that I hoped to expand to more, but if I can't get reliable network connectivity I don't think I can invest more in network connected options...
 
Next I guess I have to try defaulting the entire panel to factory settings and see what the communications performance looks like with no programs running.
 
Do you have an older FW chip that you can try? 
 
The panel I have in FL is still on an older FW revision level and I have no issues with the network connectivity via the WAN link here in the midwest.  I do have a FIOS combo router there but the HAI panel is connected to a smaller 10/100 switch.
 
I did a quick look at the newest panel and it is running at 10 Mbs. That actually could cause issues maybe with newer switches.
 
Here my network fix was disconnecting the network cable for a time then reconnecting it.  I do also have a serial link to the panel which I can utilize with PCA.
 
I started to notice the network issues after I added HaikuHelper to the mix:  since it notifies me when the network connection drops and returns it became obvious that there is something going on with that connection.  HaikuHelper requires firmware 3.0 at a minimum, so I can't really roll back further without breaking that functionality.
 
Apologies in advance...but have a read through this thread; diagnostics et al.  I am guessing you might have already read it though.  Not totally sure though if getting very granular with diagnostics will help you though.
 
http://cocoontech.com/forums/topic/21333-new-firmware-310a-old-commtime-problems-back/page-3
 
Here I was able to test because I am still in both legacy and newer modes with the HAI OPII panel.  IE: I have serially connected Omnitouch 5.7's via an Omnitouch touchscreen hub and now newer Omnitouch 5.7e's.  I also have now both analog video using the older analog HAI video hub and now IP cameras. 
 
So here the comm issues and network issues would go away with a "minutes" disconnection of the network interface.  Odd though that this would fix my issues.
 
When do you see the "Comm Status Retrying (100)" messages? You said while downloading and uploading but more specifically is it when you are starting to upload programs? If so you do not have a problem.

Let me explain. When you are writing to the controller, the controller must first erase the existing items. So when PC Access tells the controller it is going to write programs, for example, the controller goes through and zeros out all program memory before telling PC Access it is OK to begin. This often takes longer than the timeout period, so PC Access retries. This is normal and almost guaranteed for programs. Depending on how many names, voices, etc it is possible to see this at the beginning of other sections too, but far less common. PC Access should retry, but continue normally as soon as the erase has completed.

This situation can be exacerbated when the controller is very busy or has multiple connections polling for data while it it trying to do the erase. Also some inexpensive "home" class switches and routers occasionally drop packets which can cause this problem. This is especially true if you are running PC Access on a wireless laptop, but based on your description I do not suspect this is the problem in your case.

If you are actually timing out and disconnecting then there is a problem. But occasional retries are normal. All network programs have retries, I guess the problem is that PC Access reports them where as other programs like IE don't usually report problems unless it fails completely.

In summary occasional timeouts are normal and you should not have to roll back your firmware. In general we recommend that you use the latest firmware because we are constantly adding features, fixing bugs and tweaking performance. I have an OmniPro II in my own home and always test and use the latest firmware. One tip though, before upgrading your firmware read your entire configuration into PC Access and save it. After upgrading your firmware I recommend writing it back. The new firmware should upgrade your data area automatically, and it works 99.99% of the time but, just to be sure I write my configuration back to the controller after a firmware update. Better safe than sorry...

I see some other replies with other, I believe unrelated problems, and strongly suggest that they contact HAI tech support.
 
Yes, I see the Retrying message during upload/download and that isn't a problem.  When it entirely times out though that is a problem...
 
The panel is the chip based firmware and is currently at version 3.0; I would be happy to upgrade to a newer firmware if there is any chance that might improve the network communications.
 
Thanks
 
Here too I have seen the same.  That personally also was never an issue.
 
I have though never seen a total network disconnect as I saw with Firmware Version 3.11A.  I haven't seen it since my last post.  That said though too now I am not seeing the thermostat communications problems with firmware 3.11A. 
 
I don't know what is worst though; having the serial comm issue with the thermostats accompanied by a the time going off minutes by the hour or losing the network connection.
 
Personally though now with Omnitouch 5.7e's and IP cameras connected the network issue is worst.  I can though fall back and replace the Omnitouch 5.7e's  with the legacy serially connected 5.7's and I do have a video hub in place for video on these guys.
 
The above noted if you go to Firmware 3.11A you will not see the serial comm issues with the thermostats (from what I can tell). You may though have a network issue; from what I have seen so far once only though.
 
Back
Top