Dealing with the headache of business critical systems management
The widespread proliferation of smart phones, Blackberries, instant messaging, IP telephony and e-mail as legitimate business communication channels is putting increasing pressure on companies’ IT resources. Faced with this increased responsibility for managing these business-critical applications and with no extra, or an even tighter, budget to support it, IT departments may need to reassess their operating procedures to achieve the balance they need. Not only is there an issue of providing the bandwidth to handle video, voice and data, there is also the knock-on effect of managing these resources alongside the regular demands of the network. This places a massive strain on IT resources, both in terms of the people managing and controlling the conditions within environments such as the data centre. Particular challenges include: Home working The security implications associated with these trends are, in themselves, a major challenge for IT teams to overcome, but one the modern enterprise must address. With the lights out it's less dangerous Under such circumstances, there are a number of key factors influencing the growth in popularity of centralised, remote management tools. For example, with more servers and serial devices to manage, remote management enables administrators to gain secure, centralised systems access from any location through a standard web interface. The ability to identify and resolve issues without requiring a physical presence in the data centre or at the appliance, a significant impact on response times and helps the IT administrators and managers to meet the business’ growing demand for real-time resolution. The question of who should have access to what in the data centre is growing in importance thanks, in no small part, to the demands of regulatory compliance. However, the true extent to which remote management technologies can resolve the issue has yet to be explored to any great extent. The ultimate goal of data centre managers is to create a ‘lights-out’ scenario, controlled remotely and without any need for physical access, but there have been a number of stumbling blocks in attempts to achieve this. Feeling faint? Such deployments increase the infrastructure's capacity but they also have a huge impact on the ambient temperature of the facilities and the cooling measures required in order to keep the system up and running. To further compound this, blades are now being shipped and / or deployed in pre-configured racks, custom-built to maximise available space in the data centre, but leaving no place for conventional server management tools such as at-the-rack KVM switches. Because of this, there is a growing demand for environmental monitoring facilities, especially granular tools such as IPMI. Already embedded within a significant proportion of the servers currently being shipped and implemented in the data centre, it remains a largely untapped resource. However, as blade servers become increasingly common place, the ability for administrators to exploit this functionality and gain greater control will play a key role in managing servers and maintaining availability. Administrators need to know about potential problems before they impact the business; and hardware monitoring, both internal and external to the server, is paramount to enable more proactive data centre management. But the future role of IPMI may not just be limited to monitoring components of the IT infrastructure. There is a compelling case for introducing embedded monitoring capabilities into other elements of the data centre to increase the level of control that can be exercised over its non-tech components. For example, to attain true lights-out status, it is essential that the data centre manager is able to control every aspect of the facilities without physical intervention. But how does that individual know if the air conditioning units are on the verge of a major malfunction? They will be able to proactively monitor every aspect of the IT infrastructure but will have no insight into the status of the cooling facilities which are a fundamental factor in keeping systems up and running. Typically, air conditioning is managed by facilities rather than IT but, should embedded monitoring capabilities be introduced to this vital aspect of the data centre environment, it would be a significant step closer to the lights-out infrastructure. And the same approach could be adopted for other aspects, such as the internal conditions of server racks and CCTV. Developments of this nature would add a further level of control and reduce the financial cost of downtime. The range of issues that can be resolved remotely is such that only a hardware failure should result in the need for physical entry into the data centre. Using remote management tools enables organisations to restrict access to authorised personnel in these circumstances, but it also creates audit trails to support businesses in compliance with regulations such as Sarbanes Oxley and Basel II. By providing a company standard portal to manage every network device, it also forces people to go through certain processes, thereby facilitating further compliance with standards such as ITIL. One of the other critical areas that data centre managers need to examine is how they go about resourcing at a human level. We've already established that budgets are likely to shrink, making additional staff recruitment highly unlikely. Instead, proof of expertise is likely to be required in the form of training and certification of staff, with access, roles and responsibilities determined by relative, accredited skill sets. It's highly advisable at this stage to look at the various accreditation schemes offered by vendors and service providers and to plan resources accordingly. As I outlined at the beginning of this piece, shrinking budgets and increased levels of responsibility for business critical applications are causing IT managers to re-evaluate their systems and resources. The flexibility of remote management technologies, both in terms of supporting future expansion and the development of new applications, the ability to increase data centre security and monitor environmental conditions could provide them with a cost effective, long-term strategic solution to support the ongoing management of the infrastructure. Marc Marazzi is EMEA director of marketing at Avocent, a supplier of connectivity solutions for enterprise data centres, branch offices and small to medium size businesses worldwide. http://www.avocent.co.uk/
•Date: 21st Dec 2006• Region: World •Type: Article •Topic: IT continuity |
