Senior Software Engineer

Step 1