Posts Tagged ‘cloud economics’

Newer Entries »

Coping With Shrinking IT Budgets: Automated High Availability for Offsite Backup

Wednesday, October 27th, 2010

Some of the biggest data management challenges facing IT stem from having to deal with smaller budgets for staff, storage capacity, and tools - at least according to a recent Storage Networking World survey.  With less IT dollars to spread around, I’ll pile on by including in the fray the inability to economically and reliably protect vital backup and restore operations from unplanned downtime situations. In the SME world this can easily become complex and costly at the same time. The ability to affordably bring enterpirse class high availability to your data protection operations can begin to eliminate user downtime.  And if you can automate the whole process the business becomes more efficient and disaster resilient with a stronger bottom line to show for it. 

The recent announcement that CloudArray has been qualified to work out-of-the-box with VMware HA solutions allows you to better protect your busines information 24×7 without major investments in costly resources. With support for multiple VMware availability options, TwinStrata’s CloudArray ensures your off-site data protection operations are uninterrupted and cost-effective with continuous data access in the cloud.

Additionally, when CloudArray is configured with VMware HA or Fault Tolerance you can also think about eliminating complex manual recovery steps (staff), reducing storage acquisitions (capacity), and flat out improve productivity (tools).

Sounds like you may not need to worry too much about shrinking IT budgets afterall.

Tags: , , , , , , , , , , , ,
Posted in Uncategorized | 1 Comment »

Implications of Thin Provisioning in a Cloud Storage Environment

Thursday, September 30th, 2010

 

by Greg Roody, 30 September 2010

 
What if a developer came to you and wanted a 350TB volume for a day?  They wouldn’t actually need to write a lot of data to it, but they need to prove their app would be compatible.  How would you accomplish that?

With Thin Provisioning of course.

Thin provisioning is a form of storage virtualization that is implemented as a basic function within CloudArray.

What it does is allocate physical storage on-demand, only consuming real  resources at your cloud storage provider (CSP) when you actually write something.  Merely creating and provisioning a volume in CloudArray doesn’t consume physical resources in the Cloud until you write some data to it.

As a result, you can easily create that 350TB volume, which will appear to your servers as physical storage, but which consumes no real storage at your CSP (and thus doesn’t generate a large bill). As you layer a file system or volume manager on it, a small amount of storage would be consumed for metadata directories and inodes.  As you write new data to the volume, it will grow accordingly and you will pay for what you actually use.

For example, here is a screenshot after creating a 350TB volume in CloudArray.

Notice that it is utilizing  zero MB.   Until we actually mount it and write some metadata to it, it is nothing more than an empty shell, yet any server mounting it would see what looks like a 350TB physical volume.  In CloudArray, each volume can be as large as 384 TB, and you can configure as many volumes as you need.  Of course, you will have to manage your cache accordingly, so there are limits, but not in what you can create.

This has a very positive effect on performance (as well as economics).  You don’t have to read back unallocated blocks when an application calls for a range of blocks for example.  Less data to read and write is very good for Cloud Storage performance.

This sounds great on the surface (and actually is very powerful), but there are some cautions you need to take to make the most of the functionality.

More after the fold… (more…)

Tags: , , , , ,
Posted in Uncategorized | No Comments »

Storage Cloud-o-Nomics

Thursday, November 5th, 2009

You may have read a recent blog entry by a leading industry analyst addressing cloud storage ROI: 20TB of public cloud storage as a service at Amazon’s S3 rate of $0.15/GB per month totals $3,000 per month, or a whopping $36,000 per year. Some do-it-yourselfers responded they can build 4 storage arrays for this one-year price. But can acquiring a glut of storage capacity give your business offsite data protection? Can it instill confidence that your business will be able to survive a complete loss of your primary site data?

We recently used Clarity AP to conduct a total cost of ownership (TCO) analysis of cloud storage as an offsite replicated tier of storage accessible via our cloud gateway, CloudArray. For the analysis, we assume a company with a single site for data storage and linear capacity growth over 3 years. We calculated the 3 year TCO of an in-house implemented offsite tier of replicated storage versus replicated storage to the cloud. We plotted the capacity crossover point up to which cloud storage holds a cost advantage.

While we invite everyone to look at the full report, here’s a brief summary:

Cloud storage may make sense for capacities up to 60TB for replicated data. At these capacities, cloud storage benefits from a pay as you go model that does not suffer the underutilization experienced in storage arrays. As a remote site data protection solution, cloud storage substantially reduces the need for offsite infrastructure and management.

The chart below illustrates the crossover point:

3-yr TCO for offsite replicated storage

3-yr TCO for offsite replicated storage

For the initial TB of capacity, costs for an in-house solution over a 3 year period are $233,492, compared to $79,794 for cloud storage. For 20 TB, there is a $100,000 savings when data is replicated to a cloud provider.

The report lists results, configurations and cost assumptions. The analysis purposely neglects to factor local site floor/power/cooling since some may argue these are sunk costs. It also does not address application/compute failover which is a separate analysis.

Conclusion

Based on the results of this analysis, cloud storage can be very compelling for companies replicating up to 60 TB to a public cloud. For a mid-sized company or departments within a larger organization, this represents a substantial capacity range across which cloud storage presents a strong ROI. Watch for this range to expand even higher over the course of the next 3 years with more favorable cloud pricing and tiers of service.

Can your business benefit from cloud storage? Let us know.

Tags: , , , , , , , ,
Posted in Uncategorized | 1 Comment »

Newer Entries »