Before writing about pausing (and resuming) Azure SQL Data Warehouse (DW) it makes sense to discuss the architecture of this product. At a high level it involves a control node, a MPP (Massively Parallel Processing) engine compromising of compute nodes and storage. Perfectly summarised by this image.
There are a few ways to scale a SQL elastic pool. For this blog post I show you how to scale up. It can be done via the Azure portal and Azure PowerShell but not T-SQL.
I would say the PowerShell route is the easiest. Connect to your account and issue the below code. Here I am going from a 100 edtu pool to a massive 2000 edtu pool whilst tweaking the min/max setting.
The below image is a beautiful picture, now it could be worse. The red line and the green line could peak at the same time and for a very long time or the blue line could behave the same as the red line and peak at the same time as the green line. Regardless of the situation the point of this blog post is when you are hitting your eDTU (elastic database transaction unit) limits within your elastic pools, tune your queries and do not knee jerk and just scale up (straightaway that is).
Yes, a really nice new shiny feature where we have the ability to suspend and resume the encryption scan for TDE – Transparent Data Encryption which is available in SQL Server 2019. (Tested against the latest version CTP 2.4)
There is a new way to setup Azure elastic jobs to run against a target group of databases (targeting an elastic pool). I actually found the process quite messy ( especially when trying to setup all the security access via T-SQL). Setting it up via the Azure portal is not currently an option. The first element remains the same, the need to create a “Job database”. This is the central database, think of it as the master. (Not really the master database in SQL Server but holds a lot of metadata etc) Then you need to define the group (usually the elastic pool), create credentials for access and then the job itself.
All About Tech 