Zeus Technology Journal on Ulitzer

Zeus Technology, the Anglo-American concern whose web server ran 3% of the world's web sites back before the dot.com bubble burst, third in the race behind Apache and Microsoft, was asking around among CIOs, IT director and senior IT managers at large organizations and came to find out that although 75% of them, give or take, are either in the cloud or soon will be, only 27% have a solution in place to manage it and so are basically flying without a net.

Which is bully for Zeus because it's got just the ticket - or soon will have - in its back pocket and claims to be the only application traffic manager that can handle virtual resources and the cloud because it's pure software.

In 2004 Zeus released Zeus Extensible Traffic Manager (ZXTM), based on its earlier Zeus Load Balancer to make up for its web server losses. ZXTM manages application traffic, inspecting, transforming and routing requests as it load balances them across the application infrastructure. Building on this widgetry - which is used by a lot of brand names - it's gonna release the Zeus Cloud Traffic Manager in Q1, which is supposed to work in all clouds.

The widgetry is supposed to monitor cloud usage in real-time and manage application traffic up there. Once you have some idea what resources are being used, you can begin to manage traffic and control costs. It says as the more dynamic web applications and services are expected in the cloud, traffic management will grow in importance.

It means to offer policy-based traffic management and a control point for transactions with the cloud based on availability, cost, performance and locale.

First, as a stepping stone, it's just released its L4-L7 Traffic Manager 6.0, which is supposed to make sure enterprise web apps are always fast and available even during peak demand across physical, virtual and cloud environments no matter how massive.

It's got a bunch of new features like multi-hosted IP addresses to scale clusters more easily and solid state caching to improve performance by caching more content more cheaply.

It's supposed to log every transaction failure and fully explain why any connection was terminated prematurely. It's also location-sensitive, applying different policies to users from different locations like restricting content for non-export locales or redirecting traffic to a more local data center. It says large responses won't be buffered up until processing has completed any more; data can be flushed and written to the client when ready. It also got fault tolerance, a new SOAP API and bandwidth management.

It runs on Linux, Solaris, VMware, Xen, Oracle and Amazon's EC2 and there are virtual appliances for Oracle VM, VMware vSphere, Citrix XenServer and Amazon Machine Images (AMIs). It supports Microsoft's Hyper-V as a Linux guest.

Zeus is 100% channel. It's got strategic partnerships with Dell, HP, IBM, Intel, Joyent, Oracle, Sun, Qualcomm and VMware and customers include the BBC, BT, Comic Relief, Domino's Pizza, ITV, Play.com and Virgin Media.