Hierarchical Storage Management is back. And this time it's cloudy
KompriseCloud can now shunt data between different cloud storage operators
Data management software vendor Komprise has added extra cold cloud tiers to its data lifecycle manager, which moves data to slower access storage tiers without affecting its accessibility.
The company has also introduced analytics to its metadata library, data confinement and ONTAP 9 support.
Komprise's software moves data between tiers without leaving stubs*. It instead uses dynamic links, making search, management and access easier than before.
With this announcement the Komprise Intelligent Data Management (KIDM) product moves to v2.6. It already supports AWS S3 and Google Nearline as public cloud tiers, and adds colder data tiers behind them: S3 Infrequent Access and Glacier from Amazon, and Coldline from Google, calling this cross-cloud tiering.
It's intended as a replacement for existing public cloud archiving products that Komprise believes are complex to set up, don't scale on demand, don't run in the background, use static and error-prone stubs and cost a lot. KIDM v2.6, it says, fixes these issues.
Since access times for files in Glacier are much longer than for files stored on other tiers – meaning on-premises, AWS S3, or Google Coldline or Nearline public cloud stores – Komprise's non-use of stubbing may not deliver much faster access over stub-based HSM products, but the archived data exists in the same namespace as non-archived data making overall access easier. Coldline, Google claims, is just as fast as Nearline in terms of latency.
A data confinement feature can use customisable policies to move data to be kept unaltered by segregating it into a separate namespace and then, if called for, deleted. This means KIDM can meet GPDR concerns, such as the right to be forgotten.
A metadata analytics feature is being beta-tested and this can be used to check metadata for so-called zombie files; files no longer owned by anyone or needed by anyone which are still taking up space. This metadata analytics feature checks metadata, not file data contents so it can be used to place files on tiers more appropriately or group files using some criterion, but not, as we understand it, much more than that, yet.
KIDM v2.6 also adds support for NetApp's Data ONTAP 9 to the existing support for ONTAP 7, 7/8 mode and 8. ®
* A stub is a micro-file remnant that tells the accessing entity the file is not where it was, but has been moved to an identified alternate location. With Komprise the system file metadata replaces a static link address for a file with a dynamic link address. If and when a file is moved this dynamic link address is altered to point to the new location, avoiding the wasted storage drive access that a stub remnant entails.