Page 1 of 2 12 LastLast
Results 1 to 13 of 14
  1. #1
    Join Date
    Sep 2007
    Location
    Kenilworth NJ
    Posts
    1,580
    Post Likes

    LON nvi's not updating

    I have a large Distech LON network that is 95% LON links doing Trim and respond, occupancy, and so on.
    I am new to LON, getting throwed in the deep end.
    There are nvi bindings for occupancy that regularly do not seem to communicate. The source device is sending occOccupied and the receiving device says occNull. The points are not overridden, but if I command them to Auto, they do pick up what they are supposed to be doing... for 5 to 30 min, and fall back to their default again.
    I heard someone say that the parameter is on the network variable point, in GFX, in the max read and min read times.
    There are other nvi points that are also doing this, but I also heard nviOccupancy is more often a culprit for some reason.
    I have been reading up on this but I need a hand here. I know what these parameters are on paper, but I would like to understand them better, so I can read what is happening better.
    Thanks guys.
    Hmmmm....smells like numbatwo to me.

  2. #2
    Join Date
    Sep 2013
    Location
    Chicago
    Posts
    46
    Post Likes
    Sounds like a Lon link binding issue, but I'm not familiar with Distechs programming so someone might be able to help on that end.

    Sent from my moto z3 using Tapatalk

  3. #3
    Join Date
    Sep 2007
    Location
    Kenilworth NJ
    Posts
    1,580
    Post Likes
    Thread Starter
    Quote Originally Posted by poptop22 View Post
    Sounds like a Lon link binding issue,
    The bindings are made, and many work. But the values do not carry through. How would you make this happen in your particular programming system? Lay it out and I will try to apply it to Distech.
    It sounds like the Max Read Time on the nvi needs to be more than the max write time on the matching nvo? Am I on the right track here?
    I am looking for an APDUtimeout or some other BacNet style parameter and I am not finding one... because it is LON.
    The more I work with LON the more I respect it - but it is still fundamentally foreign to me for now. Me and LON need to have a few beers an get to know each other better....
    Last edited by numbawunfela; 11-23-2019 at 01:15 PM.
    Hmmmm....smells like numbatwo to me.

  4. #4
    Join Date
    Jun 2012
    Posts
    328
    Post Likes
    Quote Originally Posted by numbawunfela View Post
    The bindings are made, and many work. But the values do not carry through. How would you make this happen in your particular programming system? Lay it out and I will try to apply it to Distech.
    It sounds like the Max Read Time on the nvi needs to be more than the max write time on the matching nvo? Am I on the right track here?
    I am looking for an APDUtimeout or some other BacNet style parameter and I am not finding one... because it is LON.
    The more I work with LON the more I respect it - but it is still fundamentally foreign to me for now. Me and LON need to have a few beers an get to know each other better....
    Are these GFX or some older legacy controllers? If they are GFX, never ever duplicate a controller. Your lon bindings and com sensors will not work correctly. If you have duplicated devicesg, I suggest saving the gfx file and starting over.

    If you are using a BCPLonNetwork from the Distech pallet, the points will use the BCPPolicy for the tuning policy. You can change this on individual points. The BcpPolicy does not allow a max write to allow values to be resent if unchanged past a certain time. Inside of GFX there is a max receive time function for the NVIs and a max send time for NVOs

  5. #5
    Join Date
    Sep 2007
    Location
    Kenilworth NJ
    Posts
    1,580
    Post Likes
    Thread Starter
    Quote Originally Posted by JSLLC4Life View Post
    Are these GFX or some older legacy controllers? If they are GFX, never ever duplicate a controller. Your lon bindings and com sensors will not work correctly. If you have duplicated devicesg, I suggest saving the gfx file and starting over.

