If you fail to plan for your scale, you’ll end up with no workload at all

targetToday is “Cyber Monday”, the artificial “holiday” that stores have decided to create in order to separate customers from their money on the first normal work day of the holiday season. You’d think that as online retailers created this holiday, they would be prepared for it.

Apparently you would be wrong if Target’s webpage was any indication from earlier today. On the biggest online shopping day of the entire calendar year Target’s website couldn’t handle the workload and went offline telling users:

“So sorry, but high traffic’s causing delays. If you wouldn’t mind holding, we’ll refresh automatically and get things going ASAP.”

Planning for peak workloads is VERY important. If your applications are hosted on-premises this can be harder to deal with as you have to have hardware sitting around that you can run the workload on. If however you are running in the Microsoft Azure cloud you can easily scale your databases up and down without any major configuration changes. You just change the size of the databases in the portal and give the Azure backend processes some time to move things around and you are back up and running.

If you are using SQL Server or Oracle in a VM and you need to scale up, simply change the size of your VMs to a larger size. The VM will need to restart (so hopefully you have HA in place so you’d scale up the passive node, restart it then failover to it) and you’ll have more CPU and memory to work with.

If the web farm is what needs to scale up and down, the Microsoft Azure platform can handle that for you as well simply by using the auto-scaling features for web farms which allows Azure to add more web servers as needed as load increases, then remove servers when load decreases. For workloads that burst like they do on Cyber Monday, this is a perfect use case.

Now all this requires some pre-planning as you can’t just flip all of this on if you aren’t working in Microsoft Azure already. But if you are, your scale problems become very easy to solve.

If you are interested in moving to Azure so that you can have this flexibility, we would love to help you move from your current hosting provider to the Microsoft Azure cloud. Contact us and we will schedule some time to talk about your systems today, the problems that you are seeing, and how Azure can solve those problems.

Denny

The post If you fail to plan for your scale, you’ll end up with no workload at all appeared first on SQL Server with Mr. Denny.

Share

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Trust DCAC with your data

Your data systems may be treading water today, but are they prepared for the next phase of your business growth?