Portworx Backup: Tuneable Backups for Kubernetes Applications
Portworx sponsored this post.
When it comes to backup and recovery, one size does not fit all.
It’s one thing for a mission-critical but a non-customer-facing application like a CI/CD pipeline to go down, and another for an e-commerce site to go down on Black Friday. And not all backups ever need to be “recoverable.”
According to an analysis by 451 Research, many companies need even “non-critical” applications and data — like the CI/CD pipeline — to be recoverable in under 24 hours, and to have minimal data loss in case of a failure. Specifically, 47% expect a recovery time objective (RTO) of under 24 hours, meaning the application should be down less than 24 hours. At the same time, 53% of companies expect a recovery point objective (RPO) under 24 hours. In other words, they expect to lose less than 24 hours worth of data.
Yet a full-blown synchronous disaster recovery solution would likely be overkill for applications that could be down for six, eight or ten hours. This kind of disaster recovery solution is too expensive to run and unnecessary for non-critical apps. It also doesn’t address another common backup use case: ensuring data is stored for compliance reasons.
To address this gap, container-native storage company Portworx launched PX-Backup, allowing companies to back up — and relaunch if necessary — Kubernetes applications’ data and configuration in a couple clicks. Users have control over how often individual applications should be backed up and can back up to any S3-compatible object storage. Because the product works in a container-granular, application-centric way, it can back up multinode distributed databases like Elasticsearch, Cassandra, Kafka and Cassandra.
Portworx released PX-DR earlier this year, a Kubernetes-optimized disaster recovery solution for mission-critical applications. With PX-Backup, the company is hoping to address the full suite companies’ backup needs. Backup is available separately from the rest of the Portworx platform, and is currently in tech preview.
There are two primary use cases for PX-Backup, according to Michael Ferranti, vice president of product marketing at Portworx. The first is for that CI/CD pipeline — any application that doesn’t need to be back up and running within minutes it it goes down, but still should be up recoverable within a couple of hours.
The second addresses an entirely different enterprise need — visibility for compliance reasons. With PX-Backup, it’s possible to backup data collected by applications running on Kubernetes in a way that’s easily searchable. If you need to prove that a customer’s data has been deleted within a certain time frame following a deletion request, for example, you’ll be able to log into a complete replica of the application on a certain date in the past and search the database on that data to ensure the customer data has been deleted.
“They have a mandate to take a backup, but they might never need to ‘recover’ that application,” Ferranti said. They do, however, have to be able to search the data, ensure it’s stored securely and is accessible if needed.
“The disaster recovery use case is, if this thing ever goes down, I absolutely need to bring my app back up,” Ferranti said. “Backup addresses a different market. It’s more of the operational issues, as well as the compliance use case, where you need to keep three copies of the data at all times, you need to keep it for three years and if it’s deleted you need to know who deleted it.”
Feature image by Gerd Altmann from Pixabay.