looking to learn after a disastrous interview onsite

background : C/CPP Modem SW eng got asked a question at my onsite say you have an array of numbers (assume size 10 for now) , find the minimum number which occurs atleast 2 times. i was asked to write code. i iterated over the array and created a frequency map (hashmap of number and their count) . then iterate over the map and for every (key,value) pair such that if the value is >= 2 , i will throw it into a min_heap (priority_queue ) . the top element of the min_heap will give me my answer. was this the correct approach ? i was really nervous at this point . the interviewer was ok with it and asked me to run my code against a sample (which i had already done ) now he asked me to scale my code 1 . what changes will i make to my code to make sure millions of servers can access the data and get the minimum value. which is turn can provide output to another millions of servers. 2. if there are multiple ppl trying to read and write at same time, make changes in code to handle that. for #1 . i drew a blank. (i had learnt system design on how to scale but dont know what code can i write ) for #2 i said it was a concurrency issue and that i would protect data access using a mutex . He said me what is a mutex !? . i explained that to him. He was not satisfied. He ended the interview saying “you need to think of scale when writing code” . And of course i got a reject. i am looking to find about ways to solving such questions. how would you respond ? it would be great to get a list of questions which i can prepare with for such issues on future also got asked “LFU cache” . Talked thru the design . interviewer asked “have you seen this question before” i said not exactly the same but have done a ton of similar questions for practise as said in the recruiter prep material. was i wrong in saying that ? please note i had only 15 mins to think of an answer and write code because the other 30 mins were taken by mandatory behavioral questions. tc : 120k 5 yoe