Services > Business Continuity and Recovery

Your servers, systems and data are the lifeblood of your business. There are hundreds of servers and sub-components that can fail and many more ways environmental or human factors can trigger these failures. While impossible to predict how and when the next failure will occur, it is surprisingly straightforward to implement appropriate protections to limit the occurrences and impact of these inevitable failures.

The 7 Deadly Sins of Backup and Recovery

  1. Backing up only the data
  2. Allowing backups to go untested

    Organizations often spend an enormous amount of time making backups (weekly masters, nightly incrementals, and so on). If the backup volumes being created cannot be restored on a reliable basis, the process has effectively failed. The basic rule of data protection and disaster recovery systems is that they are unproven until they have been fully tested, and shown to be effective, on a frequent and ongoing basis.

    The typical organization checks backup schedules to make sure that they are correct; checks that the backup jobs ran; and checks error logs to be sure they ran to an apparent successful completion. With those confirmations in hand, we assume that data recovery will be possible when necessary.

    But media degrades over time; operating environments change in ways that make it unlikely that previous backups can be successfully restored; and the typical error rates associated with the media we use make it at best uncertain that the restoration will be successful.

  3. Lack of adequate recovery planning
  4. Not planning for a dissimilar recovery environment
  5. Not having offsite copies
  6. Confusing replication and vaulting

    The disadvantages to replication are that the synchronization occurs at a very low level, below the ability to see data and file system errors or corruption. This means that the inevitable errors (some of which are potentially fatal to a restoration) that occur naturally over time will be immediately replicated to the offsite location…and the hoped-for protection is thereby immediately lost.

    Replication is also highly resource-intensive. These systems create a substantial burden on the local CPU, since most replication technologies utilize a type of RAID engine to capture the block-level changes. These are implemented through software, imposing significant CPU overhead.

    This sort of replication is also extremely bandwidth- intensive. As every data change for every transaction is moved across the network to the second location, this approach can impose a huge burden on a corporation’s network infrastructure.

  7. Adopting inexible solutions

 

Are You Thinking About These Issues?

Download a list of the questions you should ask yourself and your organization, if you haven't started already.

Download the Self-Assessment Questionnaire

 

How to Choose Which Technology


  • Confused over which technology to choose?

    Direct attach (DAS) is typically when a tape drive connects directly to the server.

    Local Area Network (LAN) is typically when there are multiple clients feeding one backup server. Autoloader is directly attached to the backup server.

    Storage Area Network (SAN) is typically run through an FC switch. In most cases, a library will be used.

    Direct-Attached DAS

    LAN-Attached LAN

    SAN-Attached SAN

    iSCSi for DAS/LAN/SAN

    Single servers

    Single/Multiple servers

    Muliple servers are configured together

    Single/Multiple servers

    Attended backup

    Unattended backup

    Unattended backup

    Unattended backup

    Tape drives

    Tape Autoleaders

    Tape Libraries

    Virtual Tape Libraries

    Disk capacity to be backed up is less than 400GB

    Disk capacity to be backed up is 400GB to 2-3TB

    Disk capacity to be backed up more than 3TB

    Disk capacity to be backed up less than 3TB

    Cost conscious customer

    Consolidation is required

    Consolidation and centralized management is required

    Consolidation and centralized management

     

     

    Storage external to servers in JBOD or disk arrays

    Customers looking for rapid file restore

     

     

     

    Allows for connect to DAS/LAN/SAN without cost of fiber

     

    • Business continuity and recovery

    Terms to Know

    • Baremetal Recovery
      The process of taking a low-level snapshot of a machine's operating system partition and storing it where it can be quickly and easily accessed when required.

      There are two parts:

      The first is a program that is set up to periodically snapshot an OS partition using image backup technology. This is installed as a service and comes with a scheduler. The scheduler is then programmed to take backups of the live machine without any requirement to shut down services, close applications, or go offline. Baremetal backups are normally stored to a UNC (Universal Naming Convention) path, SAN (Storage Area Network), or NAS (Network Attached Storage) device for online storage and quick access when needed.

      The second part of a BMR solution is the process used to boot a dead machine. This enables users to connect to the online location where the image backups have been stored and initiate a restore. Once the OS partition has been restored (which can take between 5 to 30 minutes), the only remaining steps necessary to complete the disaster recovery are to remove the boot media and reboot the machine. This latter phase takes approximately two minutes before the machine is back to the exact state at which the image backup was performed.

      Static image backups are typically performed every 24 hours, with options for incremental and differential backups between full backups.
    • Dissimilar Restoration
      Clients generally fail to realize that restoring the data of an information processing environment to a single recovery point in time, assuming the original hardware environment no longer exists or is currently unusable, is challenging. While some software and operating systems individually may be able to run on a very limited range of hardware platforms and specific models of those platforms, the intersection of all those programs creates a real, singular hardware profile in time for recreating that environment.

      In anticipation of when when the original environment no longer exists or is unusable, many organizations may consider purchasing duplicate hardware for the most critical computers. Maintaining duplicate hardware for an entire site is so cost prohibitive that only the most critical IT operations can justify. Besides, it is even difficult to guarantee that the same model can be available some time after the initial purchase, which implies that purchases must be done at the time of the original system acquisition, doubling costs.

      Dissimilar restoration implies the ability to save and restore servers on any system and offers to provide a great flexibility and constitute a significant savings. Ideally, you would be able to recover a single processor computer to a multi-processor computer, or recover from expensive Fibre Channel or SCSI drives to ATA or SATA storage. Be skeptical. There are so many variables involved, it is actually very difficult to guarantee hassle free dissimilar storage. If promised, test to verify with regularity over time if dissimilar storage is a “must have” for your organizations disaster recovery plan.
    • Dislike to Known
      The ability to do a bare-metal restoration to a different kind of hardware than the source was taken from, but only to a pre-defined set of destination hardware devices. This can work reliably if drivers for those devices are already embedded in the backups themselves.

     

    Technology Decisions, Inc.

    Phone: +1 847.361.3633   Email: matt@technologydecisions.net

    © 2003-2008 Technology Decisions, Inc. All Rights Reserved.