In Azure PAAS , When I create a SQL Database , it asks me to create a SQL Server why is this? Also as there are elastic pools why need a SQL Server?
Azure SQL Database and SQL Server PAAS - why need both?
Before answer this, let's see what is an Azure SQL logical server.For more information about Azure SQL logical server, please refer to this tutorial.
A logical server acts as a central administrative point for multiple single or pooled databases, logins, firewall rules, auditing rules, threat detection policies, and failover groups. A logical server can be in a different region than its resource group. The logical server must exist before you can create the Azure SQL database. All databases on a server are created within the same region as the logical server.
So the Azure SQL server is the dependency of the azure database. If you want to create an Azure sql database, an existing Azure SQL server is required. It is similar to a SQL Server instance that you may be familiar with in the on-premises world.
Related
We are looking at options for moving our on premise SQL Server(s) to Azure and trying to understand whether we will be able to run cross database queries should we have data residing across multiple database technologies both in Azure ( specifically Azure Managed Instance, Azure Synapse Analytics, Azure SQL Database), and in an on-premise SQL Server instance.
We cannot find much information anywhere on whether these are supported and would appreciate if any of ye could help in filling out the table below:
TO->
Azure SQL DB
Azure Managed Instance
Azure Synapse Analytics
On Premise SQL Server
Azure SQL DB
Supported through Elastic Search Query (Ref: https://azure.microsoft.com/en-us/blog/querying-remote-databases-in-azure-sql-db/)
?
Azure Data Share supports sharing of both tables and views from Azure SQL Database and Azure Synapse Analytics (formerly Azure SQL DW), and sharing of tables from Azure Synapse Analytics (workspace) dedicated SQL pool. Sharing from Azure Synapse Analytics (workspace) serverless SQL pool is not currently supported. (Ref: https://learn.microsoft.com/en-us/azure/data-share/how-to-share-from-sql)
Azure SQL database doesn't support the linked server property so you wont be able to access on prem tables in Azure SQL database and the elastic query in Azure SQL database is to query tables between 2 Azure SQL databases and not On prem. (Ref: https://learn.microsoft.com/en-us/answers/questions/289105/how-can-i-query-on-premise-sql-server-database-fro.html)
Azure Managed Instance
?
?
?
Available through the use of Linked Servers (Ref: http://thewindowsupdate.com/2019/03/22/lesson-learned-81-how-to-create-a-linked-server-from-azure-sql-managed-instance-to-sql-server-onpremise-or-azure-vm/)
Azure Synapse Analytics
?
?
?
?
On Premise SQL Server
?
?
?
Using a linked server you can query data in an Azure SQL database from an on premised SQL Server (Ref: https://learn.microsoft.com/en-us/answers/questions/289105/how-can-i-query-on-premise-sql-server-database-fro.html)
Azure SQL database need elastic query to achieve cross database query. It doesn't support create linked server.
Azure Managed instance has almost same feature with on-premise SQL server, you could use USE statement to execute cross database query. It's same with local SQL Server.
Azure Synapse Analytics also doesn't support cross database query.
Per my knowledge and experiense, I will put ✔ or X in table to express support or not supported. please ref:
TO->
Azure SQL DB
Azure Managed Instance
Azure Synapse Analytics
On Premise SQL Server
Azure SQL DB
Supported through Elastic Search Query (Ref: https://azure.microsoft.com/en-us/blog/querying-remote-databases-in-azure-sql-db/)
X, not supported
Azure Data Share supports sharing of both tables and views from Azure SQL Database and Azure Synapse Analytics (formerly Azure SQL DW), and sharing of tables from Azure Synapse Analytics (workspace) dedicated SQL pool. Sharing from Azure Synapse Analytics (workspace) serverless SQL pool is not currently supported. (Ref: https://learn.microsoft.com/en-us/azure/data-share/how-to-share-from-sql)
Azure SQL database doesn't support the linked server property so you wont be able to access on prem tables in Azure SQL database and the elastic query in Azure SQL database is to query tables between 2 Azure SQL databases and not On prem. (Ref: https://learn.microsoft.com/en-us/answers/questions/289105/how-can-i-query-on-premise-sql-server-database-fro.html)
Azure Managed Instance
X, not supported
✔,same with local SQL Server
X, not supported
Available through the use of Linked Servers (Ref: http://thewindowsupdate.com/2019/03/22/lesson-learned-81-how-to-create-a-linked-server-from-azure-sql-managed-instance-to-sql-server-onpremise-or-azure-vm/)
Azure Synapse Analytics
Azure Data Share
X, not supported
X, not supported
X, not supported
On Premise SQL Server
✔,linked server
✔,linked server
X, not supported
Using a linked server you can query data in an Azure SQL database from an on premised SQL Server (Ref: https://learn.microsoft.com/en-us/answers/questions/289105/how-can-i-query-on-premise-sql-server-database-fro.html)
Please choose the suitable Azure database version according your request.
HTH.
AFAIK there is no cross-DB facade that provides a single interface to talk to multiple Databases at the same time. Be it on-prem/in-cloud or SQL-Server/Synapse/MySQL/...
There are individual ways and means by which you can access a single Database from somewhere/anywhere. E.g. accessing an on-prem DB from code in cloud or access a cloud DB from code running on on-prem "servers". List of interfaces available is specific to each "source" and "target" combination.
