Hi – back again to discuss the requirements for hosting primary, unstructured enterprise data in the Cloud. For your convenience, I have reprinted the list of requirements from my initial post again – note that I will continue to do this and will link from this list to previous posts as well:
- Accessed like traditional storage
- Easy to use, enterprise features
- Comprehensive data integrity/protection
- Data security/privacy
- Continuous availability
- Non-blocking performance
- Administrative transparency and control
- Good investment value
In this post I’m going to discuss the position that in order for a cloud storage service to be a viable option to host primary enterprise data sets, it must provide easy-to-use, enterprise features.
On the face of things, this seems obvious, almost tautological: of course a solution for storing enterprise data should provide enterprise features; however, existing HTTP-centric cloud storage solutions lack key features that should be required for enterprise adoption.
So what are some examples of these “enterprise features?” Here is a list of things our enterprise users told us to include in the Zetta Cloud Storage service — things that they expect to find in a cloud storage solution hosting their primary data:
- Parity to Traditional Arrays — in essence, the cloud storage solution should come with the features you’ve come to expect from any robust NAS array you have running as a bump on your network today – things like snapshots, mount-and-write, integration with external systems (e.g. LDAP), support for existing ACLs, etc. Without these, something will need to be written to replace these features, and enterprises don’t have the resources on hand to simply replicate features they typically get from their storage solutions.
- File-based geo-replication — replication is something that IT administrators have come to expect to be facilitated by their storage technologies. And I’m not talking about the type of replication common among the HTTP cloud object stores – those services typically rely on replication as their sole form of data protection, and employ a solution that is opaque to the user. What our enterprise customers asked us for was a form of replication that results in a mountable, readable volume in another identified data center, with all of the visibility and transparency they would get if they constructed their own solution.
- Capacity management & visibility — An enterprise solution should provide real-time presentation of exactly what is happening on your volumes – usage trends, system performance, and real-time access to events (including bad ones!). The fact that the volumes are resident at a service provider shouldn’t change the fact that you want transparent visibility into what is happening with YOUR data!
- Instant provisioning — In this particular case, you should actually expect your cloud service provider to provide much better performance than you would find with a traditional array – with a traditional array, you need time to take down space and power, negotiate with the array vender or VAR on upfront capital cost, and install, configure and test the array. This can take weeks or months. With a cloud service provider like Zetta, you can be up and running within minutes or hours.
- Native support for file-based apps — this is kind of a short restatement of my last post – an enterprise service should provide a full-featured file system that walks and talks like any other filer on your network, making it plug-and-play with existing enterprise architectures and file-based applications.
There is more that I could touch on here, but this should give you an idea of some of the things Zetta provides to our enterprise customers.
Once again, contrast this set of features with your typical HTTP-centric object stores – by design, those solutions do not provide any of the enterprise features I’ve discussed above, since by design those were built to meet a simpler set of requirements, targeted at a non-enterprise customer.
I’ll be back soon to discuss some of the core strengths of the Zetta solution, beginning with a discussion of the Zetta data integrity solution.
Tweet This
RSS