Master to Master - can only see Device 0?
true
Posts: 307
in AMX Hardware
Hey all.
Trying to connect two masters and having some issues. I've done a number of other similar units just fine, but this one is causing issues.
Master 1 is an NI-3100, System 1. URL for Master 2 is on this NI.
Master 2 is an NI-900, System 2. No URLs defined.
The masters are on their own network and can see eachother fine, and connect. However, only device 0 on the remote master is visible to each NI. The NIs can talk to their own devices, but not to eachother - in code or via telnet.
Below is with no code running.
EDIT: Tried an NI-3000 in place of the NI-900, and it does the same thing. I'm thinking it is some setting or bug with the NI-3100, but don't know what it could be...
Trying to connect two masters and having some issues. I've done a number of other similar units just fine, but this one is causing issues.
Master 1 is an NI-3100, System 1. URL for Master 2 is on this NI.
Master 2 is an NI-900, System 2. No URLs defined.
The masters are on their own network and can see eachother fine, and connect. However, only device 0 on the remote master is visible to each NI. The NIs can talk to their own devices, but not to eachother - in code or via telnet.
Below is with no code running.
Master 1: >show system /min Local devices for system #1 (This System) ---------------------------------------------------------------------------- Device (ID)Model (ID)Mfg FWID Version 00000 (00299)NI Master (00001)AMX Corp. 00380 v3.30.371 05001 (00286)NI-3100 (00001)AMX Corp. 00383 v1.13.6 10001 (00323)MVP 8400 (00001)AMX Corp. 00592 v2.81.21 Local devices for system #2 ---------------------------------------------------------------------------- Device (ID)Model (ID)Mfg FWID Version 00000 (00285)NI Master (00001)AMX Corp. 00296 v3.30.371
Master 2: >show system /min Local devices for system #1 ---------------------------------------------------------------------------- Device (ID)Model (ID)Mfg FWID Version 00000 (00299)NI Master (00001)AMX Corp. 00380 v3.30.371 Local devices for system #2 (This System) ---------------------------------------------------------------------------- Device (ID)Model (ID)Mfg FWID Version 00000 (00312)NI Master (00001)AMX Corp. 00341 v3.30.371 05001 (00286)NI-900 (00001)AMX Corp. 00382 v1.12.140AMX tech support basically told me that it is a "known bug" with no possible workaround. I have recently installed a number of these systems using an NI-3100/NI-900 combination without a problem though... what gives? Any advice?
EDIT: Tried an NI-3000 in place of the NI-900, and it does the same thing. I'm thinking it is some setting or bug with the NI-3100, but don't know what it could be...
0
Comments
It sounds like a networking problem to me. What do you mean by the masters are on their own network? What are the IP configurations (IP/subnet mask/gateway) for each master? What do you mean by device 0?
Paul
10.0.110.0/24
NI-3100 is .10
NI-900 is .11
WAP-250G is .240
MVP-8400 is .51
notebook is .250.
Gateway, when unset, gave UDP errors; set to master NI (no communications are out of the subnet anyway). Wired connections are through unmanaged 5-port Linksys workgroup switch, which is how the NIs are. Just for giggles, with only the NIs hooked up directly to each other, the problem persists. IP settings are the same as what my other working units are using.
Exactly what I said - if you look at the system information in my post, each master can only see device 0 (NI Master) on the networked master. Specifically, they can't see device 5001 - that's all I need
IIRC having nothing or incorrect data in the gateway field will cause problems although I don't know exactly what they are or why this happens. You may need to experiment with the gateway address, but making it the IP address of the other master I don't think will work. You should be able to leave them blank but I don't think the NIs like that so you many need to try a few things. You might try 127.0.0.1, or the APs IP address.
Paul
The gateway is not used if you are communicating within the same subnet, so technically, it shouldn't matter.
I don't see how this would break for these two specific NIs anyway, when it works fine for dozens of others.
Exactly. As stated above, this should only cause problems with talking to the outside world, or with AMX's broadcast implementation. Since I'm not even dealing with broadcast traffic, it shouldn't make any difference.
Yet again, it shouldn't matter, since I'm not talking to the outside world.
With all that said, I hooked up the NI-3000 in the NI-3100's place and still have the same issue. I can play with the gateway all day and it makes no difference - same IP as NI, IP of other NI, IP of anything else, 127.0.0.1, 0.0.0.0 - doesn't matter. Even directly connected I have the issue. I can talk to both NIs fine over ethernet, and they do this with a direct cable connection, so I am rather stumped.
Fine, but I didn't know that you knew that so I was just asking to make sure.
Why are you using an IP address in the Class A range but then a Class C subnet mask? Change you subnet masks to 255.0.0.0.
Paul
Thanks to subnets, we can partition off pieces of networks as we wish - since 10.x.x.x is a class A _private_ network, we have that entire range, although via subnets, we can partition it as we wish. RFC1918 declares that we can do with these ranges as we wish.
