Guest Author: Randy Weis
Date: June 14, 2022
There are different levels of Disaster Recovery in an IT organization depending on the acceptable data loss (RPO or Recovery Point Objective) and acceptable time to recovery (RTO or Recovery Time Objective). In addition, there are other variables involved when planning DR. Such as:
Disaster Recovery Types
Business Continuity – covers more than IT, such as facilities and communications
Disaster Recovery – recovery of business applications
Disaster Resiliency – avoiding disasters with resilient infrastructure such as clustering at platform, datacenter or network level (SDWAN)
High Availability – application-level redundancy and failover
Scope of the Disaster
Partial Site Disaster – Some portion of the IT infrastructure is taken offline by human error/malice, burst pipe, hardware failure
Total Site Disaster – An entire site (datacenter, server room, server closet) goes down due to natural disaster, fire, flood, local power outage/generator failure, and will be down for more than a few hours (days, weeks)
Disaster Recovery Location
Veeam works at the Disaster Recovery Level and complements the other levels of workplace disaster recovery. Veeam provides orchestrated Disaster Recovery from one location to another. In addition, Veeam offers Direct Restore to AWS. GCP or Azure for VMs and physical machines. Regardless of workload, Veeam meets the level of DR necessary in one license and UI.
Level 0: Recovery of Backup Server
Scope: Partial or Total Site Disaster
The Veeam Backup & Replication Server can be recovered quickly from a Configuration Backup. A new VM can be created (from template, preferably) at the same or alternate site depending on the scope of the disaster. Once the configuration database is recovered, or the local repository of backup copies is rescanned, restores can begin (see Level 1).
This is an internal recovery option for IT, prior to site or application recovery.
One of the main differentiators of Veeam though compared to the market is you can still easily restore your data if there is no configuration backup. All backup files are portable and have the metadata needed for recovery. Simply install a new Veeam server or turn on the idle one in your DR site and rescan the primary or DR repository whether it's on-prem or in the cloud.
Level 1: Instant Recovery
Scope: Partial Site Disaster
RPO = hours
RTO = minutes
This is an “on-site” disaster recovery. Veeam offers Instant Recovery to VMware/Hyper-V/AHV for backups of:
Virtual Machines/disks in VMware/Hyper-V/Nutanix AHV/RHV
VMs in Storage Snapshots
Physical Windows and Linux machines (Veeam does the P2V conversion)
NAS File Shares: R/W for SMB, Read-only for NFS
Databases for Oracle or SQL
Backups of VMs in Azure/AWS EC2/GCP to VMware
Level 2: Recovery from Backups at another location
Scope: Partial or Total Site Disaster
RPO = hours
RTO = hours/minutes
Veeam copies backup files from one repository to another in a Backup Copy Job. A Veeam Backup & Replication Server (VBR) at that site can restore the backups from the target repository for Backup Copy Jobs. If log shipping for SQL or Oracle is enabled, those logs will also be available at the recovery location. Backup Copy Jobs can copy & Restore any backup that is stored in a Veeam repository:
VMs from VMware, Hyper-V, Nutanix AHV, RHV
Agent backups for Windows, Linux, Solaris, AIX
RMAN, SAP HANA, SQL
VMS from AWS/Azure/GCP stored in an External Repository (Object Storage container)
Level 3: Replication & Orchestrated Recovery of VM from VMware Snapshots (and Hyper-V)
Scope: Total Site Disaster
RPO = 30 minutes to 4 hours for replicas, 24 hours for recovery from Backup
RTO = minutes to hours, depending on scope and hardware resources
Veeam replicates Virtual Machine files from one datastore to another using VMware Snapshots with built-in orchestration that can re-map to DR virtual networks, re-IP addressing, boot order and boot delay. Replication can use Backup Copy Jobs as a source, so that Recovery from Backups RTO can be reduced to minutes. Veeam can build failover plans that can be triggered on demand or planned; offers commit failover or failback (testing), failback and commit failback (after original site is available). Supported workloads are:
VMware
Hyper-V
Level 4: Continuous Data Replication & Orchestrated Recovery of VMware VMs
Scope: Total Site Disaster
RPO = 2 seconds to 5 minutes
RTO = minutes
Veeam replicates VMware Virtual Machine files from one datastore to another using VMWare API for IO Filtering (no snapshots), with built-in orchestration that can re-map to DR virtual networks, re-IP addressing, boot order and boot delay. Veeam can build failover plans that can be triggered on demand or planned; offers commit failover or failback (testing), failback and commit failback (after original site is available).
Level 5: Disaster Recovery Orchestrator
Scope: Total Site Disaster
RPO: 2 seconds to 24 hours
RTO: minutes
Veeam provides a higher level of orchestration and automation with Veeam’s Orchestrator software. This software can recover in an automated manner from backups, from snapshot replicas and from CDP replicas. DRO offers a more complete and customized automation of application recovery with
Runbooks that are updated as the plan is updated
Scheduled testing, bubble or actual
Audit Logs
Readiness Reports
Test Result Reports
Delegation of recovery by role and by site
Veeam’s Disaster Recovery Orchestrator can be implemented as an enterprise strategy and solution. Scenarios for orchestrated recovery are:
Replication Jobs (snap or CDP)
Backup Jobs
Storage Replication (NetApp and HPE 3PAR)
Disaster Recovery to the Cloud
Scope: Partial or Total Site Disaster
RPO = 24 hours
RTO = hours
Veeam copies backup files from a local repository to a Capacity Tier repository (S3 Bucket) in AWS/Azure/GCP. A Veeam Backup & Replication Server (VBR) in the cloud can restore the backups from the S3 repository directly into a running VM into one of the major public cloud providers.
