Drobo

Western Digital brings Advanced Format to Caviar Green

These drives DO NOT work in a Drobo. We just ran some tests and they run slower and will cause problems down the road. We are working on a fix and there will be a KB coming out soon.

Western Digital brings Advanced Format to Caviar Green

http://techreport.com/discussions.x/18115

The label on the silver anti-static bag has the following info on it:

        Advanced Format Drive.  To achieve full performance

• Windows® XP, multi partition and cloning software users – use WD Align software available on www.wdc.com/advformat

• Windows® XP, single partition – set jumpers 7-8 prior to installation or use WD Align SW

• All other OS configurations – drive is ready for use as is

If I’m reading this correctly, they’re just packing more data at the same bit-density by lessening the error-correction frequency.

Granted, this is the trend, but larger blocks = more loss potential and reduced chance of recovery.

This reminds me of the old modem days… Ymodem-1K with 1024-byte blocks was faster than standard Ymodem with 128-byte(?) blocks, but only if there was a good and stable connection. When the connection was bad, you’d “lose” 8 times more data per hiccup and your download would take MUCH longer.

Or for the non-modem-minded, imagine someone’s reading something to you for to type out for them. They stop to verify with you every sentence, if you need them to repeat, they just repeat one sentence for each time you sneeze, yawn, or get interrupted. If they only stop to verify with you every paragraph, they need to repeat the entire paragraph to you.

If you never sneeze, yawn, or get interrupted, then doing paragraphs at-a-time will be faster, but if you get interrupted often, the speaker will have to repeat a lot and the process will take a LONG time.

Bad things aside, it is clever…

actually bhiga, the recovery rate dramatically goes up, larger blocks - increased chance of recovery.

Stolen from an online article:

The principle problem here is that ECC correction takes place in 512B chunks, while ECC can be more efficient when used over larger chunks of data. If ECC data is calculated against a larger sector, even though more ECC data is necessary than for a single 512B sector, less ECC data than the sum of multiple sectors is needed to maintain the same level of operational reliability. One estimate for 4K sector technology puts this at 100 bytes of ECC data needed for a 4K sector, versus 320 (40x8) for 8 512B sectors. Furthermore the larger sectors means that larger erroneous chunks of data can be corrected (burst error correction), something that was becoming harder as greater areal densities made it easier to wipe out larger parts of a 512B sector. As a result, the need for the larger sector is born.

winners all round!

nice article on it here:

http://www.anandtech.com/storage/showdoc.aspx?i=3691

Thanks for the link, the charts make it much easier to digest. :slight_smile:

Seems recovery speed is definitely faster. There’s less ECC data, though not less than is required for reliability. But I like redundancy, that’s why I have a Drobo.

I’m just thinking worst-case in the event that a physical data recovery needs to be done on the drive. Larger “chunks” means more data lost if a particular chunk cannot be recovered. Then again, I’m no data-recovery expert, and hopefully we never have to find out… :slight_smile:

Brandon

but the flip side is according to that article, for any given damage - its more likely to be recoverable, so it may end up be a case of 6 or two threes, would you rather have less chance of damage, or the damage being worse when it does occur :slight_smile:

I just want my drive to tell me when it’s using spare sectors so I know the countdown has begun and I need to start saving for a replacement, hehe. :slight_smile:

Hear hear. What he said. We need access to more atomic S.M.A.R.T. readings.

atomic SMART? tells you when your nuclear reactor is about to go critical?

That’s MAXWELL S.M.A.R.T. to you, buddy!

Seriously, there are about fifteen S.M.A.R.T. parameters reportable by a drive, plus one overall “sick or well” status bit. Many systems only look at the latter, which is not turned on just by the use of a few replacement sectors. So if you want to track the rate at which a drive is grabbing replacement sectors, you need the full kaboodle reported. Otherwise you won’t find out till it’s too late.

@ jennifer - when you say there will be a KB soon. Do you mean an entry in the knowledge base saying dont use these,

or will there be a firmware update which allows you to use these drives (which are evidently the future)

That there will be a KB article on these drives

http://support.datarobotics.com/app/answers/detail/a_id/351

ooh ok :slight_smile:

i hope this hurries them along to released greater than 2tb drive, even with a 'pro im out of space

As long as DRI supports it and it gives me larger drives, I will be happy. It’d be nice to actually be able to reach that 16TB “ceiling” on my Drobo. :slight_smile:

For DroboElites:

Data Robotics has released a firmware update, version 1.0.1 that supports the 4K sector drives. It is important that you update your DroboElite’s firmware BEFORE you place any of these drives into your DroboElite.

oh wow

i can only presume if they have it working, then it may trickle down pretty quickly

of course since i now have eight 2tb drives, it wont affect me until i start putting 3tb drives in there

Ah jeeze. I won’t know until I’m back in the office tomorrow, but I have at least one of these just in. It’s either in an external drive used to backup the DroboPro, or it’s just recently inserted into the DroboPro.

Assuming it’s in the Drobo, what are your findings for backing out?
Is it as simple as replacing it, or has it’s integration changed the configuration/integrity of the diskpack.

Just replace it.

This problem leads me to the following three point:

  1. Do you add a flyer to the Drobos that are leaving the factory at the moment, mentioning in big letters that these drives are not supported (at the moment) and that the customers should check the KB article you’ve mentioned above?

  2. Why don’t you bring out a new firmware for all the Drobos which checks the discs before they are added to the data pool. If the disc is one of the new 4k-drives the Drobo simply shouldn’t start to add the drive to the pool and take the data at risk.
    That should be easily doable by reading the SMART info of the new disc.

  3. Why don’t you update the Dashboard app to pop up an error “Sorry the disc you just added is not supported at the moment! Please read KB article XY.”

  4. What happens anyway if you add one of the 4K drives to the drobo?

cu AssetBurned

cough four points cough

@Jennifer: Do you have any updates regarding new, 4KB-sector-compatible firmware versions for the DroboElite’s older brothers? I hope it won’t end up the “you-want-advanced-format-buy-the-elite” way… :wink: