Backups – Reducing complexity & Backup Footprint

Current State of Backups

Backups are constantly growing in size and complexity as organizations grow over time.  The complexity of backups is characterized by both the number of servers being deployed as well as the variety of applications and data types associated with each server.  The size of backups and also the total backup footprint of an organization grows with this complexity.  This is due to growth within each business unit and also the trend for critical applications to grow in size over time. 

Complexity issues

  • Number of servers (OS Types)
  • Application types (Databases, File sharing, Web Applications)
  • Disaster recovery requirements

The complexity of the environment is compounded by the need for simple and effective disaster recovery requirements and procedures.  Each critical application found within an organization needs to have a disaster recovery plan. The plan should focus on addressing the recovery point objective (RPO) and recovery time objective (RTO) of the application.  Ideally the organizations most critical applications would have the lowest possible combined RPO and RTO. 

RPO & RTO Review

  • Recovery Point Objective (RPO) – How much data can you afford to lose?
  • Recovery Time Objective (RTO) – How much time do you have to recover the application/data?

Step one in building a structured scalable backup environment will be to define tiers of backup service.  Each tier will allow you to align a servers calculated RTO/RPO objectives with a logical backup solution.  Note backup tiers of service will depend heavily on the organizations current backup solution and may require purchase of new hardware solutions.  

Tier 1: Continuous Backups

RPO & RTO Closest to Zero

  • Disk to Disk (Mirrored Storage)
  • Disk to Disk (Continuous remote replicated mirror)
  • Disk to Disk (scheduled remote replicated)

Tier 2: Daily Incremental, Weekly Full (traditional backup solution)

RPO & RTO increased

  • Disk to Disk (Point in time backup data cut – snapshot)
  • Host to Disk (Backup to Virtual tape library)
  • Host to Disk to Tape (Backup to Virtual tape library and archive on tape)

Tier 3: Daily Incremental, Monthly Full

  • Host to Disk (Backup to Virtual tape library Host to Disk)
  • Host to Disk to Tape (Backup to Virtual tape library and archive on tape)

Tier 4: Monthly Full

  • Host to Tape

Tier 5: Archive Static Data

  • Host to Disk (Possibly even single instance)
  • Host to Tape

After looking at these 4 tiers of service you’re probably asking yourself.  Gee… it looks like I just added more complexity to a solution which is supposed to save me time and money?  The tiers are simply set as options and a guideline for classifying how an application can be backed up and recovered.  Moving up the chain of tiers you attain a lower combined RPO/RTO.  

Next up… reducing the cost of backups.  Part of reducing the cost of backups is weighing the cost of losing critical company data versus investing in a proper backup solution.  This simply put is justification for the cost of backups.  If an organization believes that placing their data on a single backup medium will enable them to meet their backup and recovery objectives they are doomed to fail.  Outside of enterprise solutions this may fly… but within organizations that provide any class of service a higher level of backup analysis and support is required.  

Reducing Backup Costs $$$ & Footprint

Re-evaluate Backups

  • Assign application to backup tier. Cost will be validated by the required disaster recovery objective of the application.  
  • Separate backup of System OS & Application.  

Windows 2003 Server Backup = Tier 3
SQL Server Application Data = Tier 2 

  • Allow non business critical applications to be backed up less frequently
  • Lower Tier applications cost less due to the backup interval being reduced.
  • Employ backup profile exclusions specific to OS

Exclude Windows System Patches (Folder $NT*)
Exclude page file & non critical system folders

  • Archive static data removing it entirely from backup cycle. 


  • Patches & Service Packs
  • System Files (pagefile)
  • Non Critical Uncessary Local Apps. (ex. Symantec)


  • Incremental vs. Full Backup size (Show app. rate of change)
  • Small incremental backup size + large full backup size = opportunity for data archive or move to lower backup tier
  • Backup Length vs. Backup Size
  • Identify largest backup & longest backup (backup bottlenecks)

Optimize Backup Schedules

Backups are traditionally performed during off hours when business functions have slowed and servers have spare cycles.  As enterprises grow in size this window for performing backups is strained by the number of scheduled jobs compounded against the available resources assigned to perform backups.  

Backup resources are commonly composed of backup servers and backup media.  Backup media can be in the form of mirrored disk, virtual tape media, and traditional backup tapes.  

Before attempting to optimize your organizations backup schedules servers and applications should be evaluated and assigned to a specific backup tier.  The backup tier will dictate the priority of the backup and also how the backup will be handled within the backup environment.  

The next step is to identify those backups which can be scheduled to run during business hours.  This can be backups which fall into the lower tiers or backups which are taken from disk snapshots and then off loaded to tape. 

Disaster Recovery and Backup considerations

DR Server in place?

Rebuild Server?

Hardware in place

  • Servers
  • Power
  • Cooling

Brain dump in progress…


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s