Thank you for your kind reply. Even my expectation was the same, but before 600 sec if i hit 10 times my requests are not being blocked, instead quota is re allocated. All these limitations that we are talking about works perfectly for the first time but on subsequent cycles it is behaving differently. Not sure why, below is what I have this on my policy. Is there any extra configuration that I have to look at? Since issue is happening on my local installation which uses dashboard to configure and kub +docker install on aws which uses config files could it be a possible bug?