Something that is becoming a significant effort in working with SQL Server and other systems in the cloud (whether it be on AWS or Azure or other services) is the fact that updates are happening nearly constantly. There have been other columns about that here on SSWUG, but it’s important to take a few minutes and think through your implementations.
What we’re seeing quite often now is this concept of “Technical Debt” coming home to roost with cloud configurations, options, and best practices. It’s weird to say, but systems that have been around for 6 months or a year, working perfectly well, are starting to age in ways that can be surprising.
It’s beyond critical that you find a way to stay tuned in to the best practices, tools, techniques, and options for your installations and environments. Some examples where stagnation can be an issue include:
- Security best practices for your server – from access controls to certificates in use, to security of accounts accessing the system, to firewalls and other tools that can make your life so much better in protecting that data.
These can become “stale” quickly – and if you haven’t put into place dynamic updates to firewalls and monitoring, for example, you’re basically saying “I installed anti-virus software 2 years ago – haven’t updated the libraries and app since, but I’m covered, right? RIGHT?“
This can apply to the web application firewalls, the white- and black-listed IPs, applications and access points, and even the configuration of your servers (containers, instances, managed instances, etc.) and how that plays into your overall configuration and management.
It can also have an impact on how effective your monitoring is, and your early warning systems may be compromised if they are not managed and updated to new technologies and tools as well. - Billing – learning about how SQL Server impacts your billing and whether you’re on the trajectory you expected in terms of growth, usage and costs is important to the success of your installations. Storage creep is a big issue, as are automated processes that are enabled for a specific task, but never disabled. That processing can be a cost increment that goes unnoticed for a bit but adds up quickly. In one account, a simple review of services netted a 33% decrease, month in, month out, of the hosting bill, just by cleaning up and taking stock of used resources.
- Keep on top of new technologies. New deployment tools. New automation options. Scan feeds for products you’re using, stay on top of updates to tools you have deployed. Set up keyword scans in search engines to watch for alerts and reports of changes.
In some cases, a change can mean you need to update (like minor or major revisions of SQL Server) the engine, or other cases, supporting tools (like certificates, firewalls, exclusion lists, approaches to security, etc.). In still other cases, it’s critical to make sure you’re applying your own rules to your installations. Are you removing users that move on from posts that no longer need that reporting access? Are you watching over your systems and monitoring for any hiccups or performance type things you can be updating?
One of the biggest draws of the cloud is a managed environment. This is great when things come online because your SQL Server “just runs.” Truly a boon to business and the IT workload.
However, this very “set it and forget it” is creating some challenging upgrade scenarios as the cloud has been around long enough for several software engine iterations and other natural lifespan updates. NOT staying on top of these can introduce as much of a technological debt as not moving forward to new technologies in the first place. Make sure keeping on top of things is an active goal and one that is managed for your stakeholders.