MBG Heartbeats

[edmicra]

Hi all

Can anyone assist. I have a MBG in server only mode in a DMZ connected to a 3300 MXe11 on 4.1.

To cut a long story short I am getting a lot of delayed or missed heartbeats from remote Teleworker sets. "Mitel sets" from all remote sites. As a result I believe the sets are taken out of service after three missing heartbeats. I can see this in the 3300 logs and in the MBG tug logs.

I want to run wireshark to look for the heartbeats.

Not being a network Engineer I am not sure what I should be looking for. I have had a good look around and I believe the heartbeats are either Minet or SUDP and if so what port ?.

Thanks in advance

PS

MBG site has a  4Mb DSL un-contended link to the Internet.

It peaks at about 3.5 for short periods bit generally sits at 1.5

[wiseguy]

Hi,
I can't believe it has todo with the Keep alive message being not there, but look for a reason why it might be not getting there. Useally the Firewall setup is the part to look at.
When do these logs appear? If this is only during office hours then look at the amount of ports opened on the Firewall. If you want to check this, you can change the setup to Gateway/server setup or place the MBG outside the network before the Firewall.

[edmicra]

Hi Thanks for the advice wiseguy

The MBG logs indicate late heartbeats on a regular basis. Every few minutes as they are from all remote sites I would agree that perhaps the firewall is causing problems however the TNA tool show all ports open and the customer insists they are set according to MBG guidelines.

My understanding is that the heartbeats are sent every 30 seconds and if three are missed the ICP takes the set out of service until it sees heartbeats again. This is confirmed by logs on the 3300.

I would still like to know what to look for when tracing with Wireshark. Minet RTP, SRPT ??

Some logs from the MGB



* Current alarm status: major
* Application Event Type Value Severity Date and Time Description
* MBG set connection status indeterminate Thu Nov 25 12:26:22
* 2010 Acceptor:SSL, ID:08:00:0F:47:F6:7D, Ip:81.18.100.90:27760,
* Icp:172.19.1.10, active-connections:16, LocalIp:10.0.1.100,
* Type:200, dnlist:9456
* MBG set connection lost indeterminate Thu Nov 25 12:24:51 2010
* Acceptor:SSL, ID:08:00:0F:4B:AE:40, Ip:81.18.100.90:9137,
* Icp:172.19.1.10, active-connections:15, Reason: IS read failed -
* connection closed
* MBG set connection established indeterminate Thu Nov 25
* 12:24:51 2010 Acceptor:SSL, ID:08:00:0F:4B:AE:40,
* Ip:81.18.100.90:8292, Icp:172.19.1.10, active-connections:16,
* LocalIp:10.0.1.105, Type:201
* MBG DN event 08:00:0F:4B:AE:40 indeterminate Thu Nov 25
* 12:24:51 2010 Acceptor:SSL, ID:08:00:0F:4B:AE:40,
* Ip:81.18.100.90:8292, Icp:172.19.1.10, dnlist:9452
* MBG set connection status indeterminate Thu Nov 25 12:21:59
* 2010 Acceptor:SSL, ID:08:00:0F:47:F3:9F, Ip:81.18.100.90:36608,
* Icp:172.19.1.10, active-connections:16, LocalIp:10.0.1.102,
* Type:200, dnlist:9455
* MBG set connection status indeterminate Thu Nov 25 12:21:43
* 2010 Acceptor:SSL, ID:08:00:0F:47:F6:A5, Ip:122.55.96.154:38051,
* Icp:172.19.1.10, active-connections:16, LocalIp:10.1.1.133,
* Type:200, dnlist:9458
* MBG set connection status indeterminate Thu Nov 25 12:18:41
* 2010 Acceptor:SSL, ID:08:00:0F:47:FE:A5, Ip:122.55.96.154:46729,
* Icp:172.19.1.10, active-connections:16, LocalIp:10.1.1.76, Type:200,
* dnlist:9457
* MBG set connection status indeterminate Thu Nov 25 12:13:53
* 2010 Acceptor:SSL, ID:08:00:0F:4B:99:49, Ip:81.18.100.90:41530,
* Icp:172.19.1.10, active-connections:16, LocalIp:10.0.1.101,
* Type:201, dnlist:9450

[wiseguy]

Look at the attachement. I think the heartbeat is every 60 seconds and can be seen at 45, 105 and 165 seconds.

[Chakara]

  Either way, what is the symptom on the remote phone.  Does it "disconnect" or are we chasing phantoms?

-Chak

[edmicra]

Wiseguy thank you.

I have checked the 3300 and the documentation and there is a setting in the Controller Registry Configuration form called Link Maintenance - Heartbeat Rx Link Failure Time-out. This is set to the default setting of 300 or 30 seconds. I am reluctant to change this timer

What I can see is in the 3300 logs the teleworker sets being returned to service. Ie device detection logs as well as the MBG logs previously posted. These all seem to point to late or missing heartbeats and sets being taken out of and returned to service quite frequently. The MBG is in constant major alarm with loads of heartbeat errors.

logfiles in server manager, select tug/tug.log, in the filter section enter

[ralph]

I've shot myself in the foot before with this one.
I was screwing around with the Controller Registry Configuration and changed some timers.
The resulting symptom was the TW phones would reboot mid conversation.

At the time I blamed everybody but me for screwing up the network.

I resolved it my going into the Controller Registry Configuration, opening up the help files and setting everything back to default.

Ralph