Press "Enter" to skip to content

Category: Availability Groups

Database Mirroring Compatibility and Availability Groups

Published 2022-01-24 by Kevin Feasel

Sean Gallardy checks out the past:

Around 2005, mirroring was born. It was an evolution on log shipping, which is taking log backups, moving them around, and restoring them all in an automated fashion to different servers. Mirroring upped that game and created a dedicated network channel between servers (you could only have 1 principle and 1 mirror, so 2 total) so that there wasn’t this funny business of copying and restoring, additionally it allowed the mirror server to be a highly available copy with automatic failover. Since Microsoft marketing is terrible at naming things, it was originally called, “Real Time Log Shipping” which was then changed to “Mirroring” and in typical fashion you can find the unofficial “Real Time Log Shipping” name all over the place where it was never updated. (I can’t really blame them here, though, it’s hard to find all the little places you’re putting this moniker in and then having some other team tell you to change it all at some way later point)

Read the whole thing. It’s a fun read, a little sad, and helps us understand a bit of availability group behavior which might bite the unaware. I will definitely defend Microsoft’s backward-compatibility emphasis. This makes life so much easier for developers than a lot of other languages and environments. In the R and Python worlds, breaking changes are the norm, meaning that when you update packages, you can expect something to break and now that “20-minute” package upgrade ticket becomes 3 days of trying to sort out what went wrong.

Comments closed

Creating an Availability Group on Linux in Azure with Pacemaker

Published 2021-12-02 by Kevin Feasel

Andrew Pruski slams in all of the exciting nouns:

There are new Ubuntu Pro 20.04 images available in the Azure marketplace with SQL Server 2019 pre-installed so I thought I’d run through how to create a three node pacemaker cluster with these new images in order to deploy a SQL Server availability group.

Disclaimer – The following steps will create the cluster but will not have been tested in a production environment. Any HA configuration for SQL Server needs to be thoroughly tested before going “live”.

Click through to see how.

Comments closed

Index Usage across Replicas

Published 2021-11-17 by Kevin Feasel

Jess Pomfret does the math:

Last week, I was working on a project to analyse indexes on a database that was part of an availability group. The main goal was to find unused indexes that could be removed, but I was also interested in gaining an overall understanding of how the system was indexed.

Unused indexes not only take up disk space, but they also add overhead to write operations and require maintenance which can add additional load on your system.  We can also use this analysis to look for a high number of lookups which could indicate we need to adjust indexes slightly.

Click through to see how you can connect together index usage stats from the primary and secondary replicas of an availability group.

Comments closed

Operating System Error 995 on Adding a Database to an AG

Published 2021-10-06 by Kevin Feasel

Andrew Pruski troubleshoots a problem:

I was adding databases to an availability group (SQL Server 2017 CU20 instance) the other day and one database failed to automatically seed to the secondary.

When I looked in the SQL Server error log I saw this error message: –

BackupIoRequest::ReportIoError: write failure on backup device ‘{GUID}’. Operating system error 995(The I/O operation has been aborted because of either a thread exit or an application request.).

Read on to see how Andrew solved the problem.

Comments closed

HADR_SYNC_COMMIT

Published 2021-09-23 by Kevin Feasel

Sean Gallardy lays out what HADR_SYNC_COMMIT really tells you:

Initially I thought to myself, “this is the most misunderstood wait type that exists in the HA space for SQL Server”, then I realized maybe this isn’t the case… So, I pondered over this question, “is it truly misunderstood?” and came to the (possibly incorrect) realization that it is quite accurate in the general SQL Server’s users’ space of understanding. I also concluded that, really, it’s the way the wait is used in SQL Server coupled with how waits work in SQL Server, which leads to how it is viewed. Let me explain….

You’ll definitely want to read Sean’s explanation.

Comments closed

Creating a Distributed Availability Group in Azure via Terraform

Published 2021-09-17 by Kevin Feasel

Sandeep Arora has some scripts for us:

To create a distributed availability group, you need two availability groups (AG) each with its own listener, which you then combine.In this case, one availability group is on-premises and the other needs to be created in Microsoft Azure. This example doesn’t cover all of the details like creating an extended network setup between on-premises network and Azure or joining Azure active directory domain services to and on-premises forest; instead, it highlights the key requirements for setting up the availability group in Azure and then configuring the distributed AG between the on-premises availability group (represented as AOAG-1) and the Azure availability group (represented as AOAG-2).

Click through for the preparations you need in place and a set of scripts to do the work.

Comments closed

Against sp_hexadecimal and sp_help_revlogin

Published 2021-07-07 by Kevin Feasel

Andy Mallon says it’s time to give up a couple of procedures:

We recently ran into some performance problems with our login sync, which is based on sp_hexadecimal and sp_help_revlogin, the documented & recommended approach by Microsoft.

I’ve been installing & using these two procedures since I started working with SQL Server, back at the turn of the century. In the nearly two decades since, I’ve blindly installed & used these procedures, first on SQL Server 2000, and then on every version since… just because that’s the way I’ve always done it. But our recent performance problems made me rethink that, and dive in to take a look at the two procedures to see if I could do better, which made me realize, OHBOY! WE CAN DO BETTER!!

Read on to understand how.

Comments closed

Maximizing Availability Group Performance

Published 2021-06-10 by Kevin Feasel

Jonathan Kehayias has a few tips for improving performance of your Availability Groups:

Since Microsoft first introduced the Always On Availability Groups (AGs) feature in SQL Server 2012, there’s been a lot of interest in using AGs for both high availability and disaster recovery (HADR), as well as for offloading read-only workloads. The combination of the best features for failover clustering, the simplicity of data movement and synchronization from database mirroring, and the ability to offload read-only workloads to secondaries has given businesses a compelling reason to upgrade to leverage AGs.

But, as the saying goes, there’s no such thing as a free lunch, and there are several performance implications and considerations you must be aware of to have a successful deployment using AGs. This blog post will explore some of the considerations and look at how to plan, architect, and implement an AG with minimal latency and performance impact on the production workload.

Click through for those tips.

Comments closed

Automate Availability Group Failover for SSISDB 2012 and 2014

Published 2021-06-01 by Kevin Feasel

Alex Stuart shows how to fail over SSISDB in SQL Server 2012 or 2014:

Hopefully not many people are still configuring SSIS instances on SQL 2012 or 2014 – especially HA instances – but if you are, this post is for you.

If you’re running SQL Server 2016 or above, having the SSIS catalog function correctly in an AG is supported by built-in functionality to manage the DMK (database master key). In 2012/2014 however there is no such support. Without intervention this makes SSISDB unable to be opened after a failover, because the DMK isn’t open – leading to errors such as “Please create a master key in the database or open the master key in the session before performing this operation.”

Read on to see how to resolve this error, and then how to do this automatically.

Comments closed