Many enterprises have adopted cloud-based services over the last decade. Cloud computing allows these organizations to run their high-traffic, high-performance applications on low-cost IT infrastructure, taking advantage of server virtualization and the scalability of today’s data centers. This helps enterprises improve their operational efficiency while keeping costs down. There has been ample focus on monitoring and managing cloud applications, systems, storage and security. However, most enterprises do not monitor IP routing to the cloud, leaving them susceptible to cloud connectivity issues.
Because cloud applications reside outside the actual enterprise premises, IP routing between the enterprise and the cloud infrastructure over public IP networks is essential.
To ensure high availability, enterprises have redundant links between the different nodes in their networks as well as to their multiple service providers. This creates numerous paths between one node to another, either within the network or outside. But the dynamic nature of IP routing, where routing paths are automatically chosen in milliseconds based on the present state of the network, makes it extremely difficult for network engineers to understand what is happening in the network.
Things get even more complex when traffic exits the enterprise network and traverses the Internet. With Internet routing, traffic can pass through different Autonomous Systems, thousands of network addresses, and myriad paths before it reaches its destination. Here, numerous factors such as changes in routing paths, BGP peering failure, route leaks, etc., can result in slow application response or even complete data loss.
Considering these factors, network engineers should be able to understand IP routing behavior to quickly troubleshoot and resolve application delivery issues. You can read about even more reasons to monitor IP routing from our blog here.
Most enterprise networks are highly dependent on SNMP for network performance monitoring. Network management tools that leverage SNMP are designed to poll network devices at set intervals and gather information on device health, up/down status, CPU and memory utilization, packet drops, traffic information, etc. But SNMP-based tools that poll every few minutes are not capable of capturing IP routing changes that can occur within milliseconds anywhere in the network or result in a high volume of network overhead.
Another requirement is the need to plan network changes. A majority of enterprise network change management tools are only focused on ensuring correct syntax or comparing configurations. They do not have the capability to understand the impact of a network change before the actual change has been made. In the world of IP routing, it is possible to make the “right” configuration changes on a device but still end up with “wrong” routing behavior in the network.
Route analytics technology listens to BGP and IGP route advertisements exchanged between routers to build an always up-to-date network topology map. The information in the routing updates also helps determine how traffic is being routed over the network in real time, even when the routing path changes within milliseconds. Additionally, route analytics allows the routing information captured to be stored. It can then be “rewound and played back” by network engineers to understand how routing paths changed in the past.
This information helps engineers understand if subnets hosting the cloud apps were reachable at all times, if paths between the data center and the enterprise were up, how traffic is being routed, whether or not it is traversing high-latency links, and if redundant paths exist to the cloud network. This way, enterprises can not only quickly troubleshoot reachability issues but also keep their service providers accountable.
Route analytics combined with traffic flow data can help network engineers plan for network changes. Using such network planning tools, network engineers can understand if the existing network can handle additional bandwidth load from cloud applications, if any congestion will occur, and the impact of a planned change on the overall network performance.
Our white paper – “Restoring Visibility into the Cloud with Route Analytics” – provides even more details on how route analytics can provide visibility into IP routing to the cloud, along with many network monitoring best practices.