Storage administrators are combining tiered storage policies with data reduction technologies to manage capacity...
growth as IT budgets shrink or stay flat in 2009.
Clackamas County in north central Oregon deployed tiered storage using F5 Networks Inc.'s ARX file virtualization switch and Data Domain's DD565 data deduplication disk arrays in an effort to put off adding capacity to its tier 1 SAS-based iSCSI storage-area network (SAN) storage. In addition, Vancouver, B.C.-based Rainmaker Entertainment Inc. will use an ECO appliance from Ocarina Networks with BlueArc Corp.'s Titan and network-attached storage (NAS) systems from Isilon Systems Inc. to free up primary storage and keep more archival data online.
Clackamas County's tiered storage policy frees up capacity
Clackamas County is part of the Portland metropolitan area and home to approximately 400,000 people. When it began running out of space on its 40 TB Dell EqualLogic iSCSI SAN, the county's IT staff looked for an alternative to expanding tier 1 SAS capacity. "We're suffering from the same effects of the economy that everybody else is," said Christopher Fricke, senior IT administrator.
The county has two Data Domain DD565 data deduplication arrays that will do double-duty as backup capacity and archival storage for tier 3 unstructured data. The DD565s were originally bought for use with the county's IBM Tivoli Storage Manager (TSM) backups, but a partition on each of the boxes will also be accessed as NAS by the F5 Networks ARX switch. The county will migrate files older than 60 days to tier 2 SATA storage on the EqualLogic arrays, while files older than 120 days will be sent to the Data Domain DD565s. The archiving will be done without any changes to the way end users access files.
"Our tiered storage policy will free up about 10 TB of tier 2 storage. That, in turn, frees up SAS capacity used by database servers and our email archiving system," Fricke said. "It helps us not have to chase capacity while we go through a budget crunch -- we can focus on performance rather than capacity. In a couple of budget cycles, we should be able to look at replacing our current tier 1 with solid-state disk [SSD]."
So far, Fricke said he's learned one lesson when deploying the ARX switch: be mindful of the network infrastructure. In his case, misconfigured ports caused problems with the interfaces on the F5 Networks ARX switch. Network capacity should also be carefully assessed before centralizing on a network-based tool like ARX, Fricke said. "The device can handle it, but make sure the network you're plugging into can handle the consolidation," he said.
Rainmaker Entertainment dedupes data to keep more archival footage online
Rainmaker Entertainment, a digital animation studio, has 100 TB of storage capacity on its BlueArc Titan 3200 primary storage system and another 100 TB on an Isilon IQ 3000 clustered NAS systems for secondary storage.
The company plans to deploy Ocarina Networks' ECOsystem appliance between the BlueArc and Isilon devices to compress and deduplicate data as it's migrated from primary to secondary storage. This will allow the studio to keep data intact after projects finish and avoid restores from tape. It will also stretch the life of the secondary Isilon storage and free up space on the primary BlueArc NAS head for current projects.
"We're looking at compressing 6 terabytes down to two, and possibly storing 300 terabytes on the Isilon system in the future," said Ron Stinson, Rainmaker's director of IT and operations.
The primary goal of the compression is to deal with archive data, Stinson added. Data reduction comes with performance tradeoffs. For primary storage, "we wanted to maintain the performance we've had – devices we've tested to date have had significant lag times," he noted.
One of the trickier decisions during evaluation was whether or not to incorporate stub files for archiving. "You can do the same thing a couple of different ways," said David Algar, principal data administrator at Rainmaker Entertainment. "I think we'll end up migrating the entire data set over as a separate repository. It can be more complex with stub files," he said.