Merce is a software product. However, IS infrastructure does not become industrial-strength by choosing the right software product. Industrial-strength IS infrastructure is that which can cope with stringent demands on the following fronts:

• Reliability: Perhaps the most important requirement from industrial-strength IS infrastructure is reliability. All services must keep running, deliver acceptable performance, and not malfunction randomly. And the reliability of email infrastructure is perhaps the most difficult to maintain.
• Availability: This focuses on the specific area of service availability. All services must remain available in the presence of hardware and software faults. Availability is a key area of concern specially for key real-time business applications, i.e. share trading applications, and for WAN links.
• Scalability: Today, IS infrastructure faces constant scalability pressures to address increased usage and traffic, both genuine and spurious. Perhaps no other computer and network service faces this more than email. Today, the threat of junk mails challenges networking and email infrastructure to unprecedented levels. Between June and October 2006, worldwide spam traffice monitors reported a 3x rise in spam traffic. If spurious email traffic increases three times in three months, this alone may bring down some corporate email servers.
• Data protection: Data loss by accidental hardware failure is not considered acceptable at all today. Businesses today demand the same level of permanence from computer data that they do from paper files; in many cases, they demand more. But computer hard disks often fail within a few years, and the culture where each individual employee took his own backups has died completely. "No one has the time," we are always told. IS infrastructure must rise to this challenge.
• Security: The underlying concern for information security permeates the entire IS infrastructure landscape. Security has been discussed elsewhere in greater detail.
• A 24x7 world: This is now a fact of life. IS infrastructure now needs to work round the clock, much like telephones. IS Managers now understand that server downtimes need to be scheduled weeks in advance; even the "Sunday maintenance window" is not available in many enterprises. All usable backup strategies are online, automated, and near-real-time or real-time.

This is a complex set of requirements. All requirements must be met for unbroken stretches of many years. This cannot be met by just buying some hardware or software product, whatever their manufacturers may claim. Industrial-strength IS infrastructure has to be designed the way an architect designs a large building -- integration is a key challenge. Then an entire range of complex hardware and software products has to be purchased, configured, integrated and tested, brick by brick. Once this is done, the entire system needs to be monitored 24x7 using automated tools.

Therefore, the key attribute that distinguishes an industrial-strength IS infrastructure product like Merce is whether it has been deployed in demanding enterprise environments and integrated with the other sub-systems to deliver industrial-strength infrastructure.

Merce works with the rest of your IS infrastructure on all fronts to help you take your information assets to the level your enterprise deserves:

• Server-class hardware: Merce has been installed at various client sites on rack-mountable server-class hardware with symmetric multiprocessing (SMP), advanced disk controllers, RAID controllers, FC HBA support, GB NIC support, redundant power supply units, etc. The OS installation and system configuration processes involved for making full use of such hardware are more complex than for low-end computers.
• Advanced storage support: Merce has been configured at various sites to use SAN and NAS external storage; it manages multi-terabyte mail folders and directories of user data at those installations. These storage devices have been configured to use FC interfaces or the iSCSI protocol. External storage systems have been set up with RAID 5 or RAID 6 for data protection.
• High availability clusters: Merce servers have been set up in redundant HA clusters, where one server acts as hot standby for another. Data replication is either done through shared storage or by automatic application-level mirroring in Merce every few hours. In the simpler cases, Merce works on server pairs, where one server is purely a dedicated hot standby. In more complex server farms, various Merce components have been spread over a cluster of four or more servers, and more complex dependency matrix patterns have been set up for cross-failover among the servers.
• Firewalls and IDS: Without exception, all Merce-enabled enterprise networks till date are protected using firewalls. Various firewalls have been integrated into a Merce-enabled enterprise network, and are monitored with Merce Insight . IDS solutions have also been integrated.
• Large installations: Large Merce networks integrate dozens of servers delivering services to thousands of users. One of our first installations had one master server, twenty slave servers, and about ten Linux-based firewalls.
• Migration from legacy systems: Rarely is Merce the first messaging system used by our clients. In almost all deployment projects, our team has assisted the client in performing a 100% migration of all existing messages, mailboxes, address books, and other data from legacy systems to Merce. This has often been done without loss of a single email, and without the users being asked to mail systems in parallel. Effective project planning together with tools and technologies have allowed the migration to be executed within one weekend per site, even for sites with many hundred users. From Monday morning, users see a new messaging system with all old emails visible and no loss of email.
• Environmental controls: Merce servers have been integrated with compatible UPS systems, temperature sensors, etc. Servers shut down automatically if temperatures rise above pre-set values or UPS battery levels fall below a low-water mark. Such events are also monitored using Merce Insight. Some Merce servers run in unmanned server racks in remote locations without any local IT support staff, 24x7. At such installations, environment monitoring is needed to prevent cascading server failure in case of power problems or air-conditioning failure.

You cannot buy industrial-strength IS infrastructure. You must build it. Merce, and the technical team behind it, understands how such infrastructure is built.

At Merce, we do not believe in limiting our engagement with our clients to just delivering shrink-wrapped software CDs and manuals. Some of our customers will be comfortable handling the entire installation, configuration, testing and commissioning of their solutions themselves. Others seek our assistance and assurance in the rollout and continued maintenance processes.

The Merce Professional Services Group (PSG) works as implementation partners with our customers to handle the entire process of Merce deployment, from start to finish. The Merce Support Group works with existing customers to provide remote and on-site technical support the year round. We also have a network of Partners of all grades who take our prospective customer from early stage exploration to pre-sales detailing to closure, and provide ongoing on-site technical support after commissioning.

We believe in a high-involvement model to engage with the market. We want to be present to ensure that our customers do not have any cause for complaint. Modern IS infrastructure is getting increasingly complex; your business or academic institution should not need to acquire expertise in these areas to get the benefit of good IS infrastructure. Merce PSG and our Partner network will work with you.