This is all very nice but...
...BRING BACK QUANTUM BIGFOOT. What's that? 'No one makes 5.25 inch drive bays any more', they will if Bigfoot comes back.
Seagate sub LaCie has pre-announced a 6TB near line disk drive from its parent and it doesn't appear to use the slow write shingled magnetic recording technology. LaCie's news came out with the NAB event in Las Vegas, which opened on Friday. There are details of the drive on Seagate's website, although the drive has not been …
Actually no. From the top:
1: Card reader (OK, thats 3.5" device in an adaptor)
2: Blu-ray re-writer
3: Drawer. Yes, a drawer. Its a handy place to keep SD cards and USB sticks.
4: Empty
5,6,7: Additional fan
I think they're quite often filled on gaming rigs, either with water-cooling reservoirs or relocated hard disk racks. 5.25" drives themselves might be nearly extinct (optical drives seem to be just about hanging in there), but the bay itself isn't going away for some markets - its handy to have a largish, standard sized hole to put things, even if its not holding the thing it was originally designed for.
Technically, I was correct: 5/7 bays are empty of drives. In fact, if the card reader is seen as (a) steam-punked USB port(s), arguably 6 are empty of drives. Otherwise, you could say 'they're not empty; they're storing some extra air--you know, for later'.
Actually the physics works better as the platters get smaller. I doubt you could get this areal density working on a 5.25inch platter, because it would flex by more (vertically) in response to any external vibration source (even if all internal vibration modes could be tamed). Then you'd have to get the head "flying" right over a greater range of platter-surface velocities, including considerably faster velocities towards the edges of the larger platters. That, or drop the spin rate and suffer greater latency. In fact latency would almost certainly increase anyway, because of the larger head arm needed and (again) the greater time taken for vibrations to damp down after it's moved.
If it were possible to pack the tracks closer, there would be considerable benefits from moving everything to 2.5 inch drives spinning at 10,000 rpm or faster. Perhaps when HAMR arrives?
Had a 1TB USB external drive from seagate, shoddy case,
WD cases are much more solid plastic, case has never cracked due to being dropped or lost the usb socket inside the case forcing, me to crack it open and convert it to internal drive (not sure I could crack open a WD drive case
"me to crack it open and convert it to internal drive (not sure I could crack open a WD drive case"
Only if you want to connect via usb not sata. The last time I cracked the case open on a WD drive to do that - it had an extended logic board that included the usb socket - no sata connector to be seen.
The case never quite fitted back together properly either after opening it, but a little dab of superglue soon fixed it.
Next thing they'll come up with is gonna be a "drive" that has an RJ-45 connector on one end, S-ATA on the other and they'll offer "Cloud-HardDrive" where all data is directly streamed from the cloud. They'll go with 10TB and above at 50+ quid per month + 10 quid per month if you want the WiFi variant.
Just DOCs and MP3s, huh? Don't worry, you won't have to suffer through any more messages like this one I remotely accessed from your system:
"Your attempt to download allthepornontheinternet.zip has failed at 99% due to insufficient disk space. Free additional disk space and try again."
Yes but how long until they release say a 8TB drive...but it actually only has purely 7TB available?
They could get away with it in the olden days when you got 77GB of a 80GB HDD or 113GB of a 120GB but now the old measuring system is starting to look downright stupid when in fact you arent getting the storage of what would have been a large HDD a few years ago.
When 8TB+ drives arrive are they going to give you a free 1TB drive to make up or have to wait till they can make a 9TB drive and label it as a 8?
I really don't think they can get away with it for much longer. We need a new size system. Just tell it what it is. Usable size only please.
"but now the old measuring system is starting to look downright stupid"
You know what would look stupider? Marketing drives as 3725GiB instead of just keeping it a nice round 4000GB.
And does anyone really look at their free space on a blank HDD anymore? Most people just start shovelling in the pirated movies, TV shows and porn these days and only worry about free space when their OS starts hinting about it.
But aren't these reduced sizes based on formatting. The amount of decrease would depend on the filesystem, and the options for that filesystem when created (block size/inode count/what have you).
If someone wrote a highly inefficient filesystem that just sticks files one after another, then you might get much closer to the manufacturer's capacity.
