Disaster Preparedness Options Available with our San Francisco Computer Consulting Firm

The following are four basic disaster recovery preparedness options to consider. In deciding which option works best for you, keep in mind that the more thoroughly you prepare, the less time you will be down and the less data you are likely to lose. Both minimize revenue loss.

The scenario our San Francisco computer consulting firm generally uses for disaster recovery planning purposes is an earthquake affecting the Bay Area region. In such a scenario, we assume complete loss of access to your office. Only information you already have off-site in the form of a magnetic tape or data stored as a result of over the Internet backups is available. In such a scenario, our resources, and those of other IT companies, are likely to be stretched very thin.

In planning for recovery, our San Francisco computer consulting firm assumes:

• You redirect your main office number to a backup number, such as that of a branch office in another area. Calls are received there and forwarded to employee’s cell or home phones
• Employees work remotely from wherever an internet connection can be found. They log into the company’s terminal / citrix server.

The most basic level of preparedness is to:

• Ensure that all data is backed up to a tape which is taken off-site every day. Alternatively, data can be backed up over the internet to a remote facility using a service such as LiveVault.
• The backup selection list is periodically audited for completeness.
• Restoration of small files is tested periodically to ensure data can be recovered.
• However, full data restoration to a new server is not practiced.

It's possible, because a full recovery will not have been practiced, that some functionality will have to be recreated from scratch. Server recovery will be delayed for weeks as a disaster recovery location is secured, hardware is acquired, and problems are solved restoring server configurations to this new hardware.

The costs of Option 1 for a small- to mid-sized firm are a tape drive, tapes and software ($2,000 to $7,000) plus labor to install and maintain the system, $8,000/year.

Option 2) 1- to 2-Day Recovery Time

In addition to the costs of Option 1, this option costs very roughly $3,000 plus $5,000 per recovered server in labor, plus the cost of a duplicate tape drive ($1,000 to $5,000) and some spare workstations ($1,000 each) – one per recovered server. There is also a cost to maintain the procedure once developed, which depends on how much your network changes over time. Allow 30% of your initial cost per year for maintenance.

• You designate a branch office as a disaster recovery center.
• As practical, we use VMWare Server to consolidate your servers - reducing the amount of hardware required as well simplifying the recovery process
• In advance, our San Francisco computer consulting firm practices recovery by taking the tape to our office and rebuilding your servers on workstations. We work all the bugs out of this and write a procedure. The procedure is tested by someone other than the person who developed it to ensure that 2-day recovery can be achieved.
• You keep extra workstations at your disaster recovery location or assume they can be acquired on short notice.
• You keep a duplicate of your tape drive in this location. Tape drives are specialized items which can not generally be purchased at retail outlets.
• You ship the backup tape to the disaster recovery location every day.
• The recovery plan is periodically practiced and updated.

With the recovery procedure, a tape and the spare equipment in hand, any competent engineer can bring you up. So if your disaster recovery site is outside the Bay Area, you may not be affected by local manpower shortages.

Option 3) 10-Minute Recovery

This costs $3,000 plus $5,000 per recovered server in labor, plus the cost of duplicate hardware for all your servers ($4,000 to $8,000 each) plus $3,500 per server for replication software. Maintenance is about 30% of the initial labor per year.

• Use a co-location facility or branch office as your disaster recovery center.
• As practical, we use VMWare ESX Server to consolidate your servers - reducing the amount of hardware required as well simplifying the recovery process
• Install dormant, servers in the disaster recovery center
• Use special replication software from VisionCore or NSI Doubletake to keep the dormant servers synchronized with the live servers in virtually real time (with in a few minutes of each other)
• If the main servers are lost, log in to the disaster recovery center where it can quickly be brought to a production state.
• Test failover and failback from time to time.

This option is the most expensive but also results in the shortest recovery time and the lowest risk of data loss.

Option 4) Reduce the risk of catastrophic loss of your servers in the first place

For a one-time cost similar to the labor cost Option 2, you can relocate your servers to a safer location, such as a co-location facility – obviating much of the need to plan for catastrophic loss of your server room. You also don’t need to buy spare hardware.

• Move all servers to a co-location facility. Spend monthly co-location fees of about $500.
• Install a file server in the main office. Replicate basic files (Word, Excel, PowerPoint, graphics, etc) to/from this office to speed local access to files.
• Use Outlook 2003 or 2007 for email access; it works well over a WAN.
• Use Citrix / terminal services as required for access to accounting or other programs that don't work well over the WAN.
• At the co-location facility, provide sufficient backup disk storage to meet data archiving and hardware failure contingency needs.