Arjuna Kodisinghe - stock.adobe.
JetStream's Azure-native DR finally takes off
JetStream DR for Azure VMware Solution offers DRaaS for VMware VMs and is deeply integrated with Azure, including using Azure Blob Storage to minimize storage costs.
After months of delays, JetStream's Azure-native disaster-recovery-as-a-service product has finally made its last boarding call.
This week, JetStream DR for Azure VMware Solution (AVS) became generally available in Azure Marketplace. The software enables customers to perform on-premises-to-AVS or AVS-to-AVS failovers for their VMware virtual machines. This version of JetStream DR is deeply integrated with Azure -- it's discovered and deployed through Azure Marketplace, viewable through Azure Portal and billed directly through Azure.
JetStream DR for AVS continuously captures and replicates data from VMware environments and stores the copies in Azure Blob storage, an object storage platform for unstructured data. Continuous data capture translates to near-zero recovery point objectives (RPOs), and storing the copies in Azure Blob translates to lower costs compared to using a file system as a repository.
In a failover scenario, the software restores the copies from Azure Blob into AVS. A standard recovery from Azure Blob can take one or more hours as the data is rehydrated at the time the failover happens, but JetStream DR also supports continuous rehydration into vSAN and Azure NetApp Files for recovery time objectives (RTOs) of a few minutes. Customers can divide their VMs along these two recovery methods based on criticality.
First introduced in December 2020, JetStream DR for AVS would become generally available in "early 2021," according to JetStream Software at the time. Instead, the vendor spent time improving the software before the official launch, said JetStream Software president and co-founder Rich Petersen, which included closer integration with AVS infrastructure, making deployment more automated, building a capacity planning tool so customers could accurately assess the storage costs of using the software and ensuring the product complied with Microsoft's privacy policies.
Rich PetersenPresident and co-founder, JetStream Software
"We had to do a lot of the unsexy but essential work," Petersen said.
JetStream DR for AVS costs $35 per month, per VM and will appear as a line item in customers' Azure bills.
JetStream DR for AVS addresses an important area of need within Microsoft's cloud, said Andrew Smith, a research manager at IDC. Azure VMs can be protected with Azure Site Recovery, but VMware VMs running in Azure need their own DR tool. By offering native disaster recovery as a service (DRaaS) and enabling Azure Blob as the repository, Microsoft can position AVS as cheaper DR than its AWS equivalent, VMware Cloud on AWS.
"It's filling a gap in the Azure ecosystem," Smith said.
JetStream DR for AVS' main competition will be from VMware's native DR tools: Site Recovery Manager and Cloud Disaster Recovery, the latter of which is based on technology from Datrium, a company VMware acquired in July 2020, he added.
However, JetStream has the advantage over VMware's native DR because it's a single product, according to Petersen. Site Recovery Manager replicates VMs to an active VMware Cloud to deliver immediate availability during a failover scenario, while Cloud Disaster Recovery replicates VMs to Amazon S3. Customers would use the former for critical VMs and the latter for non-critical ones. JetStream DR for AVS can handle both RTO needs.
The rise of DRaaS
DRaaS has been growing more popular because it addresses two of customers' current biggest concerns: ransomware and rising infrastructure complexity, Smith said. Data loss is one potential impact of a ransomware attack, but customers are also worried about time loss. A good DR plan will minimize how long a company's systems are unavailable during an attack, and DRaaS has the additional benefit of offloading the burden of DR to a service provider.
Additionally, customers are turning to DRaaS because running DR in-house is too much of an IT burden, Smith said. Most customers can't -- or don't want to -- devote IT resources to develop and maintain an off-site location, schedule regular failover tests or otherwise perform all the tasks necessary for guaranteeing their DR site will work when they need it.
"Nobody wants to spend a lot of time doing DR testing, developing and maintaining runbooks and all that other stuff," Smith said.
Johnny Yu covers enterprise data protection news for TechTarget's storage sites SearchDataBackup and SearchDisasterRecovery. Before joining TechTarget in June 2018, he wrote for USA Today's consumer product review site Reviewed.com.