Difference between Storage IOPS and Storage Throughput

Storage discussions often bring up terms such as IOPS and throughput as they help in gauging the caliber of a storage system. The other day, one such discussion churned out confusion among the readers.

So, to clarify to those, StorageServers blog brings in the difference between Storage IOPS and Storage Throughput.

Throughput- Data transfer speed in megabytes per second is often termed as throughput. Earlier, it was measured in Kilobytes. But now the standard has become megabytes.

IOPS- The time taken for a storage system to perform an Input/Output operation per second from start to finish constitutes IOPS.

Now, comes the big question and that is why IOPS and throughput matter?

Historically speaking, the performance of online transaction processing activity was entirely dependent on response time. The better the storage system IOPS was, the better was the online transaction processing rate.

But, nowadays it’s a bit more complex, as some of today’s database queries can depend as much on sequential database transfers (or throughput) as on individual IO response time. Thus, this endeavor is giving us a feeling that there are large components of response time critical workloads out there that perform much better with shorter response time.

On the other hand, high throughput is also getting a lot of importance, especially, when it comes to high sequential data transfer workloads such as data warehouse queries, video or audio edition environments and for big data workloads. In these environments MB/s is taken as a standard supreme- higher MB/s can lead to much faster workloads.

7 thoughts on “Difference between Storage IOPS and Storage Throughput

  1. The definition of IOPS given here “IOPS- The time taken for a storage system to perform an Input/Output operation from start to finish constitutes IOPS. It is usually measured in milliseconds, though some subsystems have dropped below 1msec range.” is a definition of the response time.

    IOPS is a rate of the IO operations measured in 1/sec. It is a ratio of the number of operations executed by storage within a period of time to the duration of this period of time in seconds.

      1. @Sultana – Latency
        Latency is obviously a part of every disk I/O, along with the bus overhead, seek time and transfer time

        For Best Throughput – latency is usually immaterial as to achieve high throughout its best to use fewer I/O’s but larger Transfers
        For Best IOPS – Use the smallest transfer – latency can have a considerable effect if doing predominantly random IOPS

  2. @stoneflyman – it is corrected where? the article still says the wrong thing – it should say IOPS is the count of the number of Input/Output operations performed in one second.

    Using the correct definition probably trashes the premise of the entire article.

      1. @Stoneflyman – your definition of it being fixed and mine are obviously different

        IOPS is an acronym of I/O Operations per second, it is NOT a measure of time as stated in your article.
        Modern Spinning Mechanical HDD’s do 100-200 IOPS
        SSD’s do 10,000 IOPS upwards, I believe I saw a report that a new backplane device that could do 150,000+ IOPS

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s