A curious case of the shingles The topic of shingled writes as a way of boosting hard drive capacity discussed in my article Drive suppliers hitting capacity increase difficulties earned me a rebuke from HDD<, a very occasional Reg commenter. He has this to say : Shingled recording being used in the marketplace currently? If you have some proof of this, …
You noted nobody would go on record in the original article, as well as indicating this was a down the road thing. Which means Seagate isn't shipping this currently. Sure it has the potential to affect us all. I don't know a great deal about the physics of hard drives, but even reading the bit about changing cluster sizes made that bleeding obvious. During the design phase is exactly the time at which discussions about benefits and impacts needs to be discussed. When product is already on the shelves it is way too late. As Apple are discovering and MS have beaten into them repeatedly without learning.
From the original article:
'Shingle writing is used "in external drives today … It's good for long large writes; not so good for transaction type environment with many short writes." You can't use the technique for enterprise drives because of the performance impact. The Seagate 3TB FreeAgent possibly uses the technique.'
I think there is where the confusion comes in. You say that it "is used" in the first sentence, but at the end, the only evidence you give is that the Seagate drive "possibly uses the technique."
It's quite telling where your quoting of Steve Barber begins -- you start the quote right after the definite verb "is used". That tells the reader that it's not your source making the definite statement, but yourself.
I'd like to see the full original quote from Steve Barber, to verify if he indeed did say it was currently in use, or simply that it might be already in use. If he claims it is definitely in use, then you've got yourself a big story -- especially since the manufacturers won't comment on it.
If not, I'd chalk this up to a reporting error in trying to summarize the quote.
As they say somewhere in baseball, cowboy up. This isn't a courtroom and just because something is hearsay doesn't make it invalid. Besides, if you wait to have seventeen confirmatory sources for every article you won't ever get anything out until it becomes old hat. It seemed pretty clear from the statement; "Barber reckons an interim capacity boost can come from 'shingle writing'..." that this was an interesting potential to be explored and not something "coming to a store near you". I took the Barber quote pertaining to external drives and performance as potentially a reference to back up media used in data centers and not some portable hdd picked up at Wally World. I'm sorry that what HDD said got to you but really, suck it up, take the critique, including this one, for what it is and move on.
Keep those articles coming, sir. There's always a hater or five somewhere...and I found your article very interesting. Though I admit to having questions about shingling being used on extant external drives. I don't see why they wouldn't use it...but I would wonder in what products? Most external disks seem to be repackaged desktop drives. Perhaps in dedicated media designed to replace tapes? I'd love to know!
If a certain type of external drive is shingled, then simply using a desktop drive in an external case might be faster. Or might not? (Interface limited?) Worth an experiment…
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
