This was my third time doing big-N style interviews in the last few years, so I didn't have to study as much as the first time. YMMV, but my process is to solve the classic expository problems in CLRS - like lecture hall scheduling, max subarray, and rod cutting - without consulting the textbook. Of course, I fail horribly at some of the problems and end up consulting the textbook :-)
After I reread the things I forgot about - implementing things I was hazy on, like Dijkstra, on the way - I solved a handful of leetcode questions, read a few system design analyses, attended an onsite at a "practice company," and finally scheduled my big-N onsites. All told, I studied about 4 hrs per day for five days. But I think it's really important to be honest with yourself; this process seems to work well for me, but your ideal process might be different.
After I reread the things I forgot about - implementing things I was hazy on, like Dijkstra, on the way - I solved a handful of leetcode questions, read a few system design analyses, attended an onsite at a "practice company," and finally scheduled my big-N onsites. All told, I studied about 4 hrs per day for five days. But I think it's really important to be honest with yourself; this process seems to work well for me, but your ideal process might be different.