Showing posts with label Scale-out. Show all posts
Showing posts with label Scale-out. Show all posts

Wednesday, 17 February 2016

Welcome VxRail, VCE's Entry into the HCIA Market


You want the ease of a Hyperconverged appliance but still have all those advanced capabilities that service enterprise data centres, then the new VCE VxRail is for! This is not so much an evolution of the VMware EVO:Rail appliances but more a revolution.

Why do I say that? The Achilles heal of EVO:Rail was the lack of flexibility, all you could do was chose between two RAM configurations (no variant in disk or CPU), 128GB or 196GB which for some reason had the 196GB model referred to as the desktop service model... never quite got my head around that one, why wouldn't other services want a larger memory footprint? The other side was the fact that you could only grow your appliance service out by adding additional appliances, start with a 4 node 2U appliance and then grow in increments of 4 from there till you hit the 64 node ceiling (aligned to the maximum size for a HA / VSAN cluster).

Now think the VCE VxRail appliance here we have flexibility that can meet a huge range of use cases, your choices now include:


  • Starting point is now 3 nodes (today) and growth can then be in single node increments up to the 64 node maximum
  • CPU model options provide a range of 6 cores (single socket) through to 28 cores (dual socket) per node
  • Memory configurations from 64GB through to 512GB
  • Disk configurations of Hybrid (Flash + Spinning HDD) or all flash ranging from 3.6TB through to 19TB per node 
  • Even networking options will support 1G connections on the low end nodes with the standard being dual 10G 
You can also now mix different model configurations and EVO:Rail appliances under the same management. Phew... all that flexibility but with the ease of use that comes with the VxRail appliance.

Some other cool features of note are things such as:
  • All the coolness of VSAN 6.2 with Dedupe, compression, flash optimisation, erasure coding, snapshots, cross site HA (yes you can link two remote appliances into a vSphere Metro Cluster)
  • VM Level replication with EMC RecoverPoint/VM (yep included) 
  • Cloud based storage tiering EMC Cloud Array (yep also included)
  • VCE Vision software for service health and interoperabilty compliance monitoring and reporting
  • vRealize LogInsight for lag aggregation and analytics
  • vSphere Data Protection (EMC Avamar)
  • Appliance Market Place, your VxRail App Store for easy request / enablement of additional services
And that is not the end of it, too much to list :)

If you want to have a look for yourself have a look at this demo of the VxRail Manager I recorded!




More information is available at http://www.vce.com/products/hyper-converged/vxrail

Friday, 6 June 2014

Cool Solution# Hadoop File Services with EMC Isilon

I'm Sold!

First let me say, by my own admittance, I am an infrastructure guy! That said,  I have been lucky enough recently to be given the chance to dive into the world of Big Data and PaaS. As a techie the extensive technology options in this area are very impressive and as a consequence, I have had a lot of fun combined with late nights and heavy learning.

I have always worked on the KISS principle 'Keep It Simple Stupid!' and Hadoop first into that with the exception of its file services HDFS. By having that layer of abstraction to the file system the ability to manage and populate the file system is non existent without specific tools written for the task.

I know DAS and scale-out through the data nodes is great and builds a big pond for your big data by combining the compute and disk resources of a 1000+ server nodes . Yay all good! But putting my EMC and Infrastructure hat on for a moment what about the following:


  1. How do I backup my HDFS based data
  2. How do you use those other cool storage capabilities such as snapshots, auto-tierring, etc
  3. How do I get real-time analysis on data without having to move it into HDFS
  4. How could you share the data from within HDFS
  5. What if you need more compute resources in your cluster but not storage, or the other way around. 
This is where EMCs Isilon comes to the rescue through its OneFS. The flexible open access into its large single name space is invaluable. All the joys you expect from a true scale-out, enterprise class storage array are there as well as the capability of accessing the same file system in multiple ways at the same time. Think about a web site logging to an NFS mount that is part of the HDFS file system allowing for realtime analytics against it! 

Want to see this in action, look at this demo http://www.youtube.com/watch?v=Qx9BMzZa8UI
 
Further information can be found at http://www.emc.com/domains/isilon/index.htm