Ubiquity is everywhere for DevOps using Docker and Kubernetes

August 14, 2017 Leave a comment

This is a repost from an article I wrote on LinkedIN.

Project Ubiquity is gaining speed to help customers manage persistent storage for rolling out containers. One of the biggest issues in the container environment is now how to manage the persistent storage that allows you to snapshot, provision and move data around as applications move in and out of your DevOps.

IBM in an open source release now enables persistent storage for Docker and Kubernetes — which initially covered IBM Spectrum Scale only — now enables block storage as well. Now, clients can test the pre-release project using their IBM Spectrum Virtualize and IBM Spectrum Accelerate based systems as the storage back end for stateful containers.

This flexible framework allows businesses to now deploy not just IBM Storage like the new generation A9000 but also our storage virtualizing solution Spectrum Virtualize or the high performance file system Spectrum Scale. All of these solutions are able to be deployed as appliances or as software on your storage rich servers. With Spectrum Virtualize you can reuse some of your older storage that is under utilized today.

Here is more information and the link to the GitHub.


Categories: Uncategorized

Data Hoarding: How much is it really costing you?

March 13, 2017 Leave a comment

n-HOARDING-SIGNS-large640I have closet in my house that I keep all kinds of computer gear. Most are things from some fun project that I was working or a technology that is past is prime. There is everything from Zip drives to coax termination to a Ultra-wide scsi interface for an external CDROM. Why do I keep these things in a box in a closet? Great question that usually comes up one a year from some family member that sticks there head in there looking for a toy, coat or looking to make a point.
But on more than one occasion I have had to go to the closet of ‘junk’ to get something that helped me in completing a project. A Cat5 cable for my son’s computer, an extra wireless mouse when my other one died. Yes I could go through it all and sort it out and come up with some nice labels for it all, but that takes time. It’s just easier to close the container lid and forget about it until I realize I need something and its easy enough to grab it.
Now this is not a hoarding issue like those you see on TV where people fill their house, garage, sheds and barns with all kinds of things. Those people who show up on TV have taking the ‘collecting’ business to another level and some call them ‘hoarders’. But if you watch shows like “American Pickers” on the History Chanel, you will notice that most of the ‘hoarders’ know what they have and where, a meta data knowledge of their antiques.
When you look at how businesses are storing their data today, most are looking to keep as much as possible in production. Data that is no longer serving a real purpose but storage admins are too gun shy to hit the delete button on it for fear of some VMWare admin calling up to see why their Windows NT 4 server is not responding. If you have tools that can move data around based on the age or last accessed then you have made a great leap into making savings. But these older ILM systems can not handle the growth of unstructured data of 2017.
Companies want to be able to create a container for the data and not have to worry if the data is on prem, off prem, on disk or tape. Set it and forget it is the basic rule of thumb. But this becomes difficult due to the nature of data as it has many different values depending on who you ask. A 2 year old invoice is not as valuable to someone in Engineering as it is to the AR person who is using it to base their next billing cycle.
One of the better ways to cut through the issue is to have a flexible platform that can move data from expensive flash down to tape and cloud with out changing the way people access the data. If the user can not tell the difference where his data is coming from and does not have to change the way he gets to it then why not look at putting the cold data on something low cost like tape and cloud tape.scale_cloudTier
This type of system can be accomplished but using the IBM Spectrum Scale platform. The file system has a global name space across all of the different types of media and can even use the cloud as a place to store data without changing the way the end user will access the data. The file movement is policy based and allows admins to not ask the user if the data is needed, it simply can move it to a lower cost as it gets older/colder. The best part is because of a new licensing scheme, customers only pay the TB license for data that is on disk and flash. Any data that sits on Tape does not contribute to the overall license cost.
For example: 500TB of data, 100 TBs that is less than 30 days old and 400 that will greater than 30 days. If stored on a Spectrum Scale file system, you only have to pay for the 100 TBs that is being stored on disk and not the 400 TB on tape. This greatly reduces the cost to store data as while not taking features away from our customers.
For more great information on the IBM Spectrum Scale go here to this link and catch up.

What does the perfect match sound like?

February 15, 2017 Leave a comment

ipad_mymatchesThe day after Valentine’s is also known as “Singles Awareness Day” and people flock to the stores for the left over candies now priced down to move. Some people who are looking today will go on a website to find their love ‘match’. But how does that really work in the background?

One website named “Plenty of Fish” is an online dating site that claims they are the largest online dating site, with over 90 million registered users. Every day 55,000 people from every corner of the Earth sign up on their website for help finding that special person.

Here are some staggering facts from their website:

  • PlentyOfFish has over 55,000 signups every day
  • 80% of usage on PlentyOfFish takes place via a mobile phone
  • Over 3.6 million people log on to PlentyOfFish every day
  • Every 2 minutes a couple confirms they met on PlentyOfFish
  • There are over 10 million conversations every day on PlentyOfFish
  • PlentyOfFish creates 1 million relationships every year

Now this is not a promo for the site but the data points are interesting. If you were looking for someone to spend the rest of your life with or just trying to make friends, some of the factors in you paying this site would be the speed and accuracy of their match making algorithm. When I talk to clients about technology and their pain points we talk alot about making the decisions faster with more accurate data. This is very similar to the data points above.

I saw this video on Youtube and thought it was pretty cool.

Categories: Flash, Uncategorized

IBM Spectrum Accelerate Upgrade V11.5.4

February 14, 2017 Leave a comment

IBM Storage announced today an update on their Spectrum Accelerate platform. This is part of the Spectrum Family which IBM has been selling under different names for over 20 years. The ‘Accelerate’ portion is based on the XIV grid technology and runs on VMware servers in a converged or even hyper-converged environment.

IBM Spectrum Accelerate V11.5.4 delivers the following functions with this release:

  • Data-at-rest encryption enables the use of SEDs to protect user data from exposure to unauthorized personnel
  • Hot encryption enables you to activate encryption at a later stage of the system lifecycle instead of at the time of installation
  • Support for VMWare vCenter/ESXi 6.0 enables VMWare-based systems to upgrade to a newer version of vSphere
  • Support for 450 GB and 15K rpm drives increases the diverse inventory of hard drives you can install
  • Hyper-Scale Manager is the next-generation, web-based GUI to deliver a holistic and intuitive user experience while sharing the same pane of glass with the whole A-line family: A9000, A9000R, and XIV Gen3

For more information about this release click here.

A great resource for all things on Accelerate, check out these Redbooks

Categories: Uncategorized

Software Defined Storage: Client Case

February 6, 2017 Leave a comment
Categories: Uncategorized

IBM Hyper Scale Manager: Where Can I Put My Data?

January 27, 2017 Leave a comment
Categories: Uncategorized

IBM Hyper Scale Manager: Volume Creation

January 26, 2017 Leave a comment
Categories: Uncategorized