Home Forums HAast (High Availability for Asterisk) Configuration & Optimization How to configure IP for each NIC

Viewing 2 posts - 1 through 2 (of 2 total)
  • Author
    Posts
  • Avatar photoCustomer Inquiry
    Participant
    Post count: 203

    I have setup my 2 PBX nodes (each with 2 NIC’s) with IP addresses as follows:

    PBX1:
    – NIC 1 (management): IP 192.168.1.5, subnet 255.255.255.0
    – NIC 2 (VoIP): IP 192.168.1.100, subnet 255.255.255.0

    PBX2:
    – NIC 1 (management): IP 192.168.1.6, subnet 255.255.255.0
    – NIC 2 (VoIP): IP 192.168.1.100, subnet 255.255.255.0

    Traffic never seems to work on the VoIP NICs, and if I bring up and down the Management NICs then sometimes VoIP traffic flows but the nodes can find each other (management IPs). I have told HAAst to control the VoIP IP (192.168.1.100) and I see the IP move between nodes, but it still doesn’t work overall.

    What is going on with HAAst?

    Note from Telium: The same issue you are describing can also cause another symptom, where VoIP packets leaving the node appear to be coming from the wrong IP address or wrong NIC. The root cause is the same however, and the solution (below) is the same

    Avatar photoTelium Support Group
    Participant
    Post count: 265

    The problem you are encountering has nothing to do with HAAst, and everything to do with the basics of networking. You have setup 2 NIC’s in each PC, but they are on the same subnet! That means that Linux has no idea where (out which NIC) to route traffic for your 192.168.1.0/24 subnet. You should not do this as it confuses Linux. This topic is part of what’s known as “multihoming”, and I would suggest you research this topic a bit further before you continue to setup your networks. As well, reread the HAAst installation guide – there’s a bit more information on this topic there. (This isn’t the first time we’ve seen this issue in the support group).

    By shutting down NIC1 on either PC you allow Linux to figure out how to properly route, but then your management NIC is gone so the cluster fails over.

    The solution (for 2 NIC’s per PBX) is to ensure that each NIC is on it’s own subnet. For example:

    Incorrect

    Correct

    For advanced users only: It is possible to use a single subnet for management and VoIP, and move a shared IP address between nodes. However, setup is beyond the scope of included technical support (i.e. not officially supported). If you an advanced user and want to experiment independently, you can learn more here.

Viewing 2 posts - 1 through 2 (of 2 total)
  • You must be logged in to reply to this topic.