Five Things You Must Know About Data Vault

by Ian Nicholson, May 2014. Data Vault is a data warehouse modelling technique that works to the maxim “all of the data, all of the time”. This means that all your data – good or bad – lands in your data warehouse. Now while that may sound ominous, here are the top five reasons why you should consider Data Vault as your data warehouse architecture. Long-Term History Data Vault not only preserves history, but also history-of-history. In other words, you can go back and review the state of a data record at any point in time. This feature is incredibly useful for all kinds of reasons. For example, an insurance company would need to know if important data was in the data warehouse prior to a claim being made. A product may have changed an attribute (i.e. colour) but retained the same product code. Auditability Since Data Vault keeps a comprehensive history – including the ability to record where data came from, it makes the perfect choice for any environment where keeping an audit trail is paramount. Each row in in a Data Vault is accompanied by record source and load date information. Storage of all – including incomplete – Data. This purpose is served if, for example, a customer places an order, but their details are not yet entered into your CRM. The sales record can still make it through to your data warehouse and be tallied, even though there is not yet a customer record. Resilient to Change Data Warehouse projects are a journey, not a destination. Change is inevitable. The last thing you need is a brittle architecture that must be broken every time a change is required. Parallel Loading Data Vault employs an innovative table structure, involving tables called satellites and hubs that provide greater flexibility, and also the benefit of parallel loading. This means that you can deploy very large Data Vault solutions that can scale without any need for major redesign. However, as you can imagine, Data Vault’s table structure makes life a lot more complex for the ETL developer. Indeed, there are groups and forums across the internet where Data Vault ETL developers spend time asking and answering each other’s questions. Since Data Vault follows a well prescribed set of rules, it can be automated. Only BIReady offers a completely automated Data Vault solution. BIReady analyses your data source(s) and reverse- engineers your business model. From that, BIReady automatically designs a target Data Vault model for you – in seconds. Once done, it then generates all the ETL (ELT) code in the SQL of your target database (including parallel loading capability), and then a wizard takes you through the process of generating star-schema data marts for reporting. Data Vault in Days! Only with BIReady.