What is 11 Nines of Data Durability?
What does 11 Nines of Data Durability mean?
Data durability meaning is keeping data intact and not allowing it to degrade through processes such as power loss, drive loss, array loss, and other corrupting influences. 11 nines describes 99.999999999% annual durability, and it implies that only 0.000000001% of data is expected to be lost per year.
Durability is a measurement of the many small errors that often occur in files due to the underlying media such as hard disk drives and flash storage arrays. When you read and write data repeatedly individual bytes can be corrupted or lost, damaging the data.
Getting as close to 100% as possible when dealing with data durability in relation to storage is the goal. This article will delve deeper into data durability, its importance, and everything you need to know about it.
Because your cloud storage backups will be stored indefinitely, you need a formal way to measure how resilient your cloud storage provider’s platform is.
What You Need To Know About Data Durability
The level of data durability will vary depending on many different factors, these are the number of files and object fragments, failure rates, the drives used, and rebuild time. The failure rates of drives can vary greatly and as a result can be difficult to factor correctly. For these reasons, before storing any data, you need to ensure that durability is high, as this is one of the fundamental aspects of storage.
There are many ways to store data, one of these methods is cloud storage. Cloud storage is designed to have at least 11 nines of data durability per year. This number can be difficult to understand and put into perspective. It means that if you have a billion files and objects stored, even after 100 years of storage, you will not lose a single piece of data.
Availability of data also needs to be considered because of its importance. This is also often referred to as ‘uptime’. The storage system used needs to be operational at all times so it can deliver data whenever it has been requested. It doesn’t matter how durable and reliable data is if you are not able to access it when required. When looking to store data, two of the most important factors that need to be considered are data durability and availability.
When talking about data availability, keep in mind that 11 nines are irrelevant, and 8 nines are more than enough. Why is this? Well, after the 8th nine, practicality isn’t the focus anymore, it’s more about focusing on the academic aspect of availability. This is because at this point, it would be far more likely that natural disasters would destroy multiple data centres causing data loss.
Lengthening Data Durability Time
One method of increasing the durability of data is through the use of ‘erasure coding’. This method splits up data into fragments and then adds extra pieces of that data as duplicates. This means if a single file is lost or corrupted, it can be recreated using all the split up pieces of data which have been stored separately. This method is great for large-scale storage and it allows for scaling data protection as hard drives continue to grow in size.
Multiple copies of data can be stored in different locations to lengthen the durability of data. This method is an obvious one and is typically adopted by a large number of people. It helps overcome issues ranging from individual drive failure to natural disasters such as floods damaging data centres.
Risks To Data Durability
One of the biggest risks to data durability are software bugs. There are steps that can be taken to avoid the loss in data durability due to software bugs, these involve avoiding the introduction of data-corrupting bugs to begin with. Safeguards are also put in place to detect errors and bugs so they can be dealt with swiftly before data durability loss occurs. Companies update their software frequently to remove bugs, these updates are monitored closely and there are always plans for quick rollbacks if something goes wrong. This ensures data is kept safe and is protected against corruption and loss.
Transiting data can be at risk of data corruption. Transiting data includes data that is being transferred across different networks within a cloud storage service or when data is being uploaded/downloaded to/from a cloud storage service. There is a way to protect against this type of data corruption and loss, that method is referred to as checksums. A checksum is a string of numbers and letters that act as a sort of fingerprint for files. This fingerprint can be used to detect errors in data and to ensure the integrity of data is upheld.
Greatest Contributors Of Data Corruption And Loss
When talking about data durability and the risk to data, it is essential to note the biggest perpetrators of data loss. The greatest risk comes from the combination of human error, bugs in software, viruses, and malicious act from employees or outside forces. No matter how durable data is it is meaningless if these perpetrators are not addressed and safeguards are not put in place.
One way to deal with these issues is by using immutable storage. This type of storage cannot be modified or deleted by anyone. Once the data has been written, it will remain in storage for the set amount of time that you have decided upon. If deletion of the data is attempted, even by the owner of the data, an error message will be received. This is a powerful technique which circumvents issues of data security and loss.
Durability should not be underestimated. The top providers of cloud storage services are constantly improving their data durability. These providers include companies like Google and Amazon which provide customers 11 nines when it comes to data durability. Companies that take data durability seriously design their systems for failure. This is because all hard drives will fail, so systems need to be designed to mitigate failures of any kind to protect data and avoid data corruption and loss.