Call Manager Failover Cluster not working.

[rbarreto83]

Hi everyone.

We have two CM hosts parcitipating of a cluster. We have already created a DCHP option 150 mapping to the IP addressess of the Primay(Publisher) and Secundary(Subscriber) Call managers and the IP phones also get the two TFTP servers IP addresses on the configuration section of the phone. When we disable the primary CM host, the phones cannot register automatically to the secondary CM(also secundary TFTP server).

Is there any configuration on CM publisher where I we redirect the phone registrations to the secundary CM in case of primary publisher failure?

P.S.: on the phone IPv4 configuration, the section: Alternate TFTP is set to NO. I suppose this configuration comes from the Primary CM(publisher).

Thanks in advance.

Rafael.

 

[Jaime Valencia]

All you've done is provide TFTP redundancy, nothing else, for CUCM redundancy you need to configure the CUCM groups and assign it to phones via device pool.

[rbarreto83]

Hi jaime.

CUCM already configured correctly. When we set the TFTP secundary server IP address on DHCP server, the phone start registering correctly (we disable the primary server, the subscriber). The point is: we do this manually. We would like to redirect the phone registrations to the secondary server automatically since the primary server fails.

Thanks.

Rafael.

[Ayodeji Okanlawon]

You need to configure both cucm in the cucm group assigned to the device pool of the phones. This is where it knows its primary abd backup cucm server. Have you done this?

[rbarreto83]

Hi Ayodeji.

Device Pool is already created.

Regards,

[Ayodeji Okanlawon]

Ok. So do you have a cucm group with two servers in it? Is this cucm group assigned to the device pool? Is this device pool assigned to the phone?

[rbarreto83]

Yes Ayodeji. Both call managers primary and secondary are working properly. Only failover is not working. Maybe there is some configuration problem.

Thank you.

[Ayodeji Okanlawon]

Can you send us a screen shot o your cucm group and device pool?

How are you simulating fail over? Are you stopping cucm service on subscriber?

[rbarreto83]

Failover simulation: we remove the IP address of the primary server TFTP(CM publisher) on the DHCP server and set the TFTP 2 IP adress, so the iphone registration is made automatically on Secondary TFTP 2 (CM subscriber).

I am sorry about the lack of information. I just started my studies with Cisco Voice technologies.

 

[Ayodeji Okanlawon]

First of all that will not simulate fail over. Secondly your cucm group only has one server and you don't even have a secondary server available. Did you observer you only have one server in your cucm group?

The way to test fail over once you have ensured that you have two servers available is to stop the cucm service on the primary cucm (This is the easiest I can suggest to you. Its not the best way to do it though). The best way is to configure access-list to block sccp traffic to the cucm rom the ip phone subnets

[rbarreto83]

I know that won't simulate failover, but I just wanted to be sure that, the secondary server would be reached by the phones if the primary (publisher) was unavailable. So, the phones were able to find the secondary server(subscriber).

So, here you find attached another screen shot with the cluster configuration. There are two servers available.

[Carlo Poggiarelli]

Hi Rafael.

Please browse one of your ip phones pointing your browser to its ip address, click on "Network Configuration" on left menu, take a screeshot of the result page and post it here.

Thanks

Regards

Carlo

[rbarreto83]

Hi Carlo. Please check the configuration below:

I am sorry but the previous admin here set the language to portuguese.

Servidor DHCP		10.103.0.236
Endereço MAC		500604FAFC36
Nome do host		SEP500604fafc36
Nome do domínio		""internal content"
Endereço IP		10.101.48.94
Máscara de sub-rede		255.255.252.0
Servidor TFTP 1		10.103.8.10
Servidor TFTP 2		10.103.8.11
Roteador padrão 1		10.101.48.1
Roteador padrão 2
Roteador padrão 3
Roteador padrão 4
Roteador padrão 5
Servidor DNS 1		10.103.0.227
Servidor DNS 2		10.103.0.235
Servidor DNS 3		8.8.8.8
Servidor DNS 4
Servidor DNS 5
ID da VLAN operacional		4095
ID da VLAN administrativa		148
Unified CM 1		10.103.8.10 Ativo
Unified CM 2
URL de informações		"internal content"
URL de diretórios		"internal content"
URL de mensagens		"internal content"
URL de serviços
DHCP ativado		Sim
Endereço DHCP liberado		Não
TFTP Alternativo		Sim
URL de inatividade
Tempo URL de inatividade		0
URL do servidor proxy
URL de autenticação		http://10.103.8.10:8080/ccmcip/authenticate.jsp
Sincronização automática de porta		Não
Configuração remota da porta do switch		Desativado
Configuração remota da porta do computador		Desativado
Config. porta do switch		Auto negociar
Config. porta computador		Auto negociar
Localização do usuário		Portuguese_Brazil
Localização da rede		Brazil
Fone de ouvido ativado		Sim
Versão localização usuário		8.6.2.1000-1
Versão localização rede
Porta PC desativ.		Não
Alto-falante ativado		Sim
GARP ativado		Não
Capacidade de vídeo ativada		Não
VLAN de voz ativada		Sim
DSCP para controle chamada		CS3
DSCP para configuração		CS3
DSCP para serviços		DEFAULT
Modo de segurança		Não seguro
Acesso à Web ativado		Sim
Span porta do computador		Não
VLAN do PC		148
CDP: Porta do PC		Sim
CDP: Porta do switch		Sim
LLDP-MED: Porta do switch		Sim
LLDP: Porta do PC		Sim
Prioridade da potência LLDP		Desconhecido
ID do ativo LLDP
Modo de endereçamento IP		IPv4 apenas
Controle de modo de preferências IP		V4
Configuração automática de IPv6		Sim
IPv6 Servidor de carregamento
Servidor de registro IPv6
Servidor CAPF IPv6		10.103.8.10
DHCPv6		Sim
Endereço IPv6
Tamanho do prefixo de IPv6		0
Roteador padrão IPv6 1
Servidor DNS IPv6 1
Servidor DNS IPv6 2
TFTP alternativo IPv6		Não
Servidor TFTP 1 de IPv6
Servidor TFTP 2 de IPv6
Endereço IPv6 liberado		Não
Acesso à Web ativado		Sim
Nível do EnergyWise
Domínio do EnergyWise

[Ayodeji Okanlawon]

Have you enabled CUCM services on the .11 server? That server is not showing in your cucm group, implies it is not available as a call processing server? Please use the CUCM serviceability page to check that the cucm service is enabled on that server