OMNI II network ports locks-up

At least one person here had said that his time drifting was fixed with 3.13 and the only way he could keep it from drifting before was to unplug the network cable, so something there changed.
 
Since you are doing packet analysis and know the loss, it might help others if you just did the firmware upgrade and reported back if it is indeed better on 3.13.
 
Yep, will probably try that.   FYI...I just go off the phone with HAI tech support.  I spoke with an extremely arrogant guy who said they had never heard of this issue before and that if my crossover cable test worked then there was nothing wrong with the OmniPro.   Half the reason I purchased this was to be able to use it for remote access.
Very frustrating...these controllers are not cheap.
-Warren
 
Upgrading to v3.13 did not fix anything.  Have not experienced complete lockup but having anywhere from 30% to 70% packet loss.  The end result of this is that my Haiku iphone app and PCAccess run very slow and will appear to no be working while ethernet comms are frequently being re-tried.  I'm not an IT expert but that's my interpretation.
What I don't understand is that I can get nearly 0% loss when connected directly to the OPII (no switch).   However; when a simple 10/100 unmanaged switch is inserted (only laptop and OPII connected), I still get up to 30% packet loss.
Can this be some sort of auto negotiation problem with the OPII or the switch?    
Is the network port on the board an embedded ethernet/serial converter?    
Is this a hardware problem that firmware will never be able to fix?
 
Thanks.
 
Based on your description, my first thought would be an auto negotiation issue.  Do you have an old 10 mb hub that you could place between the switch and the OPII?
 
I did try using an old hub (10 mb) and it not fix my issues.  But that was me and here.
 
Once the NIC breaks; it stays broken unless you disconnect it.
 
Removing the network interface though did fix my serial comm issues (noticed mostly with the Omnitouch and Omnistat serial comms)
 
The network interface basically just quits functioning; there is nothing complex about that. (sniffing the network interface will only confirm this)
 
Personally here with the 3.13 FW I had issues initially and then I did a cold reset uploading my stuff to a blank panel.
 
I am not used to doing a cold reset with my OPII panel such that it was a PITA (but easy). 
 
Read the threads above relating to my fix of my network and timing issues and serial issues with the 3.12 FW.
 
It worked for me.  I will most likely downgrade to 3.12 and do my networking thing if the board hiccups again.
 
That said per an open ticket with HAI tech.  HAI tech folks were polite and very responsive to my concerns.
 
HAI tech told me that there was nothing wrong with my board and it related to using a legacy and promiscuous old network interface on the HAI OPII. 
 
IE: board was an old design which couldn't handle a large multidevice network.
 
I am today close to 128 devices on my network.  I did post a message relating to my open ticket and response to said ticket here on the forum.
 
I am also now using managed Gb switches and unmanaged Gb switches mostly these days.
 
So instead I just played until I got it it working fine.  (HAI OPII with FW 3.12 and  my DIY fix worked great).
 
This endeavor got me playing and modifying a micro-router. 
 
The micro router will let me connect to the panel via the regular home wired or wireless network and even do out of band management with it via whatever topology I want to utilize (network or serial). (that and I can also do the email thing stuff) .
 
Looking now too at an even smaller micro router that I can modify in a similiar fashion.
 
Now using FW 3.13 (sans hiccup) the panel is on the main sub net of my home network.  It does not have any networking issues, significant time loss or serial problems which all seemed to occur at the same time.
 
Maybe it was an issue with the firmware?  Maybe they just didn't want to repair my OPII board?
 
Maybe there has been a subtle design change in more recent OPII boards that we do not know about?
 
Maybe having the email board with a clock sync, serial connection and a second network interface fixed the issue?
 
Mine is about 4 years old now.  I do have my original board in another home and its still functioning just fine (~ 2003 OPII board).
 
Above PCT88 just replaced his board with a newer one and his problem went away.
 
Once I swapped out my board I have not had any problems since.  The old board sits in the basement awaiting further experimenting-  though I find it unlikely that I will put much effort into that given that the new one is happy to deal with the same network conditions.  I did isolate my assorted Arecont megapixel cameras from the OPII early in the troubleshooting:  they provide many tens of MB of continuous traffic on the network.
 
So pete_c...great info....but not quite sure I'm following everything in your chain of events.  
 
Is your original OPII network interface functioning normally now on v3.13 while connected to your main subnet?  And it was a hard reset that finally fixed your issue permanently?
 
Or is it some network isolation you've done that was your solution?
 
My OPII was purchased in about Sept 2013.   How do I tell if I am using the legacy NIC?  
 
I also called HAI tech.   When I told them it functioned normally directly connected to laptop with crossover cable he said the board was fine and he would not even listen to anything else I had to say.   I specifically asked if there were any known issues related to this network interface problem....answer was no.   I find this hard to believe.
 
 
 
Thanks,
Warren
 
Is your original OPII network interface functioning normally now on v3.13 while connected to your main subnet?
 
Yes and no.  I installed the firmware on 8th of May.
 
I had a spontaneous cold boot happen to me once since update on 10th of July.
 
That bugs me; even though it only happened once.

 
And it was a hard reset that finally fixed your issue permanently?
 
No.  I had odd OPII issues doing a warm 3.13 firmware update until I did a cold reset firmware update.  Many users had similiar issues when updating to FW 3.13 and posted here about said issues.
 
Or is it some network isolation you've done that was your solution?
 
I isolated the network to fix my issues (network, serial and time sync) on firmware 3.12 and it worked fine.
 
My OPII was purchased in about Sept 2013.   How do I tell if I am using the legacy NIC?
 
Your OPII sounds pretty new. 
 
The design change thing was only a guesstimate that there may have been a design change based on PCT88's observations and fix.
 
Try the cold boot firmware V 3.13 update and see if that works for you. 
 
If it doesn't and you are not using an email board then I would just (this is my opinion until you try it yourself).
 
1  - downgrade to firmware 3.12 and isolate the network.  Easy to do. 
or
2 - do a cold FW update to 3.13 (and see if that works).
 
This may be a dumb question....but what is a *cold* FW update.  Yesterday I updated to v3.13 but don't recall choosing the type of update.   Is there some procedure for first wiping the memory?   I can't find anything in manual about *cold* or *hard* reset either.
Thanks.
 
Warren,
 
Doing a cold reset of your panel sets it to defaults.  I think too it resets the IP stuff.  That said you will need to back up your configuration first (very important).  You can also connect to the panel via a serial port and PCA.
 
Steps are: (on your keypad or touchscreen)  I had to look then try it as I never use this feature for anything.
 
1 - setup
2 - misc
3 - installer setup (#)
4 - misc
5 - reset system eeprom
 
Go slow baby steps doing this.  Configure access to your panel.
 
After you do this then also do a warm reset by:
 
1 - setup
2 - misc
3 - installer setup (#)
4 - misc
5 - reset system ram
 
Then upload your firmware
 
Then do another warm reset and check your stuff.
 
The problem went away for me when I put the omni behind it's own router and only opened the ports it needs. I haven't seen any lockups since then.
 
Well, I went through the cold reset procedure and reloaded v3.13 followed by my configuration.  That didn't change anything, however, I'm not entirely sure it truly reset the board to factory defaults.   After both eeprom and ram resets the console always displayed v3.13    I would have expected to see v3.11 after the eeprom was reset, which was in the board when I first received it.
 
Anyway, I did finally have some success by trying *another* old 10/100 switch I found in the basement.  It is a Zonet ZFS3018P 8-port switch I bought from MonoPrice many years ago.    
 
So where I am at is the OPII is back on my primary subnet but with the Zonet switch between it and the router.    I wish understood why this works and the other two 10/100 switches did not.  arrgh!  Knock on wood though, it is dropping only a very occasional packet as of about 10pm last night.
 
