Keeping disaster at bay
- 14 February, 2007 12:00
Data loss and downtime is not just a dip on the power charts or loss of revenue: it could literally cost lives in the healthcare and aged care arena.
Ozcare knows all too well about the risk. The health and community service provider needed help finding a solution to deal with its data management and storage requirements, as well as potential disaster recovery (DR) and business continuity (BC) challenges. With 8000 clients and 70 locations throughout Queensland, and a good chunk of its offices in regional areas, the organisation's key requirements were to protect the fort, centrally store data and email, and add more storage muscle to the mix.
Enter Queensland-based system integrator, ordyss, which came onboard to help assess the organisation's mission-critical applications and the impact of downtime. The job, which took about three months, involved the implementation of Hitachi SAN hardware. ordyss general manager, Michael Cooke, said the upgrade gave Ozcare true availability while dealing with its data management and storage requirements.
"Ozcare's IT is distributed around the state. They have individual servers at remote locations, and they don't have an endless budget or IT staff," Cooke said. "The SAN helped centralise the technology, and offer up more control and consistency."
A key challenge was dealing with the business continuity and stability of services. Ozcare had broadly dispersed information across independent servers in branch locations. These prevented data consistency and made it increasingly hard to ensure system-wide, complete backup.
ordyss designed a geographically dispersed cluster running on Microsoft Windows Server 2003 Enterprise Edition. The cluster has the ability to deliver application failover between sites. Cooke said it maintained access to critical applications in the event of failure or disaster. "If head office fails, it seamlessly fails over to a secondary site," he said. This takes about two minutes. "It works as though all of the systems are in the same server room."
Remote branch servers were also relocated to a new data centre at Ozcare's Fortitude Valley site in Brisbane. The reseller handles the organisation's backup and offsite tape storage at its managed services centre, using the SATA features of the Hitachi modular storage. Similar stories are popping up in aged care and in healthcare because of growing regulatory conditions, Cooke said. New requirements released in 2003 required Ozcare to maintain all email records for a minimum of seven years.
"Disaster recovery is very topical in aged care, and healthcare, which has lots of challenges and where compliance is a big issue," Cooke said. "In healthcare and aged care, federal government funding is linked to how the organisation can demonstrate government compliance guidelines."
While healthcare and aged care are hot spots for the technology, the loss of data and the ensuing loss of revenue is also keeping CIOs awake at night. And while big enterprises are a likely spot for the technology, Cooke said the not-for-profit sector was also gobbling up the technology.
"The not-for profit category tends to be more savvy and more sophisticated than large corporates. If there's a pandemic, and data is lost, people could die," he said.
Corporate Australia, particularly the mid-market, is a big opportunity for DR and BC, Hitachi Data System's (HDS) marketing manager, Tim Smith, said. He claimed larger enterprises, including banks, telcos and many government sites, already had well established policies.
"The fastest growth area is the SME market, so don't treat the SME customer like a poor cousin," Smith said. They had the same worries over data loss and information management as the big guys - and they couldn't afford downtime either.
HDS had added enterprise class functionality into its mid-market gear. The AMS 200 entry-level product was a prime example, Smith said.
"This functionality only existed in the 500 and above models previously," he said.
Smith said ensuring all files were backed up and kept safe in a separate location was of paramount importance. This would be borne out in 2007 as more organisations put business continuity plans into place. "There's a definite fear of the potential loss of data and the direct business impact," he said. "This is a continuing trend because there's greater alignment with IT and business."
Page BreakConsulting services targeting disaster recovery and business continuity are the big money makers for the channel. Smith advised resellers to first assess the health of the storage environment. A poorly designed or managed storage infrastructure can put the business at risk in case of a catastrophic failure.
"Disaster happens because of human error or a virus -- rarely is it the air conditioner that leaks, which causes the IT cupboard or data centre to flood," he said. "There's a shift in Australia from simply being able to recover data to having those services available all of the time."
There's no doubt IT/storage managers continue to face major challenges efficiently executing a disaster recovery strategy, managing storage demands against the explosion of data, and making informed strategic decisions.
"Here's a horror story: a reseller went in to do a health check on a major bank and found 250 SANs. People were putting in baby SANs under their desks and connecting them to the network because they needed more capacity," Smith said.
Although DR had gained a reputation as an expensive technology, it didn't have to break the bank, he said. Smith recommended starting off small or opting for a phased-in approach.