I would like to know what extra benefits one get by choosing Azure SQL Managed Instance compared to Azure SQL DB PaaS. I know SQL Managed Instance is offered as a vCore based purchasing model only. Apart from this what is the extra add on and benefits that one gets over the other. Any reply would be appreciated.
With Azure SQL Managed Instance, you essentially get a full fledged SQL Server that you can control any way you want, just like you would control a locally configured SQL Server. All the power and access and customization you want.
With, Azure SQL DB PaaS, you are essentially getting a database service, so, you give up a lot of control.
For example, take server collation. With the database service, SQL_Latin1_General_CP1_CI_AS , is all you get. With the Managed Instance, its your server. So, go ahead and select whatever collation you want, just like how you would select the collation at the time of creatoin.
Another issue is with auditing, if that is something that is important to your setup. with SQL Managed Instance, auditing happens at server level, because, you are getting the full database server. With the database service, it only database, because, you are only getting a database.
These are just the main details that I found. more details here at this Azure doc - https://learn.microsoft.com/en-us/azure/azure-sql/database/features-comparison
Of course, Managed is going to cost you more because you are paying for the license of the SQL server too.
Lastly, for me, this is the almost (not the exact same thing) like you running your own File Server on a Windows VM on Azure (Managed Instance) versus, just using Blob Storage. In both cases, you are just trying to store some files, but its how much control you have.
I would say think about Azure SQL DB vs Azure SQL Managed Instance (MI) as
Azure SQL DB = Resources dedicated to individual DBs like a container. They are grouped under a Azure SQL Server but that SQL Server is just for grouping.
Azure SQL MI = Almost same as on-prem SQL Server except you don't have to worry about OS, backups, high availability.
Here is a good comparison.
https://learn.microsoft.com/en-us/azure/azure-sql/database/features-comparison
Hopefully this will help
Azure SQL Database - The usual PaaS way. You have vCore, DTU, and Serverless billing mode. And Elastic pool support.
Azure SQL Managed Instance - Similar as below. When you want to use instance-scoped features of Azure SQL Managed Instance like Service Broker, CLR, SQL Server Agent, and Linked servers. As if you have an SQL Server on premise, Azure is responsible for patching, upgrading version etc.
SQL Server on Azure Virtual Machines - Similar as above. But you are responsible for OS and SQL server upgrade.
Not sure if this is an option? Ideally looking at something like an openrowset query from a stored procedure within SQL server or a web job.
If so, are there any resources available to help learn how to make SQL Server and Cosmos play nicely together?
If your SQL Server in Azure is managed by you on a VM (IaaS) you should be able to add the Cosmos DB as "SQL Server Linked Server" e.g. by adding as ODBC Source.
Then it's also optional if you link the Cosmos DB PaaS service to your Vnet or access it via public endpoints.
Not 100% sure if this will also succeed with Azure managed instances (which is quite new in GA).
I have a "pay as you go" Azure subscription. I am trying to keep costs down.... Do I need the resource "SQL Server" when using the SQL database resource? As I read all the supporting documents SQL Server will be used on Azure VMs to extend on-premises SQL Server. I am not doing this, I am hosting a web application on the Cloud services (Classic) resource that is connecting to the Azure SQL database.
Thanks for the help!
Azure SQL Database gives you most of the functionalities of a "Standard" SQL Server database and is (in most cases) the choise with the lower costs.
It also provides you a set of additional functionalities (some of them needs to be enabled first or are part of the higher service tiers) like an out of the box 3-node failover cluster, geo-redundancy, automated backups, etc.
If you need additional SQL Server features like SQL Server Analysis Services, which are not part of Azure SQL DB or aren't provided as another Azure service, you need to create a Virtual Maschine with a real SQL Server installed.
The "SQL Server", which is hosting your Azure SQL DB, is just a wrapper and provides you only minimal features like user and role management or your firewall settings.
If your application is connecting only to Azure SQL database and using its features, you don't need SQL Server license. You just need to have the Azure subscription and pay for the SQL database(s) that you are using. However when you create an Azure SQL database, it will prompt you to create a "Server" resource which acts as a logical group for all of your SQL databases that you create within that "server". This server is NOT charged separately and just provides a logical grouping as well as a common connection point for your SQL databases. Your billing is always based on the SQL Database(s) that you create and use.
hope this helps.
Srini Acharya
Generating output to a Azure SQL database is supported, but I was shocked when I found that the portal does not allow to specify a SQL Server database running on a VM. Is not this supported?
We need to store lots of data coming through the ASA jobs, and use SQL Jobs, that's why we were planning to use a SQL Server VM.
Thanks!
You cannot configure the SQL Database running on VM as an output to the ASA job.
However, Azure provides SQL services with 2 variants
Microsoft Azure SQL Database (Azure SQL Database) as PaaS
where lower stack is managed by Microsoft Azure and billed as pay-as-you-go model.
and
SQL Server in Azure Virtual Machine (VM) as IaaS where user owns the VM and make any changes, including licences for the SQL database.
the Microsoft Azure SQL Database provided as PaaS is configurable as
ASA output.
one idea might be to create and Event Hub output for the ASA and then consume it from there using any sort of application to write into an IaaS SQL DB. The application that consumes the data can also be hosted as a Web App as well.
Hope this helps.