Interview OpenStack-using startup SwiftStack has got itself $16m in B-round funding, raising the total cash invested amount to $23.6m. Why should we care?
CEO Joe Arnold kindly told us. As a recap the business was started up in 2011 and has had a funding injection once a year since then, commencing with a seed round of $1.5m in 2012, an A-round in 2013 and including today's B-round.
The open source OpenStack cloud computing software system has a variety of storage software possibilities as well as its included Swift object store and Cinder block store.
El Reg: Why is Swiftstack better or more appropriate for use cases than Ceph, Gluster and Lustre?
Joe Arnold: Ceph, Gluster and Lustre are designed and built for fundamentally different use cases than SwiftStack.
The message we hear from storage architects and operators is that they prefer the best technologies to meet specific requirements and use-cases. All-in-one solutions, like a printer/copier/fax/scanner, are by design not optimized to deliver the best results for each use-case.
Swift, which was designed from the ground up to do one thing really well - object storage - has this focus. Swift is not block storage, so it is not for running VMs or databases. It is also not a file system and does not try to emulate those interfaces. Rather, Swift was designed specifically for unstructured data, the type of data that is growing the fastest with today’s applications. A big reason it serves so well for unstructured data is its specialization and design for eventual consistency to support massive scalability and geographic distribution of data.
El Reg: Why should customers take a risk on immature Swiftstack instead of the more mature offerings from the suppliers above?
Joe Arnold: We would argue that SwiftStack is a much lower risk solution for customers than choosing a 100 per cent proprietary solution. Here’s why:
- SwiftStack includes a 100 per cent open-source release of OpenStack Swift at its core... Once dismissed by their proprietary competitors as “immature,” open source operating systems, middleware, application frameworks, and databases are now standards in enterprise and Web infrastructure. ...
This same transformation is now happening to the storage tier ... led by storage architects, operators, and deployers at enterprises, Web companies and service providers alike.
By using SwiftStack, customers can take advantage of the collective work developed by +175 engineers from some of largest names in technology, including Red Hat, IBM, Rackspace, HP and many others versus getting locked in to a proprietary, closed source storage technology.
- According to Gartner, Swift is the most stable and widely adopted part of OpenStack (February, 2014). Swift already powers some of the largest object storage clouds in place today, including the HP Cloud, IBM SoftLayer, Rackspace Cloud Files and Oracle.
- SwiftStack is built for today’s applications with a modern, cloud-native architecture. As such, it can be deployed on a variety of different hardware platforms on off-the-shelf Linux distributions versus inflexible and limited hardware choices. It scales linearly and can serve as the storage tier for both deep archiving use-cases as well as web-facing applications with 100’s of thousands of concurrent users, serving content directly out to mobile devices and browsers.
- SwiftStack is used in large scale production by customers across all industry verticals, including ... some of the largest banks, media and entertainment companies, life sciences firms and the federal government, [like] Time Warner Cable, eBay and HP.
El Reg: How does SwiftStack's product compare to those from Amplidata, Caringo, Cleversafe and Scality?
Joe Arnold: Here’s how we compare:
- Decoupled Management and Control: SwiftStack offers customers an innovative model to deploy, manage and scale their object storage clusters that is fundamentally different from the proprietary object storage companies. For instance, our out-of-band controller - the SwiftStack Controller - decouples management and control from the actual storage nodes, which enables customers to deploy and manage one or more object storage cluster in any data centre...
- ... Because SwiftStack includes OpenStack Swift at its core, customers take advantage of a wide range of applications and programming libraries that already support its API.
- We enable companies to build storage infrastructure as a service. This allows developers and operators to use and consume storage internally and the business to build their next generation applications. Our storage was built for applications that can be accessed from anywhere and any device using object APIs, while ensuring file-based applications that haven’t been updated to use object APIs can be used using filesystem protocols.
- Some of our key differentiators:
- SwiftStack Filesystem Gateway(FSGW): This allows customers to mix and match file and object based workflows into a single system with unified access to the data. What’s different is that we allow an object to come in and a file to come out and vice versa, which is not the case for other object storage solutions that offer filesystem support (Scality), where data stored via the object API can not be accessed via the filesystem access and vice versa. This is also a completely unique feature compared to the gateway providers because the other FSGW vendors want their users to go in and out of their gateway to get an object locking them in, which creates isolated islands of storage.
- As for performance, we can put multiple FSGW into the system to support the bandwidth the customer needs.
- Storage policies from SwiftStack allow deployers to specifically configure their Swift cluster to support the different needs of data stored in the cluster. You have a single system, not multiple systems, reducing TCO because you have less things to manage. If [customers] have multiple data centres, they can have one cluster across geographies and create storage policies within that cluster.
- Enterprise Authentication is a big deal with us. [It] hooks up simply with LDAP/AD and provides utilisation information for chargebacks. We get major kudos from customers here.
- Baked-in DR strategy with a backup and archive utility. DR with multiple sites can be set up and when a backup goes into that system, it goes into those data centres simultaneously. Geographic distribution is built in.
- SwiftStack can run on different kinds of hardware, differently sized drives, different generations of CPU - in the same cluster. In other words, customers no longer need to over-purchase a packaged array or have a precisely defined hardware configuration or sized hard drives that cannot change over time as the cluster grows, as required by proprietary vendors such as Cleversafe. What we hear from customers is that they want to take advantage of larger drives (6TB, 8TB, 10TB etc) as those are released - which SwiftStack is natively built to accommodate.
- Global Data Distribution: This functionality enables the storage cluster to span multiple regions. A controller can manage multiple sites as one cluster with a global footprint in a single pane of glass. Example, for seeding content distribution, a cluster can be used to distribute content across multiple, geographically dispersed regions. Requests for data can be served by the local region.
(We edited Joe's replies for clarity.)
Overall, El Reg thinks this is all good and helpful stuff. There is a great head of steam building up behind the OPenStack initiative, but will market adoption be weakened by vendors pushing their own proprietary cloud software stacks?
As for Swift, it has to compete against many proprietary object stores, some with high adoption and growth rates such as Cleversafe and Scality.
We can expect to hear more on Swift as OpenStack's profile is raised. There is an OpenStack summit in Paris next week and news is bound to flow from that. ®