Lead

Online companies ga-ga over networked storage

Redundancy, zero downtime and reliable backup are the key factors driving online businesses to adopted networked storage solutions, discovers Megha Banduni.

The storage requirements of a Web-based business differs from that of a conventional enterprise. For a portal or e-commerce Web store, storage is an extension of its core business.

Although vendors have identified Storage Area Network (SAN) as the best storage solution for online businesses, Internet portals prefer Network Attached Storage (NAS). While some Internet companies have gone in for a mix-and-match option.

"The goal was to implement a solution that would enable us to consolidate equipment, simplify management, and optimise our storage utilisation while giving us the flexibility to expand capacity as needed" - Preeti Desai COO Digital Business Hungama

Hungama, for instance, shifted from Direct Attached Storage (DAS) to NAS. It deployed dual-switch redundant NAS boxes with 6 TB capacity connected to the company’s Gigabit Ethernet network. “The goal was to implement a solution that would enable us to consolidate equipment, simplify management, and optimise our storage utilisation while giving us the flexibility to expand capacity as needed,” explains Preeti Desai, COO, Digital Business, Hungama.

At Hungama, both Network File System (NFS for Unix) and Common Internet File System (CIFS for Windows Server) file protocols are used to access the NAS boxes.

“We are currently working with two vendors regarding superior file serving performance deliveries by leveraging our understanding of consumer patterns to standard server, storage and systems management infrastructure. The goals of the project are to consolidate file servers by 25 to 40 percent, reduce downtime for user file shares to zero, decrease file server administration workloads by 70 percent and decrease back-up jobs by 90 percent,” adds Desai.

The company has benefited in terms of better scalability and flexibility for its file serving environment. The storage solutions provide better modularity and freedom from proprietary hardware that is typical of stand-alone appliances.

Further, Hungama is customising its set-up to create an easy-to-use, highly available system. It comprises a true symmetric Cluster File System (CFS) that enables scalable data sharing, high availability services that increase system uptime and cluster and storage management capabilities for managing servers and storage as one.

Anand Mittal, COO, People Group, the parent company of portals such as Shaadi.com, Fropper.com, Astrolife.com states, “In order to protect ourselves from natural calamities, a weekly backup is stored away from the data centre at Iron Mountain, Washington D.C., USA for four weeks. Iron Mountain securely stores and protects data for over half the Fortune 500 companies,” he adds.

As the online consumer base broadens, the need to serve these customers and offer online products and services 24x7, while maintaining ease of manageability with stability are prompting CIOs of Web-based companies to go in for networked storage.

"The key factor driving demand is the increased need for dedicated storage solutions among enterprise customers. Data is growing exponentially due to rise in e-mail and database requirements" - Manoj Suvarna Country Manager HP StorageWorks Division Technology Solutions Group, HP

Manoj Suvarna, Country Manager for HP StorageWorks Division, Technology Solutions Group, HP India says that currently the market for online storage in India is a little over $100 million and is growing at more than 20 percent through 2009. “The key factor driving demand is the increased need for dedicated storage solutions among enterprise customers. Data is growing exponentially due to rise in e-mail and database requirements,” he adds.

As far as choosing from DAS, NAS or SAN, Indian online players largely prefer NAS. Soumitra Agarwal, Marketing Director, NetApp feels that most online service providers prefer NAS, since it is scalable without adding significantly to the fabric cost, working as it does over Ethernet. He adds, “Online storage is usually based on NAS protocols like NFS which is scalable to service large server farms without adding significant fabric costs. Most online services use an Open Source server OS like Linux. Increasingly, we will see adoption of storage grids, for which NetApp offers its Data OnTap GX system. The storage fabric (server-storage connectivity) is usually Gigabit Ethernet and increasingly, we will see 10G Ethernet being deployed to support the massive I/O loads in the online storage business.”

Much depends upon the kind of applications that the users have deployed. “If you have a large database and run too many applications on your system and require high performance, SAN is the best option. Files that need to be shared with different people and departments like software development, designing will require NAS,” says Manish Bapat, National Manager, NAS and CAS, EMC India.

Intelligent backup

Online companies are finding the need for more sophisticated backup solutions.

Explains Vivekanand Venugopal, Director, Software Solutions, APAC, Hitachi Data Systems, “Traditional backup doesn’t provide fast, reliable restoration from tape. Hence we believe in following Distributed Index architecture that allows disk-to-disk backup.”

“Virtual tape library is the key driver behind online storage picking up. Today tape libraries are becoming more intelligent and capable of leveraging the true value of disk. “We have a technology called deduplication, where the virtual tape library will store only the changed data instead of storing the same data again. It will not save data that is already stored,” adds Venugopal.

Shailesh Agarwal, Country Manager, Storage, IBM India elaborates, “Data backup has three elements—backup device, software and media. All backup devices are compatible with backup software and media. The trend is to make use of automated backup, in which the data is restored from the server automatically. This technology has matured in India. The best practice is to make a copy of the latest data in the disk itself and transfer outdated data from disk to tape.”

However, Bapat of EMC makes a valid point when considering backup technology, “Some research shows that out of the backups taken, half are never used. So, organisations need to evaluate what data they want to backup and for how long they want to retain these backups.” He identifies two challenges associated with backup—the data becomes old after some time and inaccessible after a long time, and recovery of data takes a long time. “Cache technology avoids these problems. The current data can be stored in disks and then after one to two weeks needs to be moved to tape. Disk-to-tape will help in solving these problems,” adds Bapat.

With clients like Yahoo, India Times and Rediff, NetApp offers SnapShots capability, which along with SnapRestore, enables online companies to maintain RPO or RTO (Recovery Point Objective or Recovery Time Objective. “It is important for online service providers to consider that SnapShots should not degrade NFS operations (which is a critical performance parameter) and the NetApp SnapShot technology assures that. SnapShots are important since in case of failure, the service provider should be able to recover data at least to the last SnapShot to maintain SLAs,” comments Agarwal of NetApp.

Security and storage go hand-in-hand

When we talk about storage, security cannot remain untouched. Sending sensitive information over Internet is risky. In such a scenario, encrypting data in transit and on the storage media is one way out. The technology known as ‘Storage Encryption’ can make a CIO’s concern little lesser by ensuring that only authorised employees can access it.

Vendors as well as users feel that the threat of internal security breaches is greater than external threats or the threat posed by spam. Today thefts occur during transportation of corporate data to a storage location or while bringing it back for data restoration from an on-site storage vault. This has led to leakage of sensitive information such as credit card details, bank account numbers and medical records.

Says Mittal of People Group, “Online transactions are always at risk. Our online storage system does support security features. We have implemented security systems from Cisco and regularly carry out penetration testing and vulnerability assessment to quantify our information risk and ensure protection from possible attacks.”

“Today the biggest issue in front of organisations is to safeguard their data from malicious attacks. The research shows that 75 percent of threat is internal. EMC has a solution called Centera because of which nobody can modify, delete or change the data,” explains Bapat of EMC.

NetApp has identified security solutions such as authentication, authorisation, accounting, anti-virus, and anti-spam, for online companies. Basic security features are built into NetApp’s Data OnTap storage product.

With the need to reach out to more customers, online businesses are increasing their service portfolios, which in turn is increasing their storage requirements. As Desai of Hungama pointed out, “We have witnessed a huge rise in the number of people accessing our portal, that has resulted in the growth of content, and it becomes quite difficult to manage if you do not have a good storage system. Hence we shifted from DAS to NAS. Besides, scalability, availability and better quality were the other reasons for the shift.”