Chapter Summary

This chapter explores two important networking concepts: fault tolerance and disaster recovery. Although many people think fault tolerance and disaster recovery are one and the same, they are in fact very different, but equally important, concepts.

Fault tolerance usually refers to the measures network administrators use to ensure that data and network services are always available to network users. A strong fault-tolerance strategy does not happen by accident; rather, you must consider many factors when choosing the best fault-tolerance strategies for a specific organization.

Because availability is such a huge issue and server downtime is so costly, most of the hardware components within a server need to be considered as part of a fault-tolerance solution. Hard drives typically receive the most attention because they're 50% more likely to fail than any other component. The mechanism used to protect against such failures is RAID.

Several RAID levels are available today. The most common are RAID levels 0, 1, and 5. Although RAID 0 is a RAID level, it does not offer any fault tolerance, but it does offer performance improvements over using a single disk. RAID 1 uses disk mirroring to establish fault tolerance but suffers from 50% overhead and limited storage capacity. The RAID level of choice for organizations that can afford it is RAID 5. RAID 5 stripes data and parity information over several disks. The parity information can be used to re-create data in the event that a hard drive in the array fails.

Other fault-tolerance measures include using UPSs, redundant components, and sometimes redundant servers.

Disaster recovery involves having in place measures that can be used when the system goes down. To protect data from disaster, you need backups. Three key types of backups are available: full, incremental, and differential. A full backup makes a copy of all data, an incremental backup makes a copy of the data that has changed since the last full backup or the latest incremental backup, and a differential backup saves everything that has changed since the last full backup.

In addition to backup methods, a backup rotation strategy ensures that data is sufficiently recoverable. The most common backup rotation strategy is the GFS rotation. This type of rotation requires numerous tapes for daily, weekly, and monthly backups.

In some networks, VLANs are used to segment a LAN into logical segments. This segmentation allows for increased organization and security for the logical LANs. Although VLANs have been slow to catch on, many people feel they will soon be common in large organizations.

As the volume of data used in modern networks continues to increase, better data storage technologies have been developed. Two of these strategies are NAS and SANs. A SAN solution is very costly and therefore limited to large organizations that have the money and resources to support such a strategy.