Network Computing is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Tape Canceled, Part II

10:00 AM -- My post, "Tape: Canceled Due to Lack of Interest?," caught the attention of blogger Jerome Wendt of DCIG in his blog entitled "Murphys Law and Continued Innovation Ensure Enduring User Interest in Tape." Wendt's point was while there is a lot of attention being paid to disk-based backup, tape will still remain viable in the data center for years to come.

While I don't disagree that tape will be with us for a while, my challenge to tape manufacturers remains the same -- INNOVATE. Tape's survival and continued usage is based almost solely on what it can do right now, capabilities it has had essentially forever like portability and cost. The problem is the gap those advantages have had over disk has been in continual decline while tape's disadvantages have not been properly addressed.

What type of innovation am I looking for? How about cataloging? One of the challenges with tape is the ability to quickly find out what is on an old tape, and also to quickly find the position of data on the tape. Having the right tape and then moving to the right area of that tape -- quickly -- is critical for an optimal restore experience.

Tape has had the ability to leave file marks on media for a long time to help in scanning a tape. The last thing you want to do is scan a couple of hundred tapes from end to end. The problem is only a few application suppliers have taken advantage of marking and, if they do, it is limited. Most backup applications keep a relatively short file history catalog (90 days to 1 year), which may contain file mark information. The file history database on backup applications can get quite large very fast. As they get larger, other than the obvious capacity concerns, there are also issues with database speed and corruption. As a result, after 90 days or so, automated pruning will take place to keep the file history database under control. The problem is that after a piece of tape media is out of the main media database you have to rescan the entire tape from beginning to end.

A solution? Years ago, Sony Corp. (NYSE: SNE) had the idea of putting a RAM chip on each piece of media. The idea was that this memory could store catalog information of what files were on the tape and where on the tape the files were. With the capacity of flash memory available today, imagine if this were done on LTO 5 and were actually supported by the tape software. Imagine further that this was a standard that all the backup software manufacturers supported, so that any backup application could read any tape the moment it was inserted into a tape library and know instantly not only what files were on that tape but where the files were on the tape as well.

  • 1