This is Part 2 in a series about cloud-based disaster recovery. Read Part 1 to learn more about why enterprises are moving their DR sites to the cloud.
In my previous post, I outlined five key reasons why enterprises are migrating their secondary data center (aka DR site) to the cloud.
An effective DR strategy, however, requires more than setting up a secondary datacenter in the cloud. To enjoy the benefits of cloud-based DR, enterprises need to implement the right tools and practices to ensure seamless operations and user productivity in the event of an outage.
In this post I want to focus on some best practices for a successful cloud-based DR implementation, including testing, overcoming latency, instant failover and quick recovery from a local disaster.
Testing Makes Perfect
Since for cost reasons, the cloud datacenter is kept “down” most of the time, companies adopting cloud-based DR are advised to perform frequent DR tests and drills to ensure that the DR site will work when disaster strikes.
Automation is critical for making these tests as cost-efficient and painless as possible. Enterprises need automated testing solutions that activate the DR site, execute failover, move users, test the application functionality, record the results, and then shut down the DR site. Speed is of the essence here since public cloud usage costs are billed on an hourly basis.
Overcoming the Latency Obstacle
Due to the distance between the primary and DR sites, latency is typically an issue when enterprises are forced to work in DR mode. File services are a good example – accessing a document and saving a file requires millisecond-level latency from a user productivity standpoint.
CTERA’s solution, for example, keeps a cached copy of critical data at the edge. Our edge filers utilize smart caching technology to deliver fast local access to a cloud-based global file system. Frequently used or critical files are cached locally to ensure optimal edge performance, while file changes are replicated to the cloud.
By deploying CTERA Edge Filers at your branch offices – close to your users – your enterprise can maintain high performance while increasing disaster readiness. With an edge filer, your users are likely not to even notice that the company has gone to DR mode – as the local filer masks the difference in latency, delivering predictable low latency for all frequently accessed files.
Zero-Minute Disaster Recovery
CTERA Edge Filers enable enterprises to instantly failover remote office file servers to the cloud in the event of an outage, ensuring uninterrupted office operations and user productivity. Seamless ACL enforcement on both the local gateway and in the cloud ensure that users experience the same file access rights in cloud failover scenarios. With no need for expensive, redundant failover or replacement hardware, organizations can save thousands of dollars per office.
Quick Recovery from Local Disasters
Edge filers are particularly advantageous in local disaster scenarios. Unlike a traditional file server, where restoring all the files from the cloud can take days, should the local edge filer become damaged and suffer a major data loss, a fresh edge filer can be fired up to restore service in an instant. Here’s how it works:
- Step 1 – Following a major failure involving loss of a local filer, a fresh, empty edge filer is spun up as a virtual machine or cloud instance. The filer then repopulates itself with metadata for the files (the actual mass of the data is not downloaded), and within a couple of minutes the filer is ready to serve users.
- Step 2 – Users start to work with the new edge filer. The first time a file is accessed there may be some latency, but this quickly disappears over time as the cache is warmed up.
DR as a Springboard for Full Cloud Migration
As noted in my previous post , implementing DR in the cloud can be a powerful catalyst towards full IT migration to the cloud. Once your cloud DR site is operable and well-practiced, the subsequent steps are relatively simple:
- Convert your cloud datacenter into the primary datacenter. Since you’re already periodically performing DR drills, shifting the workloads to the cloud can be done instantly, with minimal risk and effort.
- Deploy another datacenter in the cloud for DR purposes in a different availability zone or even better, using a different cloud provider altogether. This multi-cloud approach is the “holy grail” of business continuity, providing the highest levels of redundancy and business flexibility in terms of reducing costs and avoiding vendor lock-in.
- Once you’ve tested your new secondary datacenter, the last step is to shut down your physical datacenter and operate your entire infrastructure from the cloud.
So there you have it. By deploying a secondary datacenter in the cloud, together with Edge Filers supporting smart caching at your branch offices, you’ll benefit from cost-effective DR with minimal to zero disruption to users in a disaster scenario.
For more information about how CTERA can help your organization deploy DR in the cloud, please contact us.