So I was recently tasked to connect to CUCM clusters to enable internal extension dialing between clusters, lets call them Site A and Site B.
By and large this was pretty straight forward, i built a SIP trunk between Cluster A and Cluster B, route patterns with site codes routing to the trunk. this part was successful, and i am able to call extensions on either side and have the call setup complete. The problem i am hitting is the audio routing piece.
Site A and Site B have conflicting internal IP space where the phones and soft phones reside, so Cluster A doesn't know how to get to Site B's phones, and Cluster B doesn't know how to get to Site A's phones, because i cannot introduce routes because of the conflict. My firewall logging is showing me that Site A's phones are trying to use Cluster B as their gateway for Site B extensions (which makes sense).
Grasping at straws is there maybe a way to do this with each site using its local cluster as the "gateway" for calls routing over the trunk?
I also explored the idea of using an H323 gateway on either side, and having the gateway route calls to the other cluster
<site a phone> dials Site B ext. <cluster a routes to h323 gateway> <site a gateway dial peer match, targets cluster b>
But in practice this did not pan out, DNA and "show dialplan number" show the call flow as correct, CCAPI debugging shows the call gets to the gateway, but it never leaves the gateway after it arrives.
I have a case open with TAC to get their assistance, but as there is time pressure, has anyone had to stand up a similar configuration? If not, any idea how i can make this work without the sites phones having to be able to reach the opposing site's call manager?
Thanks for any help you can provide.
... View more
First off let me say i am new here so if this is in the wrong area just let me know for future reference. I am currently troubleshooting some output drops off of a 4948-10G interface facing a 2960G. What i believe is happening is occasional spiking traffic is causing the buffer queue on the interface to be overwhelmed causing drops, as the buffer space by default is not overly large. Usually not a problem on a 4948 but in this case it appears it might be, and the switch is admittedly rather full. However the command i locate after an exhaustive search, hold-queue (value) out, entered from the config interface prompt does not appear to work on a 4948, running 12.2(25)EWA14. So my question would be what the command is on a 4948 to increase the output queue size of a given interface?
... View more