The current e-mail business continuity methods deployed by today's organisations are expensive, cumbersome and more focused on disaster recovery than business continuity. A new, more effective and affordable approach is needed which also puts continuity in the reach of any IT budget.
Ask any senior manager or business owner: "How important is daily access to your e-mails?" and the answer would range between 'mission-critical' to 'we will lose R1 million per hour in lost revenue and productivity if our e-mail system goes down'!
We all know that communications is the lifeblood of any business - and e-mail has become the favoured tool in our fast-paced business world. E-mail is the new paper, now used for over 80% of written business communication. That is why I say, 'e-mail continuity = business continuity'. E-mail is arguably the most pervasive and mission-critical application of all. E-mail however has its challenges:
* Virus/worm transmittals via e-mail are an everyday threat.
* Spam - hours a day can be wasted identifying legitimate e-mails.
* Legitimate e-mails go missing (false positives) and can be hugely expensive.
* E-mail failure can halt e-mail communication with the outside world .
* Large attachments slow up everyone's work.
* IT support costs continue to rise.
* Regulations dictating e-mail storage place a tremendous burden upon companies.
According to recent international research, there is a 75% chance of unplanned e-mail outage and 14% chance of planned outages in any given year for a company. Put into perspective, the average outage in surveyed companies was 32,1 hours long.
The problem with existing solutions
Most large companies spend millions of rands each year on expensive on-premise e-mail clustering, mirroring, archiving, security and business continuity solutions. Smaller companies are often left in the cold due to the high costs. Expensive technical experts are required to integrate all the various components into the e-mail environment.
Every so often they restore their backups or conduct 'what-if' scenarios to test their contingency plans. On paper everything looks rosy and they may even pass the odd risk assessment with flying colours. In reality though when disaster strikes, it is often a very different picture. Current on-premise solutions simply move the problem to another area of the network. Here are a few scenarios of what can go wrong with traditional solutions.
Failure Point 1: Server hardware crashes
(35% of all downtime. Average outage 18 h)
* Problematic server upgrades tend to be the main cause, but hardware failure (disk, RAM) are major contributors.
* Reundancy is often built into servers at the HQ level, but generally neglected at the branch level.
Failure Point 2: Connectivity loss
(19% of all downtime. Average outage 27 h)
* LAN, WAN and telco outages are unpredictable and costly.
* This figure can well be higher in South Africa.
Failure Point 3: Replicated database and active directory corruption
(16% of all downtime. Average outage 9 h)
* When a corruption occurs in a database store or active directory, it can cause a main e-mail server to go offline.
* In most cases, replication software, which transfers data byte-by-byte, will copy the corrupted data to the backup server.
* Typically, corruptions are a slow process of degradation and may require administrators to restore many backup tapes until a tape is found before the corruption.
Failure Point 4: SAN complexity
(16% of downtime. Average outage 25 h)
* Complexity makes systems more prone to failure, more difficult to test, and more dependent on key resources that can be deployed on higher value problems.
* Increasingly complex SAN hardware, often used for primary and backup data stores, is becoming a common failure point for enterprise e-mail systems.
Failure Point 5: Natural and other disasters
(14% of downtime. Average outage 61 h)
* Natural disasters are not just floods and earthquakes, but also cable failure due to roadworks, electricity outages due to storms, flooding due to burst pipes, lightning and even theft of equipment.
Failure Point 6: Single platform dependency
* While most organisations depend on backup e-mail systems, the secondary systems are usually on the same e-mail platform as the primary system.
* This dependency on a single platform creates a point-of-failure where a virus, worm, or bug can incapacitate both the primary and backup system simultaneously.
Failure Point 7: Dependency on tape
* The very nature of tape backup is just that: 'backup'.
* Where tape backup fails as an e-mail continuity and recovery solution, is the fact that it takes anywhere from hours to days to recover a company's data from tape.
* If used as an e-mail continuity option, tape backup is too slow to meet reasonable recovery goals.
The alternative - SaaS
Recently, a research firm took a deeper look at the costs associated with on-premise and managed services approaches to the problem and built TCO models to compare the true costs of the solutions over their lifetime of use. Variables such as purchase price, maintenance, platform installation, network integration, configuration and monitoring were analysed to calculate the TCO in both large and small network environments. The figures were surprising! For the smaller company (300 users), a managed service was 10% cheaper. For the larger company (10 000 users) the managed service was 32% cheaper.
According to Gartner the concept of software as a service (SaaS) is gaining enormous popularity in European and other overseas markets. Gartner predicts that by 2009, 50% of all US companies deploying continuity systems will opt for a managed service.
I have noticed one or two 'value-add' companies with world-wide redundant data centres offering this service locally. They offer an instantly available live standby mail server at the perimeter, accepting and sending e-mail until the primary e-mail server is recovered. Another option is to have all specified e-mail records automatically and securely stored off-site. In the event of a total disaster, employees can still access and send e-mails from any location in the world. Their mails are still protected against security threats and archived as per the normal company policies. Best of all is that the impact of the disaster is not even noticed by the company's customers and suppliers. This in my mind seems to be a far better and often less costly approach to e-mail and business continuity in general. The cries coming from the IT trenches in times of trouble should be 'business as usual' not 'give us two days to restore'.