SQLGear

We have moved to cloud to leverage the benefits of Agile infrastructure​. We can scale our infrastructure​ based on current load they are serving. It gives us ability to meet the hardware need during peak hours and also to save on cost during off-peak hours. But often we do not know in advance when ​​​​will our resource be overwhelmed by user requests. Take the example of a database in cloud. We can scale it, but scaling during a business hour will disconnect the users, at least for a fraction of a moment. Assume our database was already on high usage, which automatically triggered the scale activity, means many users are connected and all of their connections are impacted due to scale up (or down). Some users might retry, but some might not and contact the support to look into the issue. We do not want to give a bad user experience to our customers, hence avoiding scaling activities during business hours. But this does not allow us to fully utilize the scaling capabilities of cloud