Enterprises have a wide variety of workloads such as databases, app/web/file servers running on-premises, or in the public cloud. In the public cloud, they could be running workloads using IaaS, such as AWS EC2, Azure or GCP VMs, PaaS databases such as AWS RDS, Azure SQL, GCP CloudSQL. They all need to be protected. Here is a cloud backup and cloud DR best practices checklist that many of the Fortune 5000 enterprises have shared:
- Cloud Backup AND Cloud DR
Most enterprises end up buying two points tools: one for backup retention, and another for low RTO & RPO DR. It increases in high license and cloud infrastructure costs.
Look for a solution that can satisfy backup retention, file/folder restores, and low RTO & RPO DR requirements in the public cloud.
- Multi-Cloud & Hybrid Cloud
The cloud backup solution must not only protect workloads in AWS, Azure, GCP, IBM cloud but also protect on-premises workloads to various clouds…all managed from a single pane of glass instead of siloed and many point tools.
- Wide platform support
The solution should support a wide variety of workloads running in on-premises VMs, physical servers, cloud VMs, cloud PaaS, and SaaS as well.
Beware of solutions primarily focused on VMware but claim physical servers, databases, cloud databases, etc.
- Extensive Database support
Enterprises build mission-critical applications using databases such as SAP HANA, Oracle, Microsoft SQL Server, MySQL, PostgreSQL, MongoDB, MaxDB, Sybase, etc.
Look for a solution that can protect all these databases on-premises as well in AWS, Azure, GCP, IBM cloud.
- Cloud Snapshot Management
Unlike on-premises storage snapshots, cloud snapshots are stored in cloud object storage, thus reducing storage costs. However, writing scripts using native snapshots is painful, error-prone, and not easy to manage. This blog articulates the details very well.
Look for a solution that can deliver a single pane of glass for managing cloud snapshots in multiple-clouds and on-premises workloads.
- Low RPO for critical databases
Unfortunately, the native cloud snapshots are not application-consistent for databases. Thus, DBAs stay away from cloud snapshots for databases running in cloud VMs.
On the other hand, if DBAs use the legacy approach of database backups to disk in the cloud, the storage costs and the RPO will be very high. The real costs of AWS backup using EBS Snapshots can be quite high.
Look for a solution that can deliver low RPO with application-consistent, incremental forever backups for all the databases that I mentioned above in all public clouds.
- Low RTO for critical databases
Legacy solutions can not deliver low RTO with instant recovery because of storing backups in a proprietary format, instead of native database format.
On the other hand, some solutions allow lower cloud disaster recovery RTO and TCO but it can come at a high cost.
Look for a solution that can deliver low RTO for a wide variety of databases in AWS, Azure, GCP, IBM cloud and on-premises.
- Recover directly from cloud object storage in minutes
It’s essential to keep the costs low while delivering low RTO and RPO. In the cloud, you can achieve the lowest cloud infrastructure costs by storing backups in cloud object storage such as AWS S3, Azure blob, GCP Nearline, IBM COS. But to deliver low RTO, you need a solution that can mount backups directly from the cloud object storage and recover databases and VMs in the cloud.
- Eliminate on-premises backup storage
Another area to look to reduce costs is to eliminate backup storage on-premises. Instead, look for a solution that can backup directly from on-premises to public cloud storage for the majority of the workloads.
- Eliminate on-premises infrastructure for DR
Legacy DR approach requires at least 75% compute and 100% storage allocation for DR at a 2nd data center, which leads to high CAPEX and low utilization.
Instead, look for a solution that can not only deliver cloud backup, but also help you with 1-click DR orchestration in any of the public clouds such as AWS, Azure, GCP, and IBM. This not only eliminates expensive data center costs but also delivers an on-demand cloud cost model for on-demand DR.
- Reuse backups for test data and analytics
Many enterprises view backups as a graveyard where the data goes to die.
It doesn’t have to be like that.
Especially in the cloud, where a fantastic number of cloud services are available, users should be able to mount the backup data instantly and use it for test/dev, source data for data warehouse & analytics, offline ransomware detection, etc.
- Available as product and SaaS offering
And last but not least, look for a solution that is not only available as a product with perpetual and subscription licensing, but also available as a SaaS with subscription payments for both on-premises and cloud workloads for all use cases such as cloud backup and cloud DR.
I hope you found our cloud backup and cloud DR best practices checklist helpful. In a related webinar, learn how to reduce costs by up to 55% with cloud backup & disaster recovery?