Enterprise Strategy Group analyst Nik Rouda says that even though Hadoop and big data can be confused as being...
the same thing, they are much different. Big data is an approach to understanding situations and Hadoop offers a set of data management tools to help with that process. But is Hadoop the only option? Rouda says no. Apache Spark is a new data analytics engine that Rouda says is a viable alternative to Hadoop. Some of the benefits he says Spark offers include:
- Familiarity to developers who already know Java, Scala or Python;
- Flexibility to handle SQL queries, streaming, machine learning, and graph analytics; and,
- Adaptability to support different data sources.
IBM has invested in Spark, a move that Rouda says will encourage more development, capabilities and maturity for the project.
Read more about why Rouda says Apache Spark is an up and coming data analytics engine.
Cisco to use fog computing to approach analytics
Enterprise Management Associates analyst Shamus McGillicuddy says that Cisco is employing a novel approach -- fog computing -- to collect and analyze big data. Fog computing is a distributed computing architecture in which data processing and analysis can be performed at the network edge rather than in the data center. It allows customers to more quickly interpret which data is important to keep and thus send to an analytics platform or a central big data analytics repository. The Cisco ISR 800 series routers, he writes, can detect anomalous data on-site and send it to whichever platform is best able to interpret it. This technology will enable an enterprise to assess the majority of its data. Most enterprises only assess 2% of the data they own, and McGillicuddy said organizations find it challenging to pinpoint an ROI for these projects because they don't know the value of the data before it's assessed.
Because this is cutting edge technology, McGillicuddy says only enterprises with a mature approach to data analytics should consider it.
Read more about why McGillicuddy says some data will remain distributed forever.
Co-opetition: Good for enterprise buyers, bad for service providers
Current Analysis analyst Brian Washburn says that if you're a large enterprise buyer looking at either an IT service provider, cloud, data center or network and technology vendor, it's a good time to buy. Why? Because service providers are commoditizing products in adjacent sectors. So it's a good time to look for a deal on a product, but service providers have their own concerns. As Washburn writes, "The mixing of cloud, data center, network and managed/professional services plays up fears that every adjacent sector is now a competitor. Those concerns are well-founded."
Read more about what Washburn says about co-opetition between service providers and how they can retain a competitive advantage.
Creating an L3 diagram on an unknown network
Packet Pushers blogger John W. Kerns gives step-by-step instructions on how to create what he terms a logical network diagram of an unknown network. He uses the term "logical" instead of Layer 3 to help people unfamiliar with the OSI model. The best place to start a logical diagram is through a core device that has plenty of subnets and routers to remote networks. Kerns uses generic topology icons to represent network components and advises tapping libraries from either Cisco or the Network Topology Icons homepage. When illustrating the subnets, Kerns says to include three things: virtual LAN (VLAN) name, VLAN ID and assigned IP block in the classless inter-domain routing format. He goes on to explain connectors, grouping boxes and device labeling. Kerns provides a document to download for further assistance.
Read more of how Kerns goes through the steps of creating a logical networking diagram on an unknown network.