This is the second part of an extended interview which I did with Brendan Sullivan of SullivanStrickler and Fred Moore of Horison Information Strategies, Inc. in which we discussed the revival of tape as an archive medium. The first part is here and the third part will be published shortly.
The subject of this part is the reasons why tape declined as a popular archive until about 2000 and then revived. The technical deficiencies of tape proved inadequate for the changing demands – not just increased volumes but changed purposes and pressures. Since 2000, the technology of tape has advanced to meet the demands. This, coupled with two decades of experience, makes tape a valuable, yet still under-appreciated, means of archiving data.
Part of the problem derived from issues with the tape medium itself – there were multiple differing formats, tapes were easily damaged, and the material of which they were made was susceptible to decline over time.
Now, both the tape itself and the backup software are standardised; encryption is easier to deal with; there is no need to rebuild the native environment in which the backup was made; and people like SullivanStrickler are used to working with it and getting data off it.
The poor reputation came during a period when the need to keep ever-increasing amounts of data for long-term purposes coincided with the technical difficulties then raised by tape. Enron and the Zubulake case changed the requirement from relatively short-term rotational backups to a perceived need to keep everything.
Tape was rarely the first port of call when historic data was required – organisations would deal first with more readily-available sources and only then resort to tape archives; at that point they would often find boxes of tapes of different types, created using now-redundant software by people who had long since left the organisation.
The focus then was more on the fastest and easiest way to store data and less on the records management and classification needed to retrieve it. Tape has now caught up with enterprise requirements and, as Fred Moore says in this interview, an evangelism effort is needed to demonstrate how good tape is now.
Other legal pressures have forced a focus on data retrieval, not least changes in privacy culture. Between them, the EU’s General Data Protection Regulation (GDPR) and the pending California Consumer Privacy Act give individuals in certain circumstances the rights to erasure of their data and the right to ask for information stored about them. This is not sit well with the historic ideas about restoration from tape.
Organisations are now moving from “How do I deal with this request?” to “How do we organise our data to deal with any potential future requests?”
At the same time, the recent amendments to the Federal Rules of Civil Procedure have led to the courts being less aggressive about inadvertent spoliation. As a result, organisations are less fearful about embarking on planned reviews of their data.
The SullivanStrickler approach is to add intelligence about unstructured data while migrating it, making decisions while moving it about what to keep and what not to keep. This approach shrinks the number of tapes and adds intelligence to the data.
As well as reducing storage and retrieval costs, this helps prepare for moving data to the cloud.
Part 3 of this interview will be published shortly.