    If you are using a BCPLonNetwork from the Distech pallet, the points will use the BCPPolicy for the tuning policy. You can change this on individual points. The BcpPolicy does not allow a max write to allow values to be resent if unchanged past a certain time. Inside of GFX there is a max receive time function for the NVIs and a max send time for NVOs
    These are GFX. All ECL controllers.
    Thanks for the tip on Duplication, I would have totally done that. I do not know if these were done that way... the station was done in 2013. The contractor walked away, so it seems like it worked. Then another contractor was having their low level guys do PMs on the site... When I got here, the AX station was on a JACE8000 - rebooting pretty much every day or 2. Turns out that someone after the original install had used program services to add histories until there are over 4,500 of them. (if memory serves, the AX File Descriptor Limit is around 1600). When the thing locked up they walked away and it was that way for about a year+ when I came around... so no telling what else happened to this thing. I just upgraded the Station to N4 and it runs like a champ - even with all those histories. Has 163 controllers total.
    These are on a BcpLonNetwork. these links are being made directly from LON object to LON object. So the proxy point side of this is completely bypassed (for the purposes of answering the question as to what Tuning policy is on the points - no Tuning policy on LON objects that I can find).
    Some( but not all) of these points that are passing data were mapped as proxy points, but not linked from these. I mapped the rest as Proxy points, so I could easily get them into a component grid and watch them to see which ones were not working.
    Name:  LonOut.png
Views: 96
Size:  26.2 KBName:  LonIn.png
Views: 97
Size:  49.7 KB
    Here you can see the LON objects with the green links - comes out of one with a value, and the other side is not seeing it.
    CPU usage is only 30-40%. LON Trunk is about 60% loaded.
    I had someone say that the links need to have matching Nv Config Data Selectors on both ends, and if they do not match, then I need to break the connection and re-make it. I have not looked at very many connections, but I do know that the few broken ones I have looked at have numbers here that do not match:
    Name:  LONLinkNvSelector.jpg
Views: 99
Size:  87.4 KB
    The question for me is - how did this link get made without a matching selector? Is this possibly what happens when you duplicate the controller? Is this something I need to worry about? They might not match on connections that work properly as well, I have not looked, and I have not tried this fix just yet.
    Hmmmm....smells like numbatwo to me.

  6. #6
    Join Date
    Sep 2004
    Location
    SoCal
    Posts
    484
    Post Likes
    Could be in nv preemption mode. try either resetting the device or deleting and then re-doing the binding.
    A hundred million nodes - it's a LON story.

  7. #7
    Join Date
    Sep 2007
    Location
    Kenilworth NJ
    Posts
    1,580
    Post Likes
    Thread Starter
    Update - turns out when I look at these links in the LON link manager that a large number of them come up as new links, and not bound. So I been binding these as fast as I can. I do not know whether this is helping anything just yet. I will not know until next week sometime.
    Hitting Selective Bind make the workbench hang, often long enough to kick me out of the JACE which does wonders for my productivity. I remember there was a setting in the guts of a file in a folder Workbench reads that extended the Fox Connection timeout. I just do not remember where that is...
    Hmmmm....smells like numbatwo to me.

  8. #8
    Join Date
    Sep 2004
    Location
    SoCal
    Posts
    484
    Post Likes
    LNS

  9. #9
    Join Date
    Sep 2007
    Location
    Kenilworth NJ
    Posts
    1,580
    Post Likes
    Thread Starter
    Quote Originally Posted by lonboy View Post
    LNS
    Uh, you mean I should go get a copy of LNS, and use it to make the binds? Or the Niagara file that has the Fox Connection timeout setting is called LNS?
    I was of the impression that the LON links need to be made in Niagara, so that there is a record of who talks to what in the same system used to manage the network. There CAN be an LNS database used to make the system talk, but then it adds a layer of confusion as that LNS database is not what you inherently have on the site, it is not the JACE. you always need to go digging around to find it, open it, and so on. More 'possible but not ideal'. I suppose one could bind in LNS and then quick learn the resulting connections into Niagara...
    Hmmmm....smells like numbatwo to me.

