1/18

Social Media Q&A with Jean-Paul Bergeaux

A few weeks ago, we sent out a call on the SwishData Facebook page and Twitter feed for questions from our fans and followers. I love answering questions people have about data performance and sharing SwishData’s perspective on how to handle the challenges government agencies face in a constantly evolving IT environment.

My colleagues and I keep up-to-date on the latest technology design and implementation strategies and tactics, and our Data Performance Architects all have specialized expertise in government and with the technologies we use.

I really appreciate your questions, comments and feedback, so keep them coming! Here are a few from the latest round on our social media channels:

How can an agency know its data is secure in the event of a disaster?

First, let’s define a disaster. There seems to be two kinds of situations people are referring to when they use the word disaster. The first is an IT failure that causes data loss. This could be an array or server outage that corrupts the data. The second is a physical disaster by weather or human that destroys the capacity of a data center to operate. This could be to the application servers and data itself, but could also be a long-term power or network outage as well.

For the first type of disaster, the best way to ensure you will recover your data is to have an off-host copy. By off-host, I am referring to being on separate storage accessed by separate servers from the production servers. Snapshots are not going to fit this description. This data could be at the same site or another site. However, in this case, the same site would be preferable. Most organizations have the majority of their IT staff local to the production data center and restoring from a secondary site can be more difficult, depending on the design, and less-performing, depending on the user connectivity to it.

For the second description of disaster, it is important to have a secondary site with the data required to recover the most important operational functions of production applications.In a perfect world, all production applications would be ready at a moment’s notice upon a failure, leaving very little downtime. Capital costs and administrative costs can prohibit this if the environment is not designed correctly, so often people only have a fraction of the environment ready at a second site, with the rest of the data and applications available once some administrator work is done.

So you can see that in order to be protected efficiently from both types of disasters, a local and remote off-host copy of the data is the best way to be secure in the event of a disaster.

If you want to learn more about my thoughts on backup and recovery, check out this previous blog post.

Can moving to the cloud help my agency save money?

That’s the goal of cloud for most organizations. Whether or not the cloud will save money in the long run will be determined as cloud offerings mature. The low initial outlays and the removal of capital investment certainly are attractive when an organization is facing a large tech refresh.

At SwishData, we recognize that in the tight fiscal climate that agencies face, external cloud subscriptions fit a need for lower cost today. The hope is that the cloud providers can realize enough savings from multi-tenancy that the monthly subscription costs end up being less than the amortized costs of a capital investment and administrative staff required traditionally.

SwishData has ITaaS offerings to fit the external model, and we believe the cloud will end up showing savings for organizations. One thing we believe in is that cloud computing can be delivered both over the Internet and intranet. In order to determine whether or not internal or external cloud services are right for your program, division, office or department, you really should sit down and do an analysis of what resources it will take to host those services internally versus going to external vendors. If several programs can contribute funds to an internal IT department to effectively cover the costs of some services, therefore reducing the spend each program would need by going to an outside vendor for the same service, then using an internal delivery system for those applications makes a lot of sense. A great resource to look at is NetApp’s white paper on Five Steps to Choosing an Enterprise-Class Cloud Service Provider and a checklist to help you choose a service provider.

SwishData is dedicated to helping your organization make the right choice, internal or external cloud, by analyzing the true cost of both options. We believe if the choice is made to go internal, it is helpful to set up a cloud interface similar to one that an external cloud provider would use, deferring some of the tedious administrative duties that are currently forced upon IT administrators.

If you’re still a little cloudy confused about the difference between cloud computing and cloud services (or just trying to get a handle on when “cloud” terminology applies at all), check out the primer I posted previously on the Data Performance Blog.

Keep the questions and comments coming here on the blog and on our social media channels! (Click the social media icons at the top of the page to go to our social media sites instantly.)

Image courtesy Flickr user kdonovan_gaddy.