One of the biggest requirements for businesses is to provide always-on, 100% available customer-centric experience. May that be for your favorite eCommerce platforms/websites, banking platforms, retail platforms, social media platforms, or even product development / content management applications. As a result, the challenge for IT organizations to be able to meet this always-on business requirement is protection against the loss of an entire data center. Increasingly, customers have applications spanning across continents (for example, Americas and EMEA), but they need their data to be backed up and restored at the granularity of an individual data center location (e.g. EMEA only). This challenge becomes greater when dealing with modern, distributed, cloud applications built on modern databases, like MongoDB, where the components of a database cluster are distributed across multiple geo-locations. Also, note that with international regulations such as General Data Protection Regulations taking strong foothold in compliance requirements, it is imperative for customers to protect all their data assets local to their region.
For increased resiliency and protection against data center outages or to meet GDPR standards or other compliance standards, MongoDB database clusters can be deployed in geo-distributed topologies in which some of the shard primaries of a cluster are deployed in one geo-location (e.g. Americas) and other shard primaries deployed in another geo-location (e.g. EMEA). Local data is kept on local shards in the distributed cluster by using tag aware sharding. In these geo-distributed configurations shard primaries are located in specific geo-sites and secondaries of each shard are distributed across the geo-sites. While local data is stored locally, all the contents of the sharded cluster is available at both geo-distributed data centers and continues to be operational in the event of loss of any data center.
Data Center Aware Backup and Recovery
With our latest RecoverX 2.5 release, DBAs and IT organizations can now delivers on this challenge of protecting geographically distributed data with support for local backup of geo-distributed modern applications. Thus enabling local data center recovery for faster and more flexible RTO.
Specifically, with RecoverX 2.5, we are introducing data center (DC) aware backup and recovery capability, enabling geo-distributed MongoDB sharded database clusters to be backed up at any of the geo-distributed sites and restored at any of the sites. RecoverX can backup an entire geo distributed cluster at each geo-location even though the shard primaries are distributed across geographies. The ability to perform granular restores, by collection or sub-collection (using Query-able restore or Incremental time-range restore) is available with DC aware backup as well.
In the example below, Shard_CA is located in DC_CA with one of its 2 secondaries located in DC_UK. Shard_UK is located in DC_UK with one of its 2 secondaries located in DC_CA
In this deployment scenario, RecoverX can be deployed in either data center to protect the complete MongoDB cluster (both Shard_CA and Shard_UK) without requiring shared storage across the sites. If RecoverX is deployed in site DC_CA, the secondary node of Shard_UK is used for backing up data from that shard. Likewise, RecoverX deployed in site DC_UK uses the distributed secondary node of Shard_CA for backing up data from that shard. For fail-safe data protection in a complete disaster scenario, RecoverX can can deployed in both data centers to fully protect web scale applications deployed on distributed MongoDB database clusters locally in both data centers.
For additional insight on all of these new features check out our new RecoverX 2.5 product page and register to watch product webinar. You can also check out all of our customer case studies here. For any questions or queries, feel free to reach out at email@example.com
If you’re ready to try Datos IO RecoverX for yourself, simply register here for a free trial. Take The Tesla Of The Backup for a ride on Amazon Web Services (AWS), all on us!