Written by W. Curtis Preston
Thursday, 02 June 2011 00:46
Tape is inherently a more stable magnetic medium than disk when used to store data for long periods of time. This is simply "recording physics 101," according to Joe Jurneke of Applied Engineering Science, Inc.
I had heard rumblings of this before, but it was Joe that finally explained it in almost plain English in a post to this thread from hell on LinkedIn. Here's the core of his argument:
By the way, the time dependent change in magnetization of any magnetic recording is exponentially related to a term known as KuV/kt. This relates the "blocking energy" (KuV) which attempts to keep magnetization stable, driven by particle volume (V) and particle anisotropy (Ku) to the destabilizing force (kt) the temperature in degrees kelvin (t) and Boltzmans constant (k). Modern disk systems have KuV/kt ratios of approximately 45-60. Modern production tape systems have ratios between 80 and 150. As stated earlier, it is exponentially related. The higher the ratio, the longer the magnetization is stable, and the more difficult it is to switch state.....Recording Physics 101....
I had to call him to get more information. He explained how this came about. Disk drives have been pushed for greater and greater densities, which caused their vendors to create a much tighter "areal density." Tape, on the other hand, mainly got longer and fatter to accomodate more data in the same physical space. (Yes, it increased areal density, too, but nowhere near as much as the disk drive folks did.) The result is that the tape folks have more room to play, allowing them to use magnetic particles with a bigger particle volume (the V in the equation). The bigger the particle volume, the more stable the magnetism is, according to the KuV/kt equation. In addition, tapes are generally stored outside of the drive, which means their temperature is lower than disk drives. That means they have a lower k volume (degrees kelvin), which is one of the "bad" numbers in the KuV/kt equation. Having a higher V value and a lower t value is what translates into tape systems having ratios of 80-150, vs disk systems that have ratios of approximately 45-60. While I don't have an exact cite to point to in order to show these exact values, what he's describing makes perfect sense to me.
Add to this the fact that tape drives also have a lower bit error rate than disk. SATA disk is 1:10^14, FC disk is 1:10^15, LTO is 1:10^16, and IBM 3xx0 and Oracle T10000s are 1:10^17.
Add to this the fact that tape drives always do a read after write, where disk drives do not always do this.
Sooo...
Tape drives:
-
Write data more reliably than disk
-
Read it after they've written it to make sure they did (where disks often don't do that)
-
Have significantly less "bit rot" or "bit flip" than disk drives over time.
Like I said in a previous post, I think we've put these guys out to pasture a little too soon.
Comments
They tried RAIT once, it didn't take. Among other things, it makes tape's number 1 problem (streaming) worse. It also makes tape management a nightmare.
WRT your point about cost, I'll take "RAID1" on tape vs RAID5 on disk, since tape media is at least 50 times cheaper than tape.
RAID for tape is usually RAID1, which is more expensive than RAID5 or RAID6 from a media perspective. What is the current availability of RAID5/RAID6 (or RAIT) technology for tape? It has the obvious disadvantage of requiring as many tape drives as the stripe width, which has probably discouraged its use, but perhaps some sites would use it if it works well.
What else would you vote for? It can't be disk. Because if you can't afford a tape system sufficiently large enough to handle your storage needs without human intervention, then you definitely can't afford a DISK system to do that. The acquisition cost of the latter tends to be 3-4 times the former.
If cost wasn't an issue, then disk would totally be better. You can solve its long term viability (discussed in this article) by automating data movement ever 2 years. You wouldn't worry about the incredible expense of powering & cooling all of those disks when they're just storing data no one wants to read (or you'd pay for spin-down technologies). You wouldn't pay for a "dude in a truck" service, you'd pay for unlimited bandwidth to replicate all data to two backup sites.
But cost is always a factor. It's the reason people think they can't afford separate backup and archive packages (although they couldn't be more wrong). It's the reason people don't buy big enough tape libraries. It's the reason people keep plodding along with a really inferior backup product.
And as to tapes getting lost/damaged during transit, all of that risk can totally be mitigated IF SOMEONE WANTED TO PAY FOR IT. But they don't, and then they blame the tape.
No one would think of storing any important data on non-RAIDed disks these days, but we put data on tape without having multiple copies and then ship it around all the time. Why are we so hard on tape and nice to disk?
Here's the rub. People are involved in this process and they fail. It's life. Therefore the more people need to do to deploy and maintain a process the more errors there will be. That includes tape transportation (no not all of us can afford automated tape systems), media errors and not migrating from old to new technology when needed. Not to mention that test recovery from tape is one of the first things to go by the wayside when the pressure is on.
When we can solve those issues, I'll vote for tape as well.
I totally forgot to respond to your comment on line after our two hour phone con about it. ;)
If will boil your arguments down to:
1. Tapes do indeed fail; you have seen it
2. Tapes get mishandled
Tapes do fail; so do disks. But for some reason a lot of people only make one copy of their important archive or backup on tape. They would never do that on disk, but they do it on tape.
I concede your point. Tapes fail. That is why anything worth having on one tape is worth having on two.
Tapes get mishandled. I will address this in two parts.
1. Tapes get mishandled onsite
I said this in my previous comment. A properly designed onsite tape system should not have any tape handling, other than to put new tapes in. Problem solved.
2. Tapes get mishandled going offsite
There is no question that tapes being sent offsite are a risk in a number of ways. This risk has to be managed in light of the cost/risk of either not sending this data offsite at all, or the cost/risk of going to a dedupe-enabled, replication-based offsite delivery mechanism.
Let's say that you go to a completely disk-based replication system for your daily backups and to get your archive data offsite as well. The point of this article is that the long term (many years) copy of that data on both sides is better off on tape than it is on disk -- assuming my caveat on point 1 is taken care of.
You do not have to handle tapes in a properly sized tape storage system any more than you have to handle disk system. The fact that people undersize their tape libraries (and thus require tape handlind) is not a ding against tape; it's a ding against the designers.
"[disk systems] have software to deal with bit rot." Tape systems do too. What's your point?
You must have me confused with someone else. I am not a blogger for hire. The only compensation this website provides is banner ads, and there is no quid-pro-quo. They are simple advertising, as there would be in any other publication.
As to the comment to which you responded, this person put a blatant ad to his product in a comment, with a link. While I'm not a pay-to-play blogger, I don't appreciate people trying to use my platform (which I pay quite a bit of money to maintain) to advertise their products without compensation.
Hey, Norm. If you want advertising on Backup Central, we'd be happy to sell you some banner ads.
RSS feed for comments to this post