There has to be some perfect storm of circumstances that creates this issue, otherwise lots more people would be screaming at HAI wouldn't they?  since most people use their home network for streaming video and other high bandwidth applications?    
 
If things change I will report back.  THANK YOU ALL FOR YOUR HELP AND SUGGESTIONS.  THIS FORUM IS AN INCREDIBLE HELP!
-Warren
 
Warren,
 
Good news; well sort of.
 
Yup if you had v3.13 on the panel then a cold restart would not change this piece.  It would remain at firmware V3.13
 
The "cold" / "warm" resets just relate to a sort of cleansing of old left over stuff from prior firmware versions.
 
Yup with V3.12 firmware the fix utilizing the router worked relating to the issues with the network, time sync and serial comm stuff.  This is my DIY fix relating to my network issues that worked for me.
 
OPII312.jpg
 
With the currently installed V3.13 I have removed the connection to the router and have installed it on the old subnet.  This is the way it should work and historically did work. (except for my one hiccup)
 
OPII313.jpg
 
There has to be some perfect storm of circumstances that creates this issue, otherwise lots more people would be screaming at HAI wouldn't they?  since most people use their home network for streaming video and other high bandwidth applications?
 
Over the years I have added more and more devices to my network.  I did have very occasional / random similiar issues a couple years before.
 
A simple warm reset always would fix the issues.  Only recently the issues would not go away.
 
Here I have upped the number of streaming devices and there is more traffic nowadays on the network way more today than just a couple of years ago.
 
While the main network numbers have gone up; I have also added autonomous networks and VLANs. 
 
That said I am currently stable with firmware V3.13 and connectivity to the main network except for one hiccup.
 
I asked HAI for the settings for the OP II port:
 


Sorry for delay, the Ethernet port on the Omni panels is full duplex.
 
My system auto detect's it at half, should I force the port to full? What's everyone else get for auto negotiation?
 
I always thought it was 10 full duplex until I looked.  Mine is 100Mbs full duplex.
 
Today I am current firmware 3.14a.  Thinking the default on the TP-Link is set to auto-negotiate.
 
Settings for eth0: ===>  TP-Link to HAI OPII NIC
        Supported ports: [ TP MII ]
        Supported link modes:   10baseT/Half 10baseT/Full
                                100baseT/Half 100baseT/Full
        Supported pause frame use: No
        Supports auto-negotiation: Yes
        Advertised link modes:  10baseT/Half 10baseT/Full
                                100baseT/Half 100baseT/Full
        Advertised pause frame use: No
        Advertised auto-negotiation: Yes
        Speed: 100Mb/s
        Duplex: Full
        Port: MII
        PHYAD: 4
        Transceiver: external
        Auto-negotiation: on
        Current message level: 0x000000ff (255)
                               drv probe link timer ifdown ifup rx_err tx_err
        Link detected: yes
 
[   23.710000] device eth1 entered promiscuous mode
[   23.720000] IPv6: ADDRCONF(NETDEV_UP): br-lan: link is not ready
[   23.750000] IPv6: ADDRCONF(NETDEV_UP): eth0: link is not ready
[   24.810000] eth1: link up (1000Mbps/Full duplex) <=== TP-Link to Gb switch
[   24.810000] br-lan: port 1(eth1) entered forwarding state
[   24.820000] br-lan: port 1(eth1) entered forwarding state
[   24.820000] IPv6: ADDRCONF(NETDEV_CHANGE): eth1: link becomes ready
[   24.850000] IPv6: ADDRCONF(NETDEV_CHANGE): br-lan: link becomes ready
[   26.610000] eth0: link up (100Mbps/Full duplex)
[   26.610000] IPv6: ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready
[   26.820000] br-lan: port 1(eth1) entered forwarding state
[  390.610000] eth0: link down
[  393.610000] eth0: link up (100Mbps/Full duplex) <=== HAI OPII NIC to TP-Link
 
 
Back
Top