Aw bless you AC. No, that was exactly my point, formatting losses are what people call the non existent capacity difference when they don't understand the difference between TB and TiB. The numbers I posted actually represent the same capacity in different units (one 1024 based binary and one 1000 based decimal). OSX actually correctly displays using the decimal units while Windows uses binary measurement but wrongly uses decimal notation.
"Yes but how long until they release say a 8TB drive...but it actually only has purely 7TB available?"
They never will, drive manufacturers have always supplied the capacity stated. As I said, a 6TB drive gives you 6TB or 5.46 TiB of capacity. An 8TB drive will supply 8TB or 7.28TiB of capacity.
They aren't getting away with anything, they do tell you what it is, and it's not their fault you don't understand. They are using standard SI units quite correctly :)
I know all about it. I understand exactly. Been toying with hard drives for over 20 years.
But how long will enough of dumb ol' Joe Public take it before they take out another class action asking where their missing 1000GB went and the Judge thinks that actually that is a fair point. Before it may have been a forgivable 5GB...
Like buying a 800 gram jar of coffee and finding 100grams is jammed solid at the bottom.
I dont see the problem with stating the actual average formatted size instead of the unobtainable. If only 5.5TB is ever usable its classed as a 5.5TB drive. Simple. Either that or as mentioned come up with a more efficient formatting solution.
I didnt mind when the amounts were within reason but give it a few years and you'll be looking at 20% or more of stated capacity.
You don't seem to understand it though since you're insisting that some space is missing:
"I dont see the problem with stating the actual average formatted size instead of the unobtainable. If only 5.5TB is ever usable its classed as a 5.5TB drive."
I have said several times that a 6TB drive gives you 6TB of USABLE space. You seem to be ignoring the i in the binary measurements.
"I didnt mind when the amounts were within reason but give it a few years and you'll be looking at 20% or more of stated capacity."
As I said, the stated capacity is EXACTLY the usable capacity with no losses whatsoever.
FWIW my current 2Gb seacrates are at 20k hours with no problems (YET)
Treat the drives like eggs, plus use decent vibration isolation and they last quite a while.
It may come as a surprise but some of the most significant drive impulse damage events happen when putting the screws in and/or tightneing them. Don't let that screwdriver slip.
I think you'll find that many of the 'Seagate' SCSI / ESDI Wren drives were actually re-badged CDC/Imprimis drives - in 1990 we had a 4766E full height 5.25 with a massive 660MB formatted that came in at a couple of grand: the CDC model number was 94196-766 and Seagate ST4766N. Nothing much has changed over the years. Latterly came in useful for keeping your coffee warm :-)
As the NSA are already hosting copies of all the world's p***n collections, perhaps they could just de-dupe the whole lot and make the resultant mongo collection available as a free global resource, a bit like the CIA World Factbook, thus freeing up I guess around a zettabyte of storage. Subject to concurrent serving capabilities they could reduce storage requirements sufficiently to resolve the energy crisis overnight, charge a modest membership fee for all access (so putting pirates out of business, but guaranteeing higher revenues for the "content generation" side). Obviously Mumsnet wouldn't like it, but they're just another hypocritical fundamentalist pressure group who could be monitored to make sure they aren't buying too much bleach.
Obviously the NSA would need a more catchy name, I believe Grumblebook isn't taken. Or Grumblr.
...the rebuild time will be on those drives in the real world. 2 days? We probably won't see any real midrange or enterprise arrays running these until they've validated the crap out of them in their labs but I wonder what Nimble could do with these and some of the latest-gen MLC modules coming out. XIV is another that could get a lot of out them with its rebuild performance and massive caches.
The problem here is that you'd then need to transfer the data over Ethernet or more likely the WAN which would take even longer. If your second NAS is local then there is far more benefit to be had with RAID 60 than with RAID 51 (effectively what you suggested with rsync). Realistically you'd need to do some maths to work out how long each recovery will take but once NAS 1 fails you'd be right back in the situation of having a RAID 5 array waiting to die - almost certainly using disks from the same batch as the failed ones, and Google's study into disk failures suggested that batch number was a bigger influence on failure than heat, humidity, vibration, or any other traditional ally recognised cause of failure, with those others being statistically insignificant.