"To have DR doesn't mean you need to invest huge bucks in a multisite solution. Start with the recovery side of backup and recovery," he said. "You can get a good data protection suite for under $10,000. If a fi le or the system is corrupted, make sure you recover it in minutes rather than hours or days."
IBM storage business unit executive, Francois Vazille, said the shift downward started about 12-15 months ago, with technology improvements such as data mirroring software now available in the mid-market. Advancements in virtualisation also meant any type of storage could be copied across different sites on myriad platforms.
Armed with the latest technology benefits, he recommended resellers help a company with business resiliency and continuity planning. The first step is to get a plan on the boardroom agenda; determine potential risks and opportunities; develop and implement a customised plan; and test, rehearse and review. Resellers can also help a company examine the mission-critical systems and processes to determine potential risks including security, availability and recovery requirements from the business perspective, Vazille said.
"As part of the BC process, make sure IT is aligned with the business. Some users want a full DR site, but hesitate when they see the cost. So start with different tiers of storage, and add more capabilities as you go," Vazille said.
Systems integrator, Global Storage, recently rolled out a new fully managed service in a bid to offer mid-sized businesses a fresh approach to disaster recovery. Dubbed storage environment replication services (ERS), it delivered benefits including enterprise class reliability, recovery and functionality, technical director, David Duncan, said. A big benefit is the fact that companies don't need to outlay the cost of the physical DR infrastructure upfront. The other plus is that the technology used is hardware agnostic.
The service, which is charged monthly, is based on three factors: amount of storage; number of servers; and service level requirements (response time).
"Large enterprise has the DR budget, but not the SME player, although they have the same DR requirement," Duncan said. Upfront costs aren't the only challenge for an SME. The inability to find a suitable second site, and having to sort through environmental factors such as air conditioning and clean power, was time consuming, Duncan said. "We take the headache away, give better geographical distance, and take away the management issues," he said. "We see this as a viable alternative to traditional DR.
"Local council is already interested, but we see the future in the SME commercial space. The larger SME is the cream - places with 100 or more staff."
NetApp marketing and alliances director, Mark Heers, said resellers undertaking storage consolidation jobs would be able to sell the DR concept as the logical next step. Resellers should initially help customers distinguish between the DR components versus the importance of a BC plan.
"DR is the IT part of a disaster issue," he said. "It's the second site, the equipment, and where the storage is the mechanism. BC is beyond IT; it's the planning and processes that keep the business going. "Most of the big guys - the banks and telcos - did this years ago, but it was too expensive for the middle player. Now it isn't. It's not so much the storage costs that have come down, but the communication (bandwidth) costs."
Symantec systems engineering manager, Pacific Region, Paul Lancaster, said many companies were opting for DR and BC in light of the stricter compliance and reporting push. Email management was also becoming a top risk. "Many businesses are run on email so we need to have 24/7 management over the email system," he said.
Mission critical application
Resellers are also transforming themselves into more of a trusted IT supplier, and DR/BC expert.
"DR is a critical part of BC, ensuring processes of a mission critical application and data work without interruption. Resellers are getting the message and becoming more business savvy, understanding the integrity and security of a company's data," Lancaster said.
He advised resellers to set realistic expectations about protecting the fort. "Make the application highly available - have the ability to failover from one server to another without the client seeing any downtime," he said.
"Have a fail-over from one site to another; ensuring the replication of data from one site to another, the non-production site."
Sun Microsystems channel director, Sam Srinivasan, said that while the mid-market was on a roll, he also still saw opportunity at the top end of town. He claimed he hadn't yet seen a huge upswing in investment in the mid-market.
"People want to implement DR, but often times only justify a business case for it once disaster hits or a mishap occurs," he said. Srinivasan said DR as a service, rather than a physical technology investment, made sense in the medium offi ce space and could help take-up.
"Organisations want the DR facility, but not always the control of it. They want to obtain a level of DR, but not the huge investment," he said.
Managed services also gave the mid-market comprehensive access to managed backup or a replicated database service. Users can now opt for an a la carte menu approach, making DR scalable.
At the top end of town, many companies are now also opting for a third site. Srinivasan said this was being driven by legislative changes. Again, he argued managed services were an attractive proposition in this situation.
For Ozcare, which opted for a mix of technology in-house and some managed services, there was no question the decision to boost its DR and BC factor helped the crew sleep a little easier at night, ordyss' Cooke said.