The Cleaning Lady and Write Amplification

Imagine you’re running a cafeteria. This is the real world and your cafeteria has a finite number of plates, say 200 for the entire cafeteria. Your cafeteria is open for dinner and over the course of the night you may serve a total of 1000 people. The number of guests outnumbers the total number of plates 5-to-1, thankfully they don’t all eat at once.

You’ve got a dishwasher who cleans the dirty dishes as the tables are bussed and then puts them in a pile of clean dishes for the servers to use as new diners arrive.

Pretty basic, right? That’s how an SSD works.

Remember the rules: you can read from and write to pages, but you must erase entire blocks at a time. If a block is full of invalid pages (files that have been overwritten at the file system level for example), it must be erased before it can be written to.

All SSDs have a dishwasher of sorts, except instead of cleaning dishes, its job is to clean NAND blocks and prep them for use. The cleaning algorithms don’t really kick in when the drive is new, but put a few days, weeks or months of use on the drive and cleaning will become a regular part of its routine.

Remember this picture?

It (roughly) describes what happens when you go to write a page of data to a block that’s full of both valid and invalid pages.

In actuality the write happens more like this. A new block is allocated, valid data is copied to the new block (including the data you wish to write), the old block is sent for cleaning and emerges completely wiped. The old block is added to the pool of empty blocks. As the controller needs them, blocks are pulled from this pool, used, and the old blocks are recycled in here.

IBM's Zurich Research Laboratory actually made a wonderful diagram of how this works, but it's a bit more complicated than I need it to be for my example here today so I've remade the diagram and simplified it a bit:

The diagram explains what I just outlined above. A write request comes in, a new block is allocated and used then added to the list of used blocks. The blocks with the least amount of valid data (or the most invalid data) are scheduled for garbage collection, cleaned and added to the free block pool.

We can actually see this in action if we look at write latencies:

Average write latencies for writing to an SSD, even with random data, are extremely low. But take a look at the max latencies:

While average latencies are very low, the max latencies are around 350x higher. They are still low compared to a mechanical hard disk, but what's going on to make the max latency so high? All of the cleaning and reorganization I've been talking about. It rarely makes a noticeable impact on performance (hence the ultra low average latencies), but this is an example of happening.

And this is where write amplification comes in.

In the diagram above we see another angle on what happens when a write comes in. A free block is used (when available) for the incoming write. That's not the only write that happens however, eventually you have to perform some garbage collection so you don't run out of free blocks. The block with the most invalid data is selected for cleaning; its data is copied to another block, after which the previous block is erased and added to the free block pool. In the diagram above you'll see the size of our write request on the left, but on the very right you'll see how much data was actually written when you take into account garbage collection. This inequality is called write amplification.


Intel claims very low write amplification on its drives, although over the lifespan of your drive a < 1.1 factor seems highly unlikely

The write amplification factor is the amount of data the SSD controller has to write in relation to the amount of data that the host controller wants to write. A write amplification factor of 1 is perfect, it means you wanted to write 1MB and the SSD’s controller wrote 1MB. A write amplification factor greater than 1 isn't desirable, but an unfortunate fact of life. The higher your write amplification, the quicker your drive will die and the lower its performance will be. Write amplification, bad.

Live Long and Prosper: The Logical Page Why SSDs Care About What You Write: Fragmentation & Write Combining
Comments Locked

295 Comments

View All Comments

  • Anand Lal Shimpi - Monday, August 31, 2009 - link

    The tables the drive needs to operate are also stored in a small amount of flash on the drive. The start of the circular logic happens in firmware which points to the initial flash locations, which then tells the controller how to map LBAs to flash pages.

    Take care,
    Anand
  • Bakkone - Monday, August 31, 2009 - link

    Any gossip about the new SATA?
  • Zaitsev - Monday, August 31, 2009 - link

    Thanks for the great article, Anand! It's been quite entertaining thus far.
  • cosmotic - Monday, August 31, 2009 - link

    The page about sizes (GB, GiB, spare areas, etc) is very confusing. It sounds very much like you are confusing the 'missing' space when converting from GB to GiB with the space the drive is using for its spare area.

    Is it the case that the drive has 80GiB internally, uses 5.5GiB for spare, and reports it's size as 80GB to the OS leaving the OS to say 74.5GiB as usable?
  • Anand Lal Shimpi - Monday, August 31, 2009 - link

    I tried to keep it simply by not introducing the Gibibyte but I see that I failed there :)

    You are correct, the drive has 80GiB internally, uses 5.5GiB for spare and reports that it has 156,301,488 sectors (or 74.5GiB) of user addressable space.

    Take care,
    Anand
  • sprockkets - Tuesday, September 1, 2009 - link

    Weird. So what you are saying is, the drive has 80Gib capacity, but then reports it has 80GB to the OS, advertised as having an 80GB capacity, which the OS then says the capacity is 74.5GiB?
  • sprockkets - Tuesday, September 1, 2009 - link

    As a quick followup, this whole SI vs binary thing needs to be clarified using the proper terms, as people like Microsoft and others have been saying GB when it really is GiB (or was the GiB term invented later?)

    For those who want a quick way to convert:

    http://converter.50webs.com">http://converter.50webs.com
  • ilkhan - Monday, August 31, 2009 - link

    so they are artifically bringing the capacity down, because the drive has the full advertised capacity and is getting the "normal" real capacity. :argh:
  • Vozer - Monday, August 31, 2009 - link

    I tried looking for the answer, but haven't found it anywhere so here it is: There are 10 flash memory blocks on both Intel 160GB and 80GB X25-M G2, right? (and 20 blocks with the G1).

    So, is the 80GB version actually a 160GB with some bad blocks or do they actually produce two different kind of flash memory block to use on their drives?
  • Anand Lal Shimpi - Monday, August 31, 2009 - link

    While I haven't cracked open the 80GB G2 I have here, I don't believe the drives are binned for capacity. The 80GB model should have 10 x 8GB NAND flash devices on it, while the 160GB model should have 10 x 16GB NAND flash devices.

    Take care,
    Ananad

Log in

Don't have an account? Sign up now