Dynamic-Dynamodb can sometimes scale-down too quickly just prior to another scale-up. This request is to add the ability to delay the scale down by specifying that 'n' number of polling intervals must meet the scale-down criteria. This could also be accomplished by having different timers for the scale-up and scale-down.
Thanks for the feature request. I do definitely think that this feature is needed and we should implement it.
I'm on the go right now, but I have one quick thought about an alternate approach (I haven't decided which approach I think is best): We could also make the the look back window when fetching metrics from CloudWatch configurable. I know that that has been discussed previously by some people.
I'll think a bit more about how to approach this.
Added docs for num-r/w-checks-reset-percent #148
Updated command line docs for #148 and #152
I am closing this issue and will prepare the release! Many thanks for all the help with this feature @johnbartels.