We'll start by reviewing the essentials of data backups, and then creating and implementing a data disaster recovery plan for your business. Accidents and disasters will happen periodically, and you're responsible for ensuring your systems and network are ready when they occur. The main purpose of backing up data is to have the ability to recover it later. Your goal should be to make data restoration as easy, quick and repeatable as possible, which requires careful planning.
Full : Backs up every file and folder in the configured backup scope to storage media, such as tape or disk. Full backups usually take the longest amount of time to complete but result in shorter recovery times. Incremental : Backs up files that have changed since the last backup of any type. Incremental backup jobs tend to complete quicker because they're smaller than full backups. However, restoring incremental backups can take a significantly longer amount of time depending on how many days it's been since the last full backup. Differential : Backs up files that have changed since the last full backup. Differential backups take longer than incremental backups as the number days since the last full backup increases. However, data restoration takes less time because you only need the last full backup and differential backup. Copy : Similar to a full backup, but the archive bit isn't changed when a copy backup is performed. This enables you to run a special copy backup at any time without disturbing your normal backup tape rotation. Continuous Backup : Files are constantly backed up as changes are made, usually to a storage system over the network or the internet. Historically, only mission-critical data, such as a customer ordering system, was deemed necessary for continuous backup. However, with the price of continuous backup solutions becoming very affordable, some businesses are switching to continuous backups for all business data.
Five-tape system : Daily Incremental, differential or full backup Monday through Thursday and Full backups on Friday / (Saturday) in case of 6 Day Week. Grandfather-father-son : The grandfather-father-son (GFS) backup system is also known as the monthly-weekly-daily system and uses three different sets of backup tapes. Use daily tapes, Monday through Thursday, to perform differential or incremental backups, although full backups can be performed on smaller systems. Use a weekly tape each Friday to perform a full backup of the system. Use 12 monthly tapes, on the last Friday of each month, to perform a full backup of the system.
Dirty or damaged tapes and tape drive heads This can result in errors being introduced to your backed up data, impairing your ability to perform a complete and accurate restore. Worse, it could result in garbage being written to your storage media instead of data, resulting in the loss of everything. Not only do tapes get dirty, but also they can become creased or otherwise damaged, resulting in lost data. Tape wear Pay attention to the manufacturer's recommendations. If the tape is rated for 1,000 recordings, don't stretch it. The risk of losing your data is too great. Buy new tapes and retire older media securely, making sure they're stored in a locked safe or destroyed. Don't just throw old media into the nearest dumpster where anyone can access them. Long-term storage If you intend to keep backed up data over the long term, make two copies in case one becomes lost or damaged. If you have data stored long-term on outmoded or obsolete media, transfer the data to a more modern storage method. Now that you've optimized your backup plan for your needs, you still need to know if it actually works. Waiting until a disaster strikes to test the plan is inviting more problems than you ever want to have.
Creating And Implementing A Data Disaster Recovery Plan
Creating and Implementing Data Disaster Recovery Plan
Backup Considerations <ul><li>How much time do you have to perform the backup? </li></ul><ul><li>How much time do you have to restore data? </li></ul>
Recovery Time Objective (RTO) <ul><li>The RTO (Recovery Time Objective) will be contingent on several factors, including: - </li></ul><ul><ul><li>The amount of data to be restored : Do you have to restore a single file or an entire data store? </li></ul></ul><ul><ul><li>The number of restores to be performed : Do you have to store one full backup plus three incremental, or just one differential? </li></ul></ul><ul><ul><li>The type of backup media in use : Are you restoring from physical media or over the network? Newer tape media such as LTO-4 work with high-speed drives that greatly reduce the amount of time required to restore data. </li></ul></ul><ul><ul><li>The location of the backup files : Is the data on media physically located in an offsite storage location? </li></ul></ul>
Recovery Point Objective (RPO) <ul><li>RPO is a measure of the acceptable amount of data you can lose, in hours. </li></ul><ul><li>Example : Organization might define its RPO as 2 hours, which means you must restore data to within 2 hours of any disaster striking. </li></ul><ul><li>Data created or modified within that 2-hour range is considered an "acceptable loss." </li></ul><ul><li>The RPO helps you determine how often you need to perform backups. </li></ul>
Types of Backup Media <ul><li>Disk-to-Disk backup. </li></ul><ul><li>Disk-to-Tape backup. </li></ul>
Laying the foundation for backups <ul><li>It begins with centralized storage on a server rather than on local PC hard drives. </li></ul><ul><li>Preserving network availability during backup </li></ul><ul><li>Backups require a significant amount of bandwidth. </li></ul><ul><li>That’s why backups are scheduled during off hours when few or no users are accessing the network. </li></ul>
Optimizing Backup Performance <ul><li>Leveraging current infrastructure. </li></ul><ul><ul><li>Databases have a habit of growing with amazing speed. </li></ul></ul><ul><ul><li>They consume storage capacity at an alarming rate. </li></ul></ul><ul><ul><li>It's prudent to monitor end users' data storage allotment. </li></ul></ul><ul><ul><li>Need to limit storage to necessary business-related data only. </li></ul></ul>