That being said, even with many individual successful installs at this complex using this subnet, I changed IPs on these NIs. No change to the system. I changed the subnet but kept the same IPs. No change to the system. This isn't an IP / networking problem. The systems can _see eachother_ fine, and can even _connect to eachother_, they just don't see any device other than device 0. I can have one ping the other all day.
Right now, I have two NIs hooked up, running no code, via one 200ft tested patch cable and I am still experiencing this issue. I'm about to just defenestrate this equipment and get new gear.
I have done many M2M installs and have never had any problems seeing the other masters devices. I will be very surprised if it ends up being a hardware/firmware issue.
Paul
The point I was trying to make is that it is perfectly valid and legal to use the IPs I am using - whether they are public, private, or anything else, the equipment doesn't care - as long as the equipment is in the same specified subnet (which I am obviously using), it should work.
(By using a broader subnet, other equipment in the 10.0.0.0/8 range could talk to my equipment, which I don't necessarily want. Subnets - more more appropriately, CIDR, which replaces Classes, are a solution. The numbers in the IP don't matter, as long as they are valid and won't cause conflicts - by using a range in 10.0.0.0/8, I know that if this network ever goes online, it won't have any conflicts. And, since 10.0.0.0/8 is a non-routable, local range, I can arrange any address in that range into as many subnets that I wish. You can do the same with 192.168.0.0/16 - in fact, many people do, as 192.168.1.0/24 or 192.168.0.0/24! For a reference, check http://en.wikipedia.org/wiki/Classless_Inter-Domain_Routing)
I could use 1.1.1.1 and 1.1.1.2 in 255.255.255.252 and it'd still communicate as it is now (although not work as it should be, as it is now). And I did try other IPs, 192.168.1.1 and 192.168.1.2 on /24, and the problem still exists.
This is _not_ a problem with the IP addressing. Give me examples and I'll set them and it still won't work. If there are any other commands, features, bugs, options, or whatever that could cause this to happen, it'd be helpful to know.
I have as well, and this is the first time I've had this issue, which is why I am asking here. All of the other installs I have completed in this building - or elsewhere, for that matter - haven't had any issues.
(PS - I've since moved the NI-900 to within 2ft of the NI-3100 with a factory 5ft crossover patch cable. The problem still exists.)
EDIT: If it helps at all, here is a copy from the serial terminal of show system, with full verbosity:
If you send an IR Pulse from one master to an IR port on the other you aren't seeing a red IR led flash?
Are you defining both devices on both systems?
IE:
Master 1 system 1
DEFINE_DEVICE
dvOtherMaster = 5001:1:2
Master 2 system 2
dvOtherMaster = 5001:1:1
I am not certain that the other masters 5001 address will show up in the online tree of the other if you only have the URL listed and no device in the define_device section of code. I have never encountered this issue so I simply have never looked to see what shows up in the online tree during a M2M install.
Paul
In code, yes. (Except for System 2, which currently runs no code - no installations at this site do (yet, they're used for port extension), and there are at least 10 that just have an empty program installed). But as I said, I can't even do this from telnet. For example, running this code on System 1 will flash the first serial port transmit LED, but from System 2 the first serial port transmit LED on System 1 will not light up:
send_string 5001:1:1, 'test'
Of this, I am not certain - I am not running code now to diagnose. I believe it does, but don't have working systems available to me to verify. I'll load up some test code that has these definitions. But I first noticed this issue when I _was_ running code, so...
I put code on both with those definitions. No change.
As for the online tree, I can't really see any good reason why it would show you the 5001 address of the other master. As far as its concerned its just another device like an IR port, correct?
Paul
It has always been my experience that master to master only works when you have the remote device defined in the code of the master to which you are connected (in the case of telnet). So, if I have masters #1 and #2 and I want to turn on the first IR port on master #2 from master # 1, I must have the IR port on master #2 defined in master #1 code. So, if I do:
on[5001:9:2,9] in a telnet session on master #1, it will not turn on the IR port on system #2 unless that device (5001:9:2) is defined on system #1 in the define_device section.
Try populating the define_device sections of the two masters with the devices on the masters that you wish to control via M2M and see what happens.
But, in Netlinx Studio, a full list of devices gets displayed using the online tree and display/refresh network whether or not the remote devices are defined locally. Axcent units are a bit different, I think. My recollection is that an Axcent3 is not totally happy unless all devices on the Axlink are properly defined in the program. Doesn't the Axcent Axlink light blink oddly if not all devices are defined? Of course, there is no M2M and remote masters with Axcents as masters.
I did populate define_device, and it still wasn't working... swapped gear, and of course everything worked on the new gear. Put the old gear back on - first the NI3100, then the NI900, and it's working now??
I don't know if it was just me or if I was going crazy or if the gear was going crazy, but it's working as it should now. I'd really like to know what was causing this, though...
Thanks, all.
Sure, sure...I believe you. Millions wouldn't, but I do...
Sounds like you maybe had a quirky connection somewhere. Glad to hear you got it working though, I knew it would be something silly, as M2M is pretty straightforward. Sometimes hitting the controller with your fist works for these problems too
Paul