SSD Storage: 2018 in review

At the SSD level, the drives are actually doing block management, not file management. You still need a filesystem to store metadata, manage the layout of the data, etc.

An example of something that a SSD's controller does that the operating system/filesystem doesn't have to worry about is managing bad blocks. If the SSD detects a bad block, it will replace it with a working block and update the data used by its flash translation layer to move the blocks around. This is completely opaque to the operating system; as far as it knows its underlying storage works exactly the same (until there are so many bad blocks that the drive can't keep up this convenient deception).

An example of something a filesystem does that the SSD doesn't provide is storing operating system-specific file metadata, such as permissions, creation times, multiple data streams, directory layouts, etc. SSDs deal only in blocks of data, not arbitrarily-sized units, nor metadata.

The reason that this behavior isn't more tightly-integrated is because some of the details of managing the underlying flash blocks tend to be specific to type of flash, or even different models of flash. For example, the article mentions QLC flash becoming mainstream - we're finally getting to this point because previously, QLC was so difficult to manage that your filesystem had to be aware that it was writing to QLC flash to use it effectively. There are a few filesystems designed for direct flash management like yaffs[0], but this isn't quite as efficient as a SSD's dedicated processor and software stack.

[0]: https://yaffs.net/

Two interesting updates based on what's been announced so far in 2019:

PCIe 4.0 will be arriving in the consumer market this year, with a new generation of AMD Ryzen CPUs providing host support, and at least one or two consumer-class NVMe SSD controllers supporting PCIe 4.0 should be ready to start shipping in retail products by the end of the year. (The enterprise/datacenter storage market's transition is already well underway.)

Seagate became the first vendor I'm aware of to start marketing a SSD to the prosumer/SMB market for NAS usage. It's a rebadge of one of their recent enterprise SATA drives and isn't even using QLC NAND so it's probably going to be pretty pricey, but the idea of a solid-state NAS is no longer completely laughable.

I'm not sure how you got that $25 = 41 years of 120GB of storage on Backblaze.

Their current pricing appears to be $0.005/GB-month... which would be $295.20 for 41 years of 120GB of storage.

> i wouldn't trust a cheap SSD to still work in 41 years

I wouldn't trust almost any company to still exist in 41 years, including Backblaze. Preserving data over that length of time will probably require active management of the data in some form or another. DVD or Blu-ray discs might last that long... but it's hard to trust any storage medium to last for 41 years. Modern storage technology just hasn't existed that long.

I've been following the news and it's amazing how SSDs are still progressing, a bit slower but similar to Moore's Law. Prices for Samsung SSD 1TB external USB products:

- 2015 $600: Samsung T1 http://www.thessdreview.com/our-reviews/samsung-portable-ssd...

- 2016 $430: Samsung T3 https://www.storagereview.com/samsung_portable_ssd_t3_review

- 2018 $200: Samsung T5 https://www.androidheadlines.com/2018/12/samsung-t5-portable...

Another incredible exciting point is the amount of competition coming up, it's not just 2-3 large vendors anymore.

If things continue at this pace it seems sensible to expect a 10TB SSD consumer drive for under $1000 within a couple of years.

Note: not afiliated at all, just a happy consumer.

Flash cells leak and degrade. Some firmwares also have bugs - not refreshing cells frequent enough even if the drive is powered. The leaks are even worse on high-write-endurance enterprise drives. Make sure you read the specs and don't expect to be able to read without errors a SSD that was offline for 1-2 years.