Meeting Future Challenges <ul><li>No matter how well you optimize your system, your needs eventually grows. </li></ul><ul><li>When increasing your storage capacity, your network infrastructure must also be able to manage the load. </li></ul><ul><li>Even when backing up in off-peak hours, you still have only so many hours to conduct the backup. </li></ul>
Meeting Future Challenges <ul><li>Because of other maintenance tasks, you can't depend on 100-percent processing power and network availability, dedicated to backups. </li></ul><ul><li>Upgrade your hardware and network infrastructure, as needed, to provide sufficient ability to back up the growing amount business data. </li></ul>
Securing Data Backup & Storage <ul><li>It's possible for your data to be at risk of interception and theft, both during the backup process and while in storage . </li></ul><ul><li>To protect data in transit , the best method is using I nternet P rotocol Sec urity (IPSec) over a V irtual P rivate N etwork (VPN) tunnel to ensure security. </li></ul>
Securing Data Backup & Storage <ul><li>To maintain the security of backup media, keep your portable storage in a tape vault or some other secure location. </li></ul><ul><li>Smaller businesses with limited budgets, can use a safety deposit box or offsite safe. </li></ul><ul><li>The location must be readily accessible to authorized staff should the media be needed for a recovery procedure. </li></ul>
Storage Media Management <ul><li>Managing backup tapes is more than just switching them out and making sure they're properly stored. </li></ul><ul><li>There are a number of issues that come with using storage media repeatedly over long periods of time. </li></ul>
Storage Media Management <ul><li>Dirty or damaged tapes and tape drive heads. </li></ul><ul><li>Tape wear. </li></ul><ul><li>Long-term storage. </li></ul>
Testing Data Restoration <ul><li>Set a regular schedule to test your recovery system. </li></ul><ul><li>Remember that you're not just testing whether the system works but also how quickly it works. </li></ul><ul><li>How long can your business afford to remain offline without access to critical data? </li></ul><ul><li>The pre-determined values of RTO and RPO will come into play. ` </li></ul>
Putting the Plan in Writing <ul><li>The first part of testing data restoration is developing and documenting a plan. </li></ul><ul><li>In a disaster, multiple parts of your network infrastructure can fail or at least be impaired. </li></ul><ul><li>Your recovery plan, should take into account all of the different aspects, of the overall system and how to respond when faults occur. </li></ul>
Putting the Plan in Writing <ul><li>One method is to create an overall disaster management plan that addresses aspects of recovery after a disaster, with the following individual sub-plans: - </li></ul><ul><ul><li>Systems : Covers handling of server faults and general restoration of data, applications and services </li></ul></ul><ul><ul><li>Network : Focuses on bringing up internetworking devices, such as routers and switches </li></ul></ul><ul><ul><li>Communications : Coordinates how different organizations are contacted, such as law enforcement, company management, hazardous materials personnel and Federal Emergency Management Agency (FEMA), if necessary. </li></ul></ul><ul><ul><li>These different parts of the plan can easily map to different teams in a larger organization. </li></ul></ul>
Testing the Team <ul><li>When delegating responsibilities, don't forget to assign the task of performing a recovery in the event of a disaster. </li></ul><ul><li>When you test the team, test the recovery plan and how the members of the team mesh in their tasks. </li></ul><ul><li>If a team member has successfully corrected the hardware fault, do they have to wait for the tapes to be made available to initiate the recovery? </li></ul><ul><li>If the tapes and the servers are ready, is there a delay in restoring the correct configuration files to the local switch? </li></ul><ul><li>Testing the team is like running a fire drill. </li></ul><ul><li>You not only find out how well they work together, but also where the faults and gaps are in performance and, to some degree, the plan itself. </li></ul>
Running Test Levels <ul><li>Because you can face different types of disasters, you should run different types of tests. </li></ul><ul><li>One of the most common tests is restoring data from tape in the event of a data loss. </li></ul><ul><li>Any test you run must be conducted in off-peak hours when few or no end users are on the system. </li></ul><ul><li>Planning for the occasional weekend testing "party" is a small price to pay for the relative security of knowing your recovery plan works. </li></ul>
Running Test Levels <ul><li>Beyond restoring data to a server. </li></ul><ul><li>You can also introduce issues to different parts of your system and see how quickly those issues are addressed. </li></ul><ul><li>Depending on how extensive you want to be, you can announce where the problem lies or allow your staff to attempt a diagnosis based on certain symptoms you announce. </li></ul>
<ul><li>Ensure everyone knows their role ahead of time so they can participate efficiently in testing the recovery plan. </li></ul><ul><li>Although backups are conducted regularly. </li></ul><ul><li>Recovery operations are rarely performed. </li></ul><ul><li>So make sure your staff is familiar with how to perform a server recovery and deal with all equipment and network connectivity. </li></ul>Running Test Levels
Summing Up <ul><li>In this presentation, you were explained, how to create and test a data backup plan. </li></ul><ul><li>Tips for restoring data, systems and your network after a disaster. </li></ul><ul><li>In addition, you were explained, best practices and many advanced systems and network administration techniques. </li></ul>