Cloud storage: Measuring up to the hype
By Tom Henderson and Brendan Allen | Network World US
Published: 16:40 GMT, 23 August 2010
In our continuing series of ground breaking tests of cloud computing services, we take a look at what enterprises can expect if they decide to entrust data to a cloud storage provider.
- Amazon S3 |
- Egnyte On Demand File Server |
- Nasuni Filer |
- Nirvanix Storage Delivery Network |
- Rackspace CloudFiles
We found that cloud storage lives up to its advance billing in two key areas: cloud storage can be fast and the pay-as-you-go model can be a real cost saver. We also found that security could be an issue for enterprise shops, and the formulas for trying to predict overall costs can be complex. The services that we tested were Amazon S3, Rackspace's CloudFiles, Egnyte's On Demand File Server, Nasuni Cloud Storage, and Nirvanix's Storage Delivery Network.
Amazon, Rackspace and Nirvanix represent the containerized/object-oriented model. Egnyte embodies the file/folder metaphor, while Nasuni offers a different twist – it's a front-end that simplifies cloud storage for enterprise customers and connects to other cloud storage vendors on the back end.
To test cloud-based storage, we accessed the cloud vendor's site through their supplied APIs, where applicable. We moved data either from virtual machines in our cabinet at n|Frame at 100Mbps, or from our lab connected via standard Comcast broadband.
Related Articles on Techworld
We pounded each site with a variety of file sizes ranging from 500KB to 1GB. We also tested in two periods, daytime and nighttime, to see if Internet congestion played a role in cloud storage performance.
We downloaded each vendor's software into our lab, and network operations centre. We used scripts running on a virtual Ubuntu 9.10 Server running on ESX 4.0 at our NOC at n|Frame to test upload and download speeds on Nirvanix, Amazon S3, Rackspace Cloudfiles and Egnyte. We performed two sets of tests with multiple files uploaded and downloaded, first during the day, then at night. The second set was run from our XServe running Mac OS X Server 10.5.8 using the same criteria (day and night uploads/downloads).
Each upload and download test consisted of the following randomly generated files: 20 x 500KB, 20 x 2MB, 10 x 50MB, 3 x 1GB multiple tests were conducted. Approximately 14,487 MB of files were uploaded and 14,487 MB of files were downloaded from each vendor's site. (There were some practice test files uploaded and downloaded from each vendor when learning the APIs).
All the upload/download scripts used were written in the Ruby programming language and bash shell scripts used to launch them. Egnyte does not have an API for its file operations, however, it does allow Webdav requests. We used curl, a Linux http request command line tool, to send Webdav requests to access Egnyte files in our Ubuntu VM.
Because Nasuni did not have an API to use, we couldn't include upload and download data for them. Also, Nasuni's product uses multiple clouds so it works a little differently since there is a local machine running with a cache. Nirvanix has a similar product, CloudNAS, but they also have an API to test, and that's what we used. Both Nirvanix and Nasuni's NAS-type products were tested via VMware ESX 4.0 as virtual machines. For other features such as creating sub-accounts, changing URLs, and data gathering, we used each vendors web interfaces where available.
Our connection speed from n|Frame is 100Mbps onto their multipeered backbone. We route to a Gigabit Ethernet switch. Servers included an Apple Xserve running Mac OS Server 10.5.8, and an HP 585 G5 server (four sockets/16 cores) running VMware ESX 4.0 and a virtual machine running Ubuntu 9.10 server edition.