Was invited to talk at Datacenter Dynamics Converged 2013.
Very interesting conference, indeed. Latest trends, issues, solutions, hard and software based around Data Centers – all from the perspective of Data Center Operators.
Cloud Computing in this case is the workload and there was frequent mentioning of Data Center Infrastructure Management (DCIM), a topic that we, the ICCLab, address in our new project GEYSER from an integrated Cloud Computing and Energy Efficiency perspective. This is a hot (pun?) topic these days and I am quite keen to see substance around this. Can’t tell much more at this stage, but with GEYSER and a new and excellent chap joining our team for that topic we should have all things together.
Here are the slides of the talk for download.
Appointed to General Chair of the 2nd international workshop on Data Center Performance, DCPerf 2012
Data centers have become an integral part of the backbone infrastructure for information technology. In particular, they provide a cost-efficient solution for provisioning a wide range of computing resources in diverse environments such as for business, scientific, or mobile purposes. Owing to the rapid growth of user-defined and user-generated programs, applications and content, the complexity of deploying and operating data centers is expected to increase. Considering the high volume of mixed workloads, the diversity of services offered, and the increasing mobility of users across geographically distributed areas, the performance optimization of data centers has become ever more necessary and challenging, especially concerning criteria such as scalability, reliability, manageability, power efficiency, area density, and operating costs. As a result, adaptive and autonomic optimization strategies are essential for managing the multi-dimensional complexity of data center performance.
The goal of this workshop is to promote a community-wide discussion to find and identify suitable strategies to enable effective and scalable data center performance optimizations. We are looking for papers that present new techniques, introduce new methodologies, propose new research directions, or discuss strategies for resolving open performance problems at all layers of a data center.
GEYSERS was presenting at the Cloud Computing and Cross Stratum Optimization Workshop, organised by the IETF Cross Stratum Optimization Task Force.
Another source on cloud networking. Slides accessible online.
The Funkschau special edition “Datacenter 2011″ features an article about *server access technologies* (network design) within datacenters. The author elaborates on pros/cons of ToR vs EoR/MoR designs. Having read a couple of papers about this topic, to me, Top-or-Rack (ToR) seemed to be de-facto industry standard. Apparently, this is not the case.
Browsing the Web for some additional info I came across an excellent source on the overall topic (networks/datacenters/cloud computing/etc). Check, for instance, ToR vs EoR for a lengthy discussion on datacenter network design choices.
Another good source on the same / a similar topic