Chapter 7 Backup & Recovery

1. Chapter 7Backup & Recovery "One important key to success is self-confidence. An important key to self- confidence is preparation." -- Arthur Ashe   Prepared by: SITI ZAINAH ADNAN If you do have any feedback or comment, please feel free to email me at sitizai@hotmail.com Your cooperation is very much appreciated !

2. Chapter 7Backup & Recovery TOPICS • Introduction • Regular Backup • Comparisons of backup types • Redundant Servers & Disks • Power Management • Finding Backup Resources • System Documentation

3. References • Book (available at the Informatics library) • Notes (available at IVC)

4. Introduction • Disaster recovery is a process of restoring the systems back into operation quickly and accurately. • Organisation must prepare to expect the unexpected events as part of routine. • Protecting computer systems from natural disaster require careful planning, testing and evaluating options to achieve such redundancy.

5. Introduction • A good disaster recovery planning has two important objectives: • Minimise the impact of a disaster on the ability to conduct business • Speedy and accurate recovery to normal operations • Two approaches to disaster planning: • Avoid it • Reduce it

6. Introduction • Avoid it: • The simplest and most effective • Protect the equipment and data against any kind of risk we can anticipate • Requires a good security plan e.g. from natural disaster or even sabotage

7. Introduction • Reduce it: • If can’t avoid it sometimes not possible to eliminate a risk completely • Do whatever necessary to minimise the impact • Try to control damage should the disaster happen in spite of our best effort • Have contingency plan ready to be put into practice e.g. if we have limited budget, concentrate on protecting the resources that would be hardest to replace

8. Introduction • Factors that contribute to disaster plan to make it a good plan: • Must provide for the provision to refine the plans • Call for regular training exercise and improvements to plans • Make provisions for maintaining the plan • Employees should be trained to execute their roles in the recovery process • Top management support should be exist it is an important factor

9. Dealing with Disaster • Organisation must make regular backups and store them off site. • A contingency site should be available for the employees to continue working. • Staff must collect enough information to duplicate the original system.

10. Regular Backup • A good backup system allows to copy data to a variety of media, such as: • Small-capacity removable disks • Large-capacity removable disks • Removable optical disks • Magnetic tape • After you choose your backup medium, you must decide what type of backup to run. • The three backup types are: • Full • Differential • Incremental

11. Backup Medium • Small-Capacity Removable Disks • Magnetic media disks - capacity < 500MB • Can be removed from the drives and replaced as they filled • Popular bcoz low cost and easy to use • Floppy disk - 1.44MB, 2.88MB • SyQuest cartridge - 44MB, 88MB, 105MB, 200MB, 230MB • Iomega ZIP disk - 100MB, 250MB • Imation SuperDisk - 120MB • Removable hard disk drive - varies

12. Backup Medium • Large-Capacity Removable Disks • Can store > 500MB per disk • Good for backing up workstation that has one/two disks • Iomega Jaz - 1GB, 2GB • Sharq - 1.5GB • Syquest cartridges - 1GB, 1.5GB • Removable hard disk drive - varies

13. Backup Medium • Removable Optical Disks • Use a laser to read and write on disk • Capacity is large - 128MB • Fairly slow • CD-ROM, CD-R, CD-RW - 650MB • Magneto Optical disk - 650MB, 1.3GB, 4.6GB • DVD - 4.7GB to 17GB

14. Backup Medium • Magnetic Tape • Oldest and most popular backup medium for offline data storage • Stores data in magnetically oriented metal particles form (copper oxide or chromium dioxide) on a polyester tape • Simple, inexpensive and high capacity • QIC - 100MB to gigabytes and above • DDS Cartridge for DAT drives - Gigabytes • DLT - 35DB and above • AIT - 25GB, 50GB

15. Full Backup • In a full backup, all network data is backed up. • To do a restore after a crash, only use one set of tapes to restore from. • Simply insert the most recent full backup into the drive and start the restore. • Normally, full backups take several hours • Requires: • Backup drive and medium with enough capacity • Backup system that can automatically changes its own tape

16. Differential Backup • A single, full backup is done typically once a week. • Every night for the next six nights, the backup utility backs up all files that have changed since the last full backup (the actual differential backup). • After a week's worth of differential backups, another full backup is done, starting the cycle all over again.

17. Differential Backup • The backup utility keeps track by using archive bit, which is simply an attribute that indicates a file's status of the current backup type. • The archive bit is cleared for each file backed up during the full backup. • Two backup sessions required to restore the system: • The last full backup • The most current differential backup

18. Differential Backup • Used when the amount of time each day available to perform a system backup (called the backup window) is smaller during the week and larger on the weekend.

19. Differential Backup • The files that changes at the beginning of the week may have been backed up several times, even though they haven’t changed since the first part of the week

20. Incremental Backup • A full backup is used in conjunction with daily partial backups to back up the entire server • Thus reducing the amount of time it takes for a daily backup. • The weekly full backup takes place, and the archive bit is cleared during the full backup. • The daily backups back up only the data that has changed since the last backup (not the last full backup).

