StorPool update adds data backup, disaster recovery options, iSCSI support

The shape of the storage market will change beyond recognition over the next five years, according to vendor.
Written by Colin Barker, Contributor

Bulgarian storage company StorPool has made a major upgrade to its software that includes support for native data backup and restore capabilities, full iSCSI support, and support for multi-threading that can concurrently execute multiple processes.

According to the company, the latter means that it can achieve per-node performance that is up to 80 percent faster, offering from 100,000 to up to 180,000 IOPS. The improved hardware performance will enable customers to build storage systems at well below $0.1/IOPS, the company says.


Ivanov: "A small number of companies are spending a lot of money and IT is their competitive advantage."

Photo: Colin Barker

"We believe that five years from now the majority of businesses will not be running a storage array anymore but instead are going to run intelligent software on standard racks," StorPool's CEO Boyan Ivanov told ZDNet.

"If you look at where budgets are going, they are going to public and private clouds. They are going to the guys who are doing mainly infrastructure. Cloud service providers who in 80 to 90 percent of cases use a Linux stack.

"A small number of companies are spending a lot of money and IT is their competitive advantage. That is our target market," he said.

Ivanov said that the decision to focus on iSCSI was "an easy one". StorPool developed its own implementation of iSCSI which still uses the iSCSI initiator required in VMware, Hyper-V, Oracle, or Unix but this is put next to a StarPool-developed iSCSI target on the server side.

StorPool is a distributed storage solution that allows companies to run data storage, alongside applications, on standard x86 servers, effectively aggregating the capacity of the drives.

According to Ivanov, this creates a virtual storage array without the need for a storage array, and one that is flexible. "If you lose a node operations continue without downtime because there are other copies of the data on different nodes," he said.

This means that you can have different operating systems running on different servers but all talking to the same underlying StorPool network.

Another feature to be launched is native data backup and disaster recovery with geo-replication. The functionality here "eliminates the need for a separate data protection solution", the company said, and simplifies the storage infrastructure.

An advantage is that StorPool clusters in remote locations can now asynchronously transfer encrypted snapshots over the public internet, the company said. This should improve data backup and virtual machine migration from one geographical region to another.

So after the first sync, StorPool will only back up new or changed data, rather than entire data sets. These incremental backups should reduce data transfers and shorten backup and recovery times by up to 60 percent, the company said.

A final feature to be added is multi-core processing. Because the software is multi-threaded it will concurrently execute multiple processes, which, according to Ivanov, should improve performance.

StorPool customers are divided evenly around the world, with a third each in the US, Europe, and Asia/Pacific.

Read more about storage

Editorial standards