The Secret Life of Tape, 2011 Update -- Actions - Page 2


Want the latest storage insights?

Download the authoritative guide: Enterprise Data Storage 2018: Optimizing Your Storage Infrastructure

Share it on Twitter  
Share it on Facebook  
Share it on Google+
Share it on Linked in  

There is no real trend here. The upside of this is the likelihood of seeing improvements. LTO is up and LTO is down, and there are only two data points for T10000, so that does not provide much information. As Imation makes neither T10000C nor tapes for newer IBM tape drives, I do not have that information. You might say that using an LTO-5 drive and media at full rate for 25.4 days is impossible, and I would totally agree with that. I wanted to provide the raw numbers, and everyone could put in the factors that make sense to them. The point is, you cannot write and read a tape forever. Honestly, 25.4 is a fairly long interval, but bear in mind that tape is usually for archival data, not backup data, these days.

What are you going to do about it?

First of all, I have said time and time again that tape drives do not provide information like disk drives provide soft errors and status. Disk drives provide and export an interface call SMART (Self-Monitoring, Analysis, and Reporting Technology), which provides information about soft errors. RAID vendors use it to determine when to down a drive and start rebuilding. Let me repeat: Tape does not have a standard framework to known information that is collected and analyzed. There are vendors that provide third-party products, and some tape library vendors support collection, but it is not a standard.

This, in my opinion, is a big mistake for the tape drive vendors, as you cannot track the media or drive issues without specialize software. I have said this for years, and there has been no movement by the drive vendors to change it. The only thing that can be done to prevent the media wear-out problem and save data is to buy a product that monitors both tape drives and media and indicates when a problem is brewing. Generally, what happens is that the number of soft errors increases before a tape fails. Also, without software that monitors tape drives and media, you cannot determine how many full tape passes have been made. So in another one of my continual rants about preservation of data integrity, if you manage archives and are not monitoring tape drives and media for soft errors, then you are playing with fire. The tape drive or library vendors should be providing this information for free, which they do not.

This leaves you with one choice: Go out and buy a product that monitors your large archival investment.

What is missing from what vendors tell you?

The big thing missing from the information I have is both IBM's TS1130 and 1140 tape media and Oracle T10000C media. Why don't vendors publish these statistics? From what I have seen, media vendors are scared of drive vendors and other media providers. They will not publish data for the most part, given that the information is difficult to explain. Additionally, vendors do not want to be compared to each other, as LTO is a midrange commodity product, and the real issue for most purchases is price -- reliability is a distraction. Drive vendors that also sell media do not want this information to come out, as it is difficult to explain and distracts the attention from the sale. More information in the sale process that is not generally part of the discussion just slows things down in most sales people's opinion. The problem, again, is most tape usage today is for archive, not backup. The old sales model does not work where people have a long-term interest in their data and often a single or a small number of copies of the information as compared to backup.

Therefore, I put forth a challenge to the industry:

  1. Whether the tape is enterprise or midrange LTO, all tape vendors should allow the publication of the media information. That means tape drive vendors, tape library vendors and the media vendor themselves ought to have all that public and let the customers decide if the data is meaningful and important to their operation.
  2. Tape drive vendors should get together and publish standardized error specifications, as the disk drive vendors and industry has done

I am not sure that either of these are possibilities as the vendors are not motivated to make these changes. The fastest thing that motivates a vendor is a customer request. If enough of you make these requests, maybe we will see some change, but I am not very hopeful.

Henry Newman is CEO and CTO of Instrumental Inc. and has worked in HPC and large storage environments for 29 years. The outspoken Mr. Newman initially went to school to become a diplomat, but was firmly told during his first year that he might be better suited for a career that didn't require diplomatic skills. Diplomacy's loss was HPC's gain.

Follow Enterprise Storage Forum on Twitter

Submit a Comment


People are discussing this article with 0 comment(s)