The modern data protection environments are very complex and demanding. There are numerous layers involved starting from bare bone hardware, firmware, device management, kernel, operating system, volume management, file system, connectivity and all the way up to very sophisticated applications. At each of these layers there are a variety of offerings. There are close to hundred different hardware vendors. There are well over dozen different operating systems like Windows, Linux and Solaris etc. There are native volume managers like Solaris Volume Manager (SVM), Linux/AIX/HP LVM. There are standalone heterogeneous volume manager offerings like Veritas Volume Manager. The file systems range from quite rudimentary FAT, UFS, ext up to ultra-advanced VMF S5, LTFS, ext4, ReFS. There are numerous connectivity approaches like IP, SAN/FC, iSCSI. The applications like Oracle, SQL Server and SAP are getting richer and richer. There are different deployment models like virtualization, cloud and appliances in addition to traditional client – server. These environments have complicated requirements like clustering, HA, strict SLAs in various forms, shapes and sizes.
Backup and Recovery systems
Backup and recovery systems evolved traditionally to address these data protection needs. They focused on making a copy of the data to be protected so that it can be used in case the original data is lost or corrupted. This copy can be stored on disk, tape, appliances, cloud etc. These backup systems are required to manage the entire life cycle of such copies to meet the business requirements. Some of these backup images are required to be that of a complete system. These backups can be full for entire protection or incremental for efficiency. These modern backups are often required to have the knowledge of the application that will be running on the data that is to be backed up. It allows restoring of data as of any past backup taken at certain time. It allows performing browsing of data using catalog information. It also allows granular restore like a single mailbox on exchange server, only a few files on the entire file system etc. These systems are enhanced further by techniques like multiplexing, deduplication, compression, encryption and change tracking. They are typically geared towards delivering Recovery Point Objective (RPO) and Recovery Time Objective (RTO) based SLAs.
Snapshots and Replication
On the other hand, storage array vendors have been evolving technologies like snapshot and replication to better address the challenges of backing up the accelerating growth in data where backup windows are disappearing and backup and recovery SLAs are getting tighter with 24/7 production environments. The snapshot is an efficient way of creating copies of data. It also enables off-host mechanisms for data protection to alleviate the additional load on the production system. The vendor can further replicate the live data or its snapshots to another storage array or disaster recovery site. This replicated data can be useful for performing simulated or actual disaster recovery whenever necessary. These snapshots and replication are sometimes also implemented in firmware or software at various levels like volume management, file system, database or virtualization. There are various types of snapshots like Copy On Write (COW), clone, mirror etc. based on the need. The vendor can perform rollback restore from such a snapshot without doing any data movement at all but simply manipulating some metadata references. The vendor in some cases also does life cycle management across multiple such copies – like local snapshots, remote snapshots, remote replicated snapshots etc. – of the underlying data.
Enriching data protection by snapshot and replication
Netbackup 7.5 Replication Director brings best of both worlds’ together. This integration is achieved by Netbackup’s Open Storage (OST) framework. It provides the single pane of glass for managing the entire enterprise global backup and recovery environment. From within NetBackup, Storage Lifecycle polices or SLPs, can be used to automatically initiate the snapshot, replicate the snapshot, expire the snapshot and move snapshots across storage tiers including tape for long term retention. In addition to simplifying the complexity of snapshot management, Replication Director enhances its recovery as well. It provides the ability to easily search, browse and restore a single file from any snapshot or replicated snapshot from anywhere. It brings all restore capabilities like catalog indexing, live-browse, granular restore, rollback, full restore to such snapshot copies. Below is an example of a Life Cycle Policy applied to a snapshot. It provides:
- Unified policy management,
- Snapshot copy management & monitoring,
- Global search and restore.
- Optimized rollback based restore from any of such copies
The optimized rollback based restore from any of such copies can be achieved as:
This integration is achieved by Netbackup’s Open Storage (OST) framework.
Netbackup Replication Director currently supports NAS file system on NetApp filers.