  10. #10
    Join Date
    Sep 2004
    Location
    SoCal
    Posts
    484
    Post Likes
    Use LNS based tool to make binds, commission devices, manage devices, create straightforward multi-vendor control networks based on LonMark standards. We run our LNS tool in the AWS cloud and only need to provide an IP connected RNI onsite. Dump the JACE!
    A hundred million nodes - it's a LON story.

  11. #11
    Join Date
    Sep 2007
    Location
    Kenilworth NJ
    Posts
    1,580
    Post Likes
    Thread Starter
    Quote Originally Posted by lonboy View Post
    Use LNS based tool to make binds, commission devices, manage devices, create straightforward multi-vendor control networks based on LonMark standards. We run our LNS tool in the AWS cloud and only need to provide an IP connected RNI onsite. Dump the JACE!
    RNI?
    Are you offering to bund my network for me?
    The site is 30 JACEs, a Websup, Envysion graphics and so on. Cannot really ditch the JACE altogether. Perhaps you mean bind the thing using LNS under the JACE?
    Hmmmm....smells like numbatwo to me.

  12. #12
    Join Date
    Sep 2004
    Location
    SoCal
    Posts
    484
    Post Likes
    Quote Originally Posted by numbawunfela View Post
    RNI?
    Are you offering to bund my network for me?
    The site is 30 JACEs, a Websup, Envysion graphics and so on. Cannot really ditch the JACE altogether. Perhaps you mean bind the thing using LNS under the JACE?
    You can do your own bunding.
    30 JACE's ? wow that's job security! How many devices? Must be a campus.

    Lon devices don't care what OS you're using whether Tridium or LNS. They simply share information in a unicast, multicast, or broadcast across some media FT-10 twisted pair being the most popular and robust.

    The details of the network variable messaging are handled well by LNS tools. The LNS tool will automagically assign the device or group ID's as well as update the device NV tables with selector ID's. These cannot be easily manually changed but is clearly exposed to the integrator.

    The message service type (ack, unack, repeated, authenticated) is also assigned but can be changed if needed.

    The main challenge you may have by doing LNS binds under the JACE is that domain/subnet/node and group ID's may change when you commission devices with an LNS tool. I suppose tho that you might be able to use NodeUtil to modify these if needed. Or, like you said simply use the JACE to discover the points after the fact.

    BTW, I just worked on an LNS site (campus) that includes 30 buildings, 30 LNS databases all hosted in one VM, 50 IP-852 routers, and over 50,000 data points all served up to a WonderWare front end using LNS DDE!

    LON LIVES ON!
    A hundred million nodes - it's a LON story.

  13. #13
    Join Date
    Sep 2007
    Location
    Kenilworth NJ
    Posts
    1,580
    Post Likes
    Thread Starter
    Quote Originally Posted by lonboy View Post
    LON LIVES ON!
    I was talking to Distech, and told them that the LON support provide varies greatly in quality... (hint: maybe some of your guys need a refresher) the response was 'Be glad we support it' in a nice way, of course. They are removing it from their trainings...
    I am happy to do it as then it means I have job security as the guy who can work on LON. That is, once I can actually work on LON.
    I am actually liking it, I have a new appreciation for it for sure.
    I REALLY appreciate your responses. Super helpful.
    I may need to go the LNS route. The devices all have Subnet and nodes assigned. There are 163 devices and some devices have over 100 links on them, I am really trying to not have to entirely rebuild the links if possible. Niagara knows who needs to talk to who, so I would really like to get Niagara to inform the devices if possible. If possible....
    But if I need to start from scratch, I will.
    If I remember right, LNS is free to download, but needs a license? Also would need a lon adapter to get my laptop running LNS onto the FT10 network? Can you point me in the right direction on that maybe?
    Hmmmm....smells like numbatwo to me.

Page 1 of 2 12 LastLast

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •