Data Preservation

This editorial was originally published on Dec 3, 2013. It is being re-run as Steve is on holiday.

Maintaining data across time isn’t something many of us think about. We work with data in the here and now, and in the database world, we typically only need to recover or restore data from a short window. Like most of you, I would usually plan on recovering data that’s only a few days old. Being forced to restore a database from two weeks on any of my systems would make me cringe. It certainly would be embarrassing for me personally if it were my fault I couldn’t restore to a point in time that was more recent than that.

In planning to recover our systems, we typically know the versions of software we have to recover from, and we can easily re-download copies of SQL Server or the patches we need. Most of us are dealing with SQL Server 2000 or later, which is good since those are the only versions still documented on MSDN. If you need SQL Server 7.0 or SQL Server v6.5 documentation, I hope you have copies.  The same goes for the media. You can still download SQL Server v6.5, and SP5, but if you needed SP3, it isn’t easily available. I ran into that situation about 10 years ago, and we had to make a special request through our TAP manager to get someone in Redmond to dig up a copy.

In some ways it might not be important to worry about long term storage. Most of us will end up transferring our data to newer systems (and formats) over time. As we upgrade SQL Server, our databases move along to newer formats, or we abandon them because they are no longer needed. That’s fine for some data, but not all.

Long term archival and storage is a challenge, as you can see in this short look at how old films are maintained. It just touches the edges of what’s being done, and doesn’t address costs. Plenty of old films have been lost forever, and perhaps that doesn’t matter, but it does concern me. I have thousands, maybe tens of thousands of digital images. While I love the ease with which I can share them with family, and make extra copies, I am worried that perhaps the lack of a physical copy means my great-grandchildren will struggle to find evidence of my generation if there is a catastrophe or storage formats change.

This is one area of our industry in which we have a lot of maturing to do, and I hope that we can come up with some new ideas for maintaining our data for the long term, across not months or years, but decades or centuries.

Steve Jones

About way0utwest

Editor, SQLServerCentral
This entry was posted in Editorial and tagged . Bookmark the permalink.