Hi,
I have a situation where the IP Trunk between 2 controllers keeps dropping. It stays down for between 30s and 5 minutes, then comes back.
The error in the logs say
Source - Heartbeat Server - ICP has lost contact with (10.78.100.2:3163), network driver cluster pool free: 2980 and low water mark: 2831
Then
Source - STSP - IP trunk to 10.78.100.2:3163 has failed.
I then get the major alarm for the ICP Comms Card.
After a random amount of time the link comes back.
The user hasn't reported any issues, yet.
The 3300s are running 10.2.0.26_2. There are 8 3300s in the cluster and 2 of these are getting this error, 1 a lot more often than the other (once per day as opposed to 5-10 times per day). They are connected to the main site via a BT MPLS network and each of the far end sites has an IP trunk connection to 2 3300s. Only one of these is causing an error.
So, site A has IP trunks to sites 1 & 2 - only site 1 is dropping out, site 2 has never dropped.
Site B also has IP trunks to sites 1 & 2 - again, only site 1 is dropping out, site 2 has never dropped.
Sites 1 & 2 have IP trunks to the other 4 sites and none of these drop out.
The logs above are taken from site A which is the far end site - site 1 is the at the main site.
BT have some sort of class map at both sites where EF traffic is prioritised. They also prioritise AF4 Citrix traffic.
The State Xnet All command comes back with Established (I haven't been logged in when the link has failed).
Weirdly, the 'Rmess Verify PBX 120 LAN' command is successful from sites 1 & 2 to A & B, and is successful from A & B to site 2, but times out from A & B to site 1.
I have tried changing the DSCP value to 46 and L2 Priority to 6 for the Voice Signalling in the LAN Policy (QoS) form in the hope that BT might not be assigning DSCP 26 to an AF queue, so it's going with the default QoS queue - the same as internet traffic. This doesn't seem to have worked.
Is there anything else I can try or ask BT to check or change?
Many thanks
M
I have a situation where the IP Trunk between 2 controllers keeps dropping. It stays down for between 30s and 5 minutes, then comes back.
The error in the logs say
Source - Heartbeat Server - ICP has lost contact with (10.78.100.2:3163), network driver cluster pool free: 2980 and low water mark: 2831
Then
Source - STSP - IP trunk to 10.78.100.2:3163 has failed.
I then get the major alarm for the ICP Comms Card.
After a random amount of time the link comes back.
The user hasn't reported any issues, yet.
The 3300s are running 10.2.0.26_2. There are 8 3300s in the cluster and 2 of these are getting this error, 1 a lot more often than the other (once per day as opposed to 5-10 times per day). They are connected to the main site via a BT MPLS network and each of the far end sites has an IP trunk connection to 2 3300s. Only one of these is causing an error.
So, site A has IP trunks to sites 1 & 2 - only site 1 is dropping out, site 2 has never dropped.
Site B also has IP trunks to sites 1 & 2 - again, only site 1 is dropping out, site 2 has never dropped.
Sites 1 & 2 have IP trunks to the other 4 sites and none of these drop out.
The logs above are taken from site A which is the far end site - site 1 is the at the main site.
BT have some sort of class map at both sites where EF traffic is prioritised. They also prioritise AF4 Citrix traffic.
The State Xnet All command comes back with Established (I haven't been logged in when the link has failed).
Weirdly, the 'Rmess Verify PBX 120 LAN' command is successful from sites 1 & 2 to A & B, and is successful from A & B to site 2, but times out from A & B to site 1.
I have tried changing the DSCP value to 46 and L2 Priority to 6 for the Voice Signalling in the LAN Policy (QoS) form in the hope that BT might not be assigning DSCP 26 to an AF queue, so it's going with the default QoS queue - the same as internet traffic. This doesn't seem to have worked.
Is there anything else I can try or ask BT to check or change?
Many thanks
M