Nasuni Cloud Filer Ongoing Product Review
Its been a few weeks since we first installed the Nasuni Cloud Filer and as simple as the installation was, the actual use of the product has been even simpler. A good storage system should be like a good referee in a World Cup match (something lacking this year), it should just do its job without any attention from you and that is what the Nasuni Filer has done all week. As you can see from the graph below we have been slowly but steadily adding data to the system and it has worked without flaw.
Thursday, July 22, 2010
For a video tour of our ongoing use with the Nasuni Filer please see below:
Lab Report
George Crump, Senior Analyst
- Status Check
Currently we have three very active continuous data protection jobs that are sending data to the Nasuni Filer throughout the day. Then the Nasuni Filer replicates that new data every hour to the storage cloud. It has handled thousands of small files and large files with little issue. As the graph below shows each hour, unless there is another snapshot already in progress, a new snapshot will take place. The snapshots are what Nasuni uses to move data to the cloud from your local cache.
The GUI also provides an excellent overview of what is and is not in the cloud. As you can see in the picture below, the yellow section represents data that Nasuni calls ‘clean’. It is data that is stored both in the cache and has been successfully replicated to the cloud. The data that is in red is data that Nasuni calls 'dirty' and still needs to be replicated. The status bar on the home screen shows you what % of the current snapshot is still being replicated to the cloud. The green section is free space in the cache.
Nasuni is a client of Storage Switzerland
Another Important aspect when working with cloud based storage is how efficiently that storage is used. After all you are paying by the GB per month, the less of it you use the better. In the Nasuni case you have the local file cache and a relatively big one at 64GB, (you can even grow this by simply expanding the underlying VMware volume) for your active data therefore taking extra time to get maximum storage efficiency from your inactive data is well worth the effort since the time needed to decompress and restore something out of the cache is not something that the user will experience on a regular basis.
As you can see from the screenshot although we have sent almost 17GB of data to the cloud, we are only consuming about 6GB of that space thanks to Nasuni capacity optimization techniques. Of course over time we expect to have much more capacity in the cloud than in the cache. The cache will eventually just represent our active data set. Another important wrinkle is the work of capacity optimization is done after Nasuni receives the data so it does not impact local performance. The result is a significant savings in capacity with no real performance impact.
One of the things I really like about the Nasuni Filer is how well this home screen is designed. In a glance I know everything that is going on with the Filer, without it being too cluttered. In GUI design it is difficult to balance having enough information without having too much information. Nasuni, in my opinion, nails that balance.
