ParaScale Storage Cloud(PCS) FAQ Registration for Trial 2 When will Trial 2 happen? We are getting very close to the release of trial two software and are currently looking for participants.
Will anyone be able to sign-up for Trial 2? What about locations outside N.America? With our SW download model, we can certainly host trials outside N.America. But we will need to be selective in participants to get quality feedback. By being efficient in the trial process, we expect to get our cloud solution to market faster.
What is expected from trial participants? Honest and frank feedback. The trial will also require hardware that can support VM images for testing.
ParaScale Cloud Storage (PCS) sweet-spot (where it fits, where it does not)
Is ParaScale Nirvana for storing any kind of data? No. PCS is designed for storing file, not block data. ParaScale is a cloud storage solution – it is not a SAN.
Is PCS software good for databases? No. If you are maintaining a database that drives your enterprise revenue stream, cloud storage is not a solution for you, and you should be looking at SAN solutions.
What kinds of data is PCS software best at storing? (Where does PCS shine?) PCS is especially great for storing large files (Web2.0 rich media, video surveillance data, large genomic data files, oil-gas SEG Y data). Both read and write performance is highly scalable. Single data nodes(with modest dual core CPUs and 2G RAM ) can deliver 70-100 MB of read and write performance, and aggregrate I/O essentially scales linearly with number of nodes.
What use-cases are not in ParaScale’s sweet spot? ParaScale is less of a fit as with the workloads handled by classical primary NAS storage. If you are looking for a few 10s of TBs of primary storage for active data you should be considering NetApp, EMC, HP, IBM etc.
ParaScale is not suitable for storage tied to high performance computing use-cases. If you are looking for blazing single-file write throughput to multiple clients as seen in animation applications, CAD/CAM etc., there are many solutions to choose from. For these applications you should be looking at NetApp GX/Spinnaker, HP-Polyserve, Panasas, BlueArc, iBrix, and open source clustering options amongst others.
Manageability How easy is it manage a PCS solution? The PCS solution is largely self managing. The GUI provides valuable information as you monitor the health, utilization and performance of your cloud. It is also necessary to set the simple policies (e.g. number of replicated copies). But most of the onerous and frequent tasks such as data migration, load-balancing are done automatically by the system. The system is designed to not require additional IT heads as you scale the cloud from a few nodes to 100s of nodes.
Platforms Can I create a PCS pool using hardware that I have? Yes. Your server should be able to run a virtualization hypervisor or Linux. In fact, we anticipate that it will be fairly common to throw older “post-warranty” servers into the PCS pool. “This is where servers go to age gracefully”.
Can I use different servers that I already own in a PCS pool? Yes. Nodes can have different hardware capabilities and configurations.
Can I create a PCS solution with just one monster-server? Yes. You can float a lot of VMs on this one very capable server. And it will give you a great feel for how the system works. But to get real resiliency with our replication model, the PCS solution gets interesting when you have at a minimum 1 control node and 3 storage nodes, or more.
I don't want to use VM-hypervisors; can I run PCS software directly on server platforms? Yes. For our GA release we are in the process of qualifying a number of server platforms for native deployment (without VMs).