topic Re: ClusterXL Sync latency in Firewall and Security Management

ClusterXL Sync latency

Mikel_Aanstoot — Thu, 09 May 2019 09:44:59 GMT

Hi,

Kind of a general question, not too much in detail: We have a 5900 cluster where the active and passive gateway are separated in 2 different server rooms but physically still in the same office. We want to move them to 2 different datacentres in 2 different cities. We'll have our own dark fibre connections between the datacentres and the expected latency between them is around 2-3ms. We want to keep the gateways as a cluster in Active/passive configuration. Is 2-3 ms latency still sufficient for the Sync network between the gateways ?

kind regards,

Mikel

Re: ClusterXL Sync latency

Wolfgang — Thu, 09 May 2019 10:21:05 GMT

That's no problem.

snip from ClusterXL AdminGuide

###################################################################

Synchronizing Clusters on a Wide Area Network
Organizations sometimes need to locate cluster members in geographical locations that are distant from each other. A typical example is a replicated Data Center, whose locations are widely separated for disaster recovery purposes. In such a configuration, it is clearly impractical to use a cross cable for the synchronization network.
The synchronization network can be spread over remote sites, which makes it easier to deploy geographically distributed clustering. There are two limitations to this capability:
1. The synchronization network must guarantee no more than 100ms latency and no more than 5% packet loss.
2. The synchronization network may only include Layer 2 networking devices - switches and hubs. No Layer 3 routers are allowed on the synchronization network, because routers drop Cluster Control Protocol (CCP) packets.

###################################################################

Wolfgang

Re: ClusterXL Sync latency

Jerry — Thu, 09 May 2019 13:43:32 GMT

as long as you provide Layer 2 connectivity between both CCP members (ClusterXL Cluster Members) there would be no problem at all. Latency is least important though, the provisioning of the traffic (layer wise) between those 2 members is vitally important! IF you however do not have L2 capability between sites then ... proper L3 design seem required.