21. Incremental Backup • The archive bit is cleared each time a backup occurs. • Only files that change since the previous day's backup are backed up. • Each day's backup is a different size because a different number of files are modified each day. • To restore, it requires: • The full backup • Every tape from the day of the failure back to the preceding full backup

22. Incremental Backup

23. Comparison of Backup Types

24. Redundant Servers and Disks • Most network servers include higher levels of backup by duplicating data on redundant hard disks. • There are three levels of protection available: • Server redundancy • Disk redundancy • RAID technology

25. Redundant Servers and Disks • Server redundancy: • Maintains the duplicate files on a completely different system. • Links two servers via a high speed connections. • If one server fails, the other keep the networking running.

26. Server • Shared disk model: Linking two or more servers to operate as one and to equally share resources Hub Server

27. Main connection Server • Shared disk nothing model: Linking two or more servers to operate as one, but with each owning particular disk, CD-ROM, and tape resources Backup connection in case of server failure Hub Server Main connection

28. Redundant Serversand Disks • Disk redundancy: • Maintains the duplicate files on separate disks in the same system. • More vulnerable than server redundancy, but less expensive. • Disk mirroring makes use of same bus adapter; a duplex drive uses a separate adapter and provides more security.

29. Mirrored Drive Primary Drive Controller or Adapter Redundant Serversand Disks • Disk mirroring: A fault tolerance method that prevents data loss by duplicating data from a main disk to a backup disk. Some operating systems also refer to this as disk shadowing.

30. Controller or Adapter 1 Primary Drive Controller or Adapter 2 Mirrored Drive Redundant Serversand Disks • Disk duplexing: A fault tolerance method similar to disk mirroring in that it prevents data loss by duplicating data from a main disk to a backup disk; but disk duplexing places the backup disk on a different controller or adapter than is used by the main disk.

31. Redundant Serversand Disks • RAID technology: • Redundant Arrays of Independent (Inexpensive) Disks, includes disk mirroring or duplexing. • Duplicate copies of the same data into different surfaces of the disk. Should the data is corrupted then recovering can be done from the other surfaces • It is widely available standard feature in many file servers and network products.

32. RAID 0: Striped Disk Array without Fault Tolerance Commonly used Requires a minimum of 2 drives to implement Provides data striping (spreading out blocks of each file across multiple disks) Only improve performance Does not provide fault tolerance

33. RAID 1: Mirroring and Duplexing Commonly used Requires a minimum of 2 drives to implement If the first disk fails, the second automatically takes over Each drive has duplicate information of the other If both fail, new drive install and restore data from a backup

34. RAID 2: Hamming Code ECC Requires 3 drives One drive store parity data – every bit Any drive fails, data rebuild from parity data from parity drive

35. RAID 3: Parallel transfer with parity Commonly used Requires a minimum of 3 drives to implement Similar to RAID 2 – but data strips in byte instead of bit Any drive fails, data rebuild from parity data from parity drive

36. RAID 4: Independent Data disks with shared Parity disk Requires a minimum of 3 drives to implement Similar to RAID 2, 3 – but data striped in block Any drive fails, data rebuild from parity data from parity drive

37. RAID 5: Independent Data disks with distributed parity blocks

38. RAID 5: Independent Data disks with distributed parity blocks Commonly used Requires a minimum of 3 drives to implement Data and parity are striped across several drives Allows for fast writes and read The parity information for data on one disk is stored with the data on another disk If any one disk fails, the drive can be replaced and its data can be rebuilt from the data stored on the other drives

39. Power Management • Power Backup and Conditioning • UPS & SPS are used to protect from power failure. • UPS (Uninterruptible Power Supply) is supplying continuous AC power • SPS (Standby Power Supply) is using DC power to maintain the power to the systems.

40. Power Management • Power Backup and Conditioning • Link conditioners smooth out the power flow. • Surge protectors block voltage surges that exceed specified values. • Line conditioners are surge protectors provide more stable power.

41. Finding BackupResources • Identify any backup resources that can be developed and maintained within the organisation. • This is most effective and least expensive way to solve problem.

42. Finding BackupResources • Co-operative Agreement: • Most traditional way to find a business partner to share the computing resources when required. • Either for a disaster situation or otherwise.

43. Finding BackupResources • Hot Sites: • Every computer system and piece of information has a redundant copy -systems must be up 100% • Budgets are typically large. • The redundant system(s) will take over for the failed system without any downtime. • The technology used to implement hot sites isclustering, which is grouping multiple computers to provide increased performance and fault tolerance.

44. Finding BackupResources • Warm Sites: • Nearline site, the network service and data are available most of the time (more than 85 percent of the time). • Warm site technologies require some administrator intervention • The commonly used technology is a duplicate server, which is a server that is available to replace any server that fails.

45. Finding BackupResources • When a main server fails, the administrator installs the new server and restores the data; the network services are available to users with a minimum of downtime. • Once the main server comes back, it is now become the spare/redundant server and is available when another server fails.

46. Finding BackupResources • Cold Sites: • A cold site cannot guarantee server uptime. • It has little or no fault tolerance and rely completely on efficient disaster recovery methods to ensure data integrity. • If a server fails, the IT personnel do their best to recover and fix the problem.

47. Finding BackupResources • If a major component needs to be replaced, the server stays down until the component is replaced. • Apart from regular system backups, no fault tolerance or disaster recovery methods are implemented. • This type of site has one major advantage: • The cheapest way to deal with errors and system failures. • No extra hardware is required

48. System Documentation • System Documentation • It is important to keep a full and timely record of LAN configuration. • This information is required to rebuild the network after disaster and also recover from network crash. • Should make sure the instruction are clear and easy to understand.