Integrated management is at the heart of Merce's design philosophy.

IS management skills: Fast-growing organisations are under constant pressure due to the scarcity of IT management resources. In-house IT infrastructure management teams are difficult and expensive to recruit, train, and retain. Such teams often waste their time and energy performing routine management and maintenance tasks on a large set of heterogenous servers, when their expertise would have been better utilised handling the technically challenging tight spots. Even conceptually simple tasks like backups of a dozen different servers with subtly different flavours of operating systems can become a laborious and error-prone manual exercise. User account provisioning across a similar server community can be even more difficult to manage.

Unix -- and now, Linux -- has been the server platform of choice for network services for almost three decades. About 75% of the Web servers on the Internet today are running some flavour of Unix or Linux. In the mid and late nineties, there was speculation about more user-friendly server operating systems seizing the leadership from the Unix family, but this promise has been belied.

How do you scale it up? The classic server operating systems do not allow easy integrated management of a farm of servers distributed over slow or unreliable WAN links. Servers in a high-speed cluster within one data centre can be integrated acceptably well. But modern mid-sized enterprises often cannot operate out of a single data centre. When you try to scale up to multi-site IS infrastructure, your management resource requirements do not scale up. If you have five small factories and each has a small data centre with a few servers, will you need five separate sets of system administrators for these factories?

The challenge of scaling up scarce and expensive IS management teams usually leads to under-staffed IS teams and poor infrastructure management.

An integrated management layer: Managing a network of such servers requires an integrated management interface. For instance, it should be possible to create a new user’s account across an enterprise network, and not on each Unix server separately. Unix has some built-in tools which help in integrated management, e.g. NIS, but they do not work well under all circumstances. And they do not provide some much-needed features, e.g. a unified UI for the administrator, unified log reporting of network usage per user, and on-line real-time network and server monitoring with exceptions and alerts. Merce fills these gaps. Unix/Linux and MERCE together make for a winning combination for managing small or large enterprise networks across a country or multiple continents, with multiple servers handling very large numbers of users.

Integrated management as offered by Merce must be distinguished from the common industry buzzword: network management systems, or NMS. The term NMS as understood by the industry refers to a system which usually monitors network components and servers and logs and reports observations. This is a far cry from the integrated management offered by Merce. Merce does not aim to monitor individual attributes (like CPU load) of individual components (like servers). Merce takes a much more high-level view of IS infrastructure as a whole, and offers the enterprise IS manager an interface to manage enterprise network services as a whole. Very often, the enterprise IS manager does not need to think in terms of individual servers or routers. He can think of higher-level concepts like "users", "locations", and "total mail flow".

What does the IS Manager want to focus on: the forest or the trees? With Merce, he does not have to choose.