CBL Data Recovery Media Coverage

Need More Information? Have a look a CBL's Data Recovery Articles, Press Releases and CBL News Features

Stay Sane With SAN: Article On RAID Data Recovery Prevention

Data Recovery Article: Business Solutions Magazine – October 15, 2001

RAID controllers are the SANE (SAN enabler) component of SANs, according to Gentry Ganote, president and CEO of 75-employee GAIN Systems, Inc. (Norcross, GA). From his experience installing SANs over the past year, he has identified the four key features that must be present in a SAN. These features are reliability, functionality, performance, and manageability. An examination of these four features, plus consulting the SAN lessons presented in the February 15th issue of Business Solutions For Mass Storage, may help avoid VAR inSANity.

Reliability: Why Store It, If You Can’t Retrieve It?
If storage is important enough to create a whole network for it, the data must be awfully important. RAID systems employ multiple hard disks to increase MTBF (mean time between failures). This increases fault tolerance, which is also known as data protection. Fault tolerance is important in the event of a component or I/O (input/output) path failure. “To address fault tolerance, all of our systems are designed with redundant SANEs (SAN enablers). The Infortrend RAID controllers are the basic building block for our SANEs,” said Ganote.

RAID controllers are electronic devices that provide the interface between the host computer and the disk array. From the viewpoint of the host computer, the RAID controller makes the array of disks look like one or more virtual disks. RAID controllers sense drive failures and can reconstruct data if needed. They write data to the disk array according to the RAID level configured for the application.

RAID levels are part of the custom fit for each application. This is a good thing, because every SAN is different, based on each customer’s individual needs. “Most of the RAID systems we install are configured in either RAID 1 or RAID 5, so the drives are redundant. RAID 5 works well in high-end database applications. RAID 1 provides high data reliability, but the cost is very high.”

For all that, you can have the most reliable RAID system ever invented, but if your power supply isn’t failproof, you’re sunk. Additional features such as cooling systems and monitoring circuitry will ensure that the RAID stays up and running. RAID arrays generate a lot of heat and need internal fans to keep them cool. If one of the fans stops, the temperature will rise and drives will fail. One of the functions of monitoring circuitry alerts administrators if a fan malfunctions.

Functionality: Run The SAN So It Actually Works
RAID controllers enable all the popular RAID levels and provide monitoring. “RAID controllers from Infortrend (Santa Rosa, CA) provide all the worldwide name mapping features of a SAN to us,” said Ganote.

Just what is worldwide name mapping, and why is it important? Ganote explained, “When we talk about SAN, we’re talking about Fibre Channel. With a SAN, host computers (servers) talk to storage via Fibre Channel HBAs (host bus adapters).” Worldwide name mapping makes sure that not all servers on the SAN try to control the virtual disk that holds the storage. Employing worldwide name mapping is valuable in a heterogeneous environment where you have a SUN box, an NT box, and a Linux box. The NT tries to grab all the virtual disks it sees. In a SAN, using worldwide name mapping, you can control which server can access each virtual drive presented by the RAID controller.

Performance: Make The Network Run Faster Than Before The Installation
When customers decide to invest in a SAN, the last thing they want is to give up performance and speed. Since every implementation is different, a knowledgeable VAR is needed to help customers decide which combination of controllers and drives will achieve the best performance. And that’s how a VAR adds value.

“An integrator must understand the customer’s network environment and the requirement of each host that is connected to the SAN,” warned Ganote. By using RAID in a SAN, network I/O operations can overlap in a balanced way.

RAID employs the technique of striping, which involves partitioning each drive’s storage space into units. These units range from a sector (512 bytes) up to several megabytes. The stripes of all the disks are interleaved and addressed in order. When data is striped across drives, all disks can be read at the same time for rapid retrieval.

More drives in the RAID array will result in greater performance. To manage this, you’ll need to integrate more RAID controllers in the network. These controllers will increase bandwidth, allowing more information to funnel through the system at once. This will also increase performance.

Manageability: Software Tools Will Monitor Health Of Storage
SANs are typically heterogeneous, meaning they consist of several manufacturers’ products. This is why they are complicated to implement and manage. It’s impossible to have a SAN without adequate management tools. These tools are found in the form of storage management software. “We use RAID controllers by Infortrend and then run our own software over that,” said Ganote. “Our storage solution is heterogeneous. This allows us to manage the SAN from any platform (computer system), whether it is Solaris, NT, AIX, or Linux.”

“It’s important to be able to monitor the health and status of your storage from anywhere in the network,” said Ganote. “We’ve implemented an alert mechanism that can be monitored through an SNMP (simple network management protocol) program like Hewlett-Packard’s OpenView. The SAN needs to be managed because if you have a drive failure, your degraded array must be rebuilt and configured properly. This way the customer won’t have any loss of data. These days, everybody needs to be up and running 24/7. Understanding the health of your storage is critical.”

If you’re going to implement SAN solutions, you must know RAID. But there are many other components of hardware and software, including the customer’s legacy equipment that you must consider. Perhaps the key to implementing a SAN is to take it in small portions, learn each component, and consider how all the components work together. When you imagine the entire storage area network, it can be overwhelming. A wise person once said, “You can’t eat an elephant in one bite.” Taking it bit by bit should help keep you sane when implementing SAN solutions.

The Science Behind Data Recovery
What if you’ve implemented a SAN with RAID storage and the physical media becomes damaged? Data striping spreads a record over multiple hard drives, but what if something goes wrong with the hard drives?

The computer data that seems solid and reliable when you see it on a monitor or printout really only exists as minute electrical impulses. These impulses cram tightly on tape cartridges, diskettes, or hard drives. Depending on how it is stored, that data will continue to exist until the medium is destroyed, scrambled by a virus, or overwritten by other information. In other words, some “physical” event prevents you from accessing your data. Data recovery companies work to reverse that process any way they can.

There is no single method to retrieve data. In fact, it is extremely rare to apply a solution more than once. To avoid damage from repeated scanning, a skilled data recovery company will find a way to duplicate client data, even if it must be done bit by bit.

Hard disk drives store data on metal oxide platters spinning as fast as 10,000 revolutions a minute. While they spin, an actuator arm reads and writes magnetic charges one millionth of an inch above the surface. Even though any contact can and does destroy data, a great deal of information can still be harvested from physically damaged media. The manufacturers of storage media, like hard drives, are constantly striving to compress more data onto smaller surfaces. A data recovery specialist must have a “clean room” to avoid dust or dirt. Contaminants can cause more damage to the delicate magnetic media. Data recovery not only requires specialized equipment, but also requires people who possess advanced technical skills. More importantly, those people must have the creativity to apply their skills in unique and demanding situations.