Skip to content

Convergence between Cloud Infrastructure management and Data Centre management

The convergence – a reality

If we look at the evolution of private cloud, we can clearly see a natural convergence between Data Center Management (DCM) and IaaS management. It is natural because it fits better to the enterprise organization. Let’s be clear, we are talking about large organization, having already data centers for their own IT or businesses. The Services and software department are focusing on the elastic applications they can build on top of the infrastructure management such as Eucalyptus [1], OpenStack[2] or commercial solutions such as VCloud Director [3].

Usually, application guys start creating a PoC on Amazon or locally on few PCs with an open source IaaS and they can quickly show the real value of the elasticity to their top management. So, afterward, when the top management is convinced by the project, they will ask for selecting and deploying an IaaS management in order to let the application guys doing their elastic stuffs. However, guess who will be asked to deploy the IaaS management …? The data center management (DCM) guys, and those guys haven’t got the same needs as the application guys. Their objective is to make their lives easier but still while  managing a park of 50.000 machines.

Then, those DCM guys will look after machine controllers, statistic systems, orchestration, VM automatic configuration such as Puppet system [8], global deployment tools, a Governance framework for managing policies – access control – CMDB – VM templates – dependencies managements, and many other DCM features. Going further, a function of DCM is also evolving. Indeed, the Application Delivery Controller (ADC) is a component coming from the DCM side which aims to control, optimize, re-allocate all the incoming and internal traffic but also between DCs according to applications location. In addition it also manages the application content caching for faster delivery. This DCM optimization framework is also evolving to control elastic decisions.  This is the case of the Zeus ADC, positioning themselves as the intelligent component that lead the elasticity decision.

As a result we can see two major evolutions, (1) the IaaS management evolving toward DCM management, and (2) DCM management components evolving to control IaaS API and service management function (claudia). They are converging.

The missing piece in research and OS projects

If we look at the most known open source projects such as OpenNebula coming from the RESERVOIR[4] project, Eucalyptus coming from the university of California, OpenStack[5] from Rackspace and the NASA, all of them are focusing on the IaaS and none of them tackle the DCM issues we present here. That is clearly a missing piece for making them strong challengers in large organizations.

However, taken separately different components you should be able to build your own convergent solution, but that’s not really a product and the maintenance of the different elements, having different life cycles, must be done by yourself. I strongly believe that is show stopper for many organizations.

On the other hand we start seeing evolutions in commercial product such as BMC, CA,  3Tera or Novell by integrating pieces of governance like approval process and SOX compliance, but still missing a complete governance picture.

The SNIA CloudScape III: a similar conclusion

These 15th and 16th of March, I was at the SIENA[6] Cloud Scape III conference at Brussels organizing with the OGF[7]. The conference aimed at defining the next steps for fostering the cloud adoption by defining (or continuing defining) standards in cloud.

One of the key requirements expressing during the conference was about the Governance framework including: security, policy management, SLA, life-cycle management, service description and dependencies, etc. If we think about it, that is completely logic. In IT system the governance is probably the most important challenge that is not completely resolved yet. As the cloud conveys a certain complexity from in-premises IT to cloud, the governance will clearly be the next war in this field.

At the end of the day, we can see a general trend in cloud: simplifying the complexity, whether it comes from Data Center Management, or from IaaS management or even from a global governance view point. As the result the real market leaders will be the ones managing a successful convergence between existing tools, technology and methodology to manage and govern properly the cloud infrastructure.

References

[1] Eucalyptus, http://open.eucalyptus.com/

[2] OpenStack, http://www.openstack.org/

[3] VMWare VCloud Director, http://www.vmware.com/products/vcloud-director/overview.html

[4] RESERVOIR research project, http://www.reservoir-fp7.eu/

[5] The OpenNebula project, http://opennebula.org

[6] Standard for Interoperability for eInfrastructure implementation initiative, http://www.sienainitiative.eu

[7] The Open Grid Forum, http://www.gridforum.org

[8] The Puppet Master, http://www.puppetlabs.com/


 

Releated Posts

IEEE Big Data 2023 – A Summary

Our CTO, Sabri Skhiri, recently travelled to Sorrento for IEEE Big Data 2023. In this article, Sabri explores for you the various keynotes and talks that took place during the
Read More

Robust ML Approach for Screening MET Drug Candidates in Combination with Immune Checkpoint Inhibitors

Present study highlights the significance of dataset size in ICI microbiota models and presents a methodology to enhance the performances of a multi-cohort-based ML approach.
Read More