Hi All,
We have inherited a clustered VMware system from a new clients previous IT provider, with a replication to a DR site.
Currently, the system runs on 5 Vyatta virtual routers and another 6 PFSense firewalls. Pain in the ass to look after, so the client wanted to get some Sophos UTM's in to replace it all, that they can understand. (we are 2nd line support)
Replacing it all is pretty straightforward with the exception of the VMware replication and performing a failover.
The servers being replicated are on 172.16.200.0/24 subnet at site A
The DR site B, is on 172.17.200.0/24
Having not tried this with a Sohos UTM before, whats the most straightforward way to get the routing correct so that users at both sites can still see the servers once they have been failed over. Ideally with minimal administration as we have to train the clients staff to perform a fail-over themselves. There will be a Site to Site VPN link.
Currently the procedure given by old support company is as follows:
The DR capability for COMPANY is based on vSphere replication of the business critical VM’s from SITEA to SITEB, These are all in the 172.16.200.0/24 subnet. In each site there is a VLAN and interface on the router for this subnet, and in the standby site the router interface has and placeholder IP address in a different subnet.
To fail over The router IP addressing for the interface in the primary site is changed to a placeholder address so the 172.16.200.0/24 subnet is removed from the intersite dynamic routing automatically. In the standby site the router interface is re-ipaddressed to match the 172.16.200.0/24 subnet which re-adds it to the dynamic routing table but it’s now based in SITEB. The VM’s are then brought on-line in SITEB where the ESX portgroups match to attach them to the correct VLAN and so they can communicate with both sites.
All external traffic to VM’s in the subnet i.e. Outlook Web Access, Access to the Citrix Netscaler, SSTP VPN is configured on both site firewalls with inbound source NAT so that no matter which firewall the traffic comes in on it returns to the firewall it cam from to avoid dropping on state.
The external DNS needs to be updated as part of the failover to point to the relevant IP’s in the DR site.
Many Thanks,
Gareth
This thread was automatically locked due to age.