Sunday 6 March 2011

Cloud initiatives from IBM

2011 has been dubbed by some as the year of the cloud – the year when cloud computing comes off the PowerPoint slides and onto a computer near you. Others have muttered wisely about similarities with client-server computing, which for many years was the ‘technology of the future’ at every presentation and never quite arrived the way we were promised.

Anyway, this week saw announcements from IBM about cloud computing. It announced Tivoli Provisioning Manager 7.2, which creates virtual server images in minutes and automates the management of virtual environments and hybrid cloud systems. The software is also claimed to help to better protect data in the cloud. And the software provides ‘image management’ of a virtual environment, which means that as computing needs change the number of virtual machines can be scaled up and down.

Provisioning Manager works stand-alone or with other Tivoli management products. It works across Power-based servers and Intel x86 instruction set-based servers. It can manage LPARs (logical partitions) or WPARs (workload partitions) on AIX servers. It also manages virtual machine images provided by VMware ESX Server, Citrix Systems XenServer, and Red Hat’s open source KVM (but not Microsoft’s Hyper-V, yet) on x86 servers.

The beta availability of new hybrid cloud management capabilities were announced by IBM at the PULSE 2011 conference. Tivoli Provisioning Manager and other components can manage virtual machines across an internal enterprise cloud and an external public cloud. However, this doesn’t extend to other public clouds, yet.

Tivoli Storage Manager for Virtual Environments (also in beta) utilizes VMware’s vStorage APIs for Data Protection. By doing so, it offloads the back-up function from the virtual machine and ESX Server host to a VMware vStorage back-up server. Storage Manager is then able to request more frequent back-ups, ensuring greater data protection without placing excessive demand on virtualized system resources. It also allows smaller and faster back-ups that replicate only data blocks that have changed since the previous back-up by making use of VMware’s Changed Block Tracking.

The same week saw a discussion about cloud computing at the CeBIT expo in Hannover in Germany. There it was suggested that the adoption of cloud computing was being held back by issues such as security and reliability. One solution to the security issue was to use private clouds or shared private clouds. Another challenge that was highlighted was bandwidth (or the lack of it). They focused on the amount of data that would need to be moved around in a cloud environment might be too much for the existing infrastructure. The fourth issue the conference identified was interoperability – and the need for standards was highlighted. My personal opinion is that too often computing (and technology in general) has been held back by standards – the very fact that there are far too many of them! Have you ever tried to use someone else’s phone charger to charge your phone, only to find it has a completely different connector (let alone different amps and volts, etc). But you remember SOA and the plethora of standards, etc – OK rant over.

I’m sure this blog will return to cloud announcements, developments, and set-backs many times during this year!

No comments: