We recently helped a customer in the insurance industry better manage their Disaster Recovery/Replication traffic that was having a negative impact on users trying to access business critical applications. The company has a 50M circuit between their two corporate locations. The larger facility also serves as the gateway to the Internet for both sites, and they have a need to provide nightly back up replication using SnapMirror.
You Might Also Like: 5 Secrets to Make Microsoft Business Apps Faster
We first installed an Exinda Network Orchestrator in monitor mode to help us understand network and application behavior on their WAN circuit. After collecting data for two days, we discovered the issue – the backup replication jobs they initiated every evening were not complete by the time users arrived in the morning. We could see that the DR traffic was still using upwards of 20%-30% of the available bandwidth well into midmorning hours. The DR traffic was causing the contention as users tried to access critical application server resources to perform daily tasks and surf the Internet.
Here’s what we did:
We edited the two default schedule objects in Exinda Network Orchestrator to reflect business hours vs. after work hours. We knew that if we could offer an easy way to put the DR bandwidth usage policies on a schedule, we could change the behavior of the DR traffic based upon the time of day.
Since we automatically classify SnapMirror, we were able to quickly create two different policies based on business hours and after hours. We allocated 90% of the available bandwidth from 7:00pm until 6:00am and 10% of the available 50M bandwidth between 6:01am and 6:59pm.
Below are the policies we created:
We found that by limiting the SnapMirror traffic to a smaller portion of available bandwidth at the start of the business day, users had the network WAN resources they needed to be productive as soon as they got to work, which was a welcome change for some departments, and a not so welcome change for others!
Exinda’s easy to implement optimizer policies tied to a schedule, is the perfect way to have applications treated differently based upon the clock.
It’s a very practical and operationally friendly approach to adapt a WAN resource to a different application personality based on dynamically changing application WAN resource requirements.
It’s one thing to be able to see what’s happening across the network, and another to be able to actually do something about it. With a modern WAN orchestration solution, network managers not only gain visibility into why applications are performing slowly, but they also get advanced control capabilities that allow them to proactively address performance issues such as misconfigurations, congestion and bottlenecks. Organizations should look for a WAN orchestration solution that have sophisticated controls that can create network polices to prioritize critical applications, fair share network resources, throttle recreational traffic or block unwanted traffic altogether.
Need help managing your network resources? Join our weekly live demo to see how Exinda can help!