February 1, 2017 at 2:21 am
Ive been investigating redundancy of apps built on azure sql database.
From what i can see (mainly in this article by MS), one would have to write a monitoring app that continually polls the databases to check availability and performs the failover if necessary.
Two things strike me and i want to confirm if i am correct:
1. I cannot find any reference to a method for automatic failover. a person or a service/app has to issue the alter database...Failover command. is this the only way to failover?
2. I cant find any connection string that can take the primary and the secondary database names ( like in mirroring), or reference to a listener ( as in AlwaysON). Is it a case that there is no such functionality?
February 7, 2017 at 6:27 am
Within Azure SQL Database, there are two different kinds of failovers. First, each database actually consists of three different databases on three different servers within the data center (all this is invisible). You have a primary and two secondaries. If something goes wrong on the primary or they need to do maintenance, there'll be a failover, all handled within the Azure fabric. In fact, you can't control this at all, in any way. It'll just happen. When you add in Geo-replication, you effectively add three more databases within another data center, all of them acting as a secondary to your primary set. Don't think about them as sets though. As far as you're concerned, it's just a primary and a secondary database. If the primary data center goes offline, again, as part of the Azure fabric, you'll get a failover. However, with geo-replication, you can actually fail over manually. Here's an article on it.
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning
Viewing 2 posts - 1 through 1 (of 1 total)
You must be logged in to reply to this topic. Login to reply