Within a Cisco voice infrastructure, DNS or (dynamic DNS) can provide name resolution between all telephony nodes and endpoints in the cluster.
DNS in the CUCM Cluster
DNS Implementation Types
Two types of implementations exist for DNS:
Corporate DNS
When the CUCM nodes are configured to use Corporate DNS, the corporate DNS infrastructure is used and default DDNS functionality will act as a cache-only service to this corporate DNS service. Before configuring CUCM to use corporate DNS, ensure that all nodes in the cluster are correctly configured within the DNS prior to beginning installation to make the process as smooth as possible.
Internal DDNS Service
When no corporate DNS service exists, Dynamic Domain Name System (DDNS) service is used to implement a clusterwide DNS infrastructure. This also serves other devices on the network that are interacting with the cluster. Each node (CUCM server) has DNS running on it. These DNS servers get configured with hostname and IP address information for all the nodes and any other devices in the cluster. The DNS on the first node in the cluster gets configured as primary DNS, while all other nodes get configured as secondary nodes.
When any change to DNS configuration occurs to the first node of Cisco Unified Communications Manager, the change automatically gets transferred to other nodes. Other devices in the network can point to any of the nodes in the cluster for the DNS lookups.
Design Recommendations
Using DNS introduces an additional failure point in your telephony system. If the DNS is not available or fails, communication between cluster nodes and endpoints will fail. For networks that require a high level of availability, Cisco recommends that you do not rely on DNS name resolution for any communications between Unified CM and the Unified Communications endpoints.
In some cases, the overall design may require the use of DNS due to the use of NAT or due to the telephony disaster recovery configuration. If DNS must be used, it should be implemented in a geographically diverse manner to limit exposure to outages.
Key Concepts For Review
- Configure Corporate DNS prior to beginning CUCM installation
- If Corporate DNS is not used, the CUCM uses DDNS internally
- CUCM DDNS changes are replicated automatically throughout the cluster
- The first CUCM node in the cluster will be the primary DDNS
- High availability designs should use IP addresses for communication, not DNS
- Some disaster recovery designs will require DNS
- In some cases, the use of NAT in the design will require DNS
DNS Protocol Details
Domain name system (DNS) generally uses UDP port 53 to serve requests. If the size of the response data exceeds 512 bytes, the TCP protocol will be used.
If you'd like to get really dirty, the DNS protocol is defined by a lengthy list of RFCs which can be on this WIkipedia page, very bottom.
Suggested Reading
- CUCM 7.x SRND, Network Infrastructure
- CUCM 7.x System Guide, DHCP & DNS Considerations
- CUCM DNS Configuration Via CLI
- Installing Cisco Unified Communications Manager 7.x, several DNS considerations
- Wikipedia - DNS
- Wikipedia - DDNS
- CUCM SRND 7.x, Full PDF
CCIE Voice Written Blueprint
The relevant section of the blueprint that includes this topic is shown below.
1.00 Infrastructure Protocols
1.01 DNS
1.02 TFTP
1.03 NTP
1.04 Power over Ethernet
1.05 Voice and Data VLAN
1.06 Troubleshooting Infrastructure Protocols
Full Cisco CCIE Voice Written Blueprint
No comments:
Post a Comment