IOPS versus Throughput

Question

1. What is the key difference between IOPS and Throughput in large data storage?
2. Does file size have an effect on IOPS? Why?

Quattlebaum

Answer 1

IOPS measures the number of read and write operations per second, while throughput measures the number of bits read or written per second.

Although they measure different things, they generally follow each other as IO operations have about the same size.

If you have large files, you simply need more IO operations to read the entire file. The file size has no effect on the IOPS as it measures the number of clusters read or written, not the number of files.

If you have small files, there will be more overhead, so while the IOPS and throughput look good, you may experience a lower actual performance.

Sheridan

Thanks, lets say I have a 100TB disk full of 200KB files. Chuck size is 32KB and I have 4GB data transfer per sec. How can I calculate IOPs for this system – Quattlebaum

@halilenver: You can't calculate it, you measure it. You can calculate the theoretical maximum by dividing the througput with the chunk size, but the actual IOPS will be lower because different IO operations take different time. If you factor in the average seek time you could make an educated guess on how close you could reasonably get to the theoretical maximum. – Sheridan

I would like to understand the difference between iops and throughput more. Is there any optimization that that could increase iops but not throughput? or vice versa? Just like decreasing latency doesn't necessarily increase througput? – Frieder

@ErbenMo: Decreasing the latency will affect the IOPS and throughput about the same, as it's just removing wait time. If you for example change the chunk size, that will generally affect the IOPS greatly, but only affect the throughput marginally. If the IOPS times the chunk size is close to the throughput, then you have a good chunk size for the size of the files that you have. If the chunk size is much too small or much too large you get a lot of overhead. – Sheridan

No necessarily the following cases, e.g., if the wanted small pieces read from/write to a file are scattered in many blocks, then you may need a lot of IOPS, while the throughput doesn't follow the IOPS. Considering the expensive track-seeking efforts on HDD, the random but small-pieces reading/writing will cause big troubles, and throughput doesn't follow IOPS. – Flighty

Answer 2

This is the analogy I came up with when talking about Throughput and IOPS.

Think of it as:

1. You have 4 buckets (Disk blocks) of the same size that you want to fill or empty with water.

2. You'll be using a jug to transfer the water into the buckets. Now your question will be:

• At a given time (per second), how many jugs of water can you pour (write) or withdraw (read)? This is IOPS.

• At a given time (per second) what's the amount (bit, kb, mb, etc) of water the jug can transfer into/out of the bucket continuously? This is throughput.

Additionally, there is a delay in the process of you pouring and/or withdrawing the water. This is Latency.

There's 3 things to consider when talking about IOPS and Throughput:

• Size (file size/block size)
• Patterns (Random/Sequential)
• Mix (Read/Write) percentage

Hypercorrection

Take an instance, if you have need high throughput you would always need higher IOPS? or is there any other mechanism in between like queue – Selfexamination

@Selfexamination I think if the chunk size is bigger (big jugs in this example), fewer IOPS could also benefit from high throughput. – Prefecture

What happens if I have high throughput but low IOPS? For example, what's the performance difference on large/small files for the following two setups: a 12 disk raidz1 which has 11x throughput but 1x IOPS, versus three 4-disk-raidz1 striped which has 9x throughput but 3x IOPS? – Cristinecristiona

Answer 3

The Disk IOPS Describes the count of input/output operations on the disk per seconds, regardless block size.

The disk throughput describes how many data may be transferred per second, so the block size play a huge role upon calculating the throughput required by app

Let's consider as the sample the 3000 IOPS and SQL database engine, the block size in terms of db engine is called the page size and for SQL Server it's equal to 8 KB. If you wish to calculate the actual throughput, if the IOPS defined, you will end up with the formula below:

throughput = [IOPS] * [block size] = 3000 * 8 = 24 000 KB/s = 24 MB/s

Caius

therefore they are related. One thing pls: is a single IO operation always exactly "big" as a one block size? – Acquiescence

are the limits of iops and throughput independent of each other? For example in your calculation you have taken blocksize as 8kb, by taking it 1024kb we could say throughput is 3072MB/s. But that may not be the case since we might hit throughput limit. In that case we would have to work backwards, lets say max throughput is 48MB/s, if our block size is 32kb we would only be able to reach 32000/48=666 iops. Can you please confirm this understanding? – Kainite

Answer 4

IOPS - Number of read write operations mostly useful for OLTP transactions used in AWS for DBs like Cassandra.

Throughput - Is the number of bit transferred per sec. i.e.data transferred per sec. Mainly a unit for high data transfer applications like big data hadoop,kafka streaming

How

Answer 5

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

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.

More about this see: What is the difference between IOPS and throughput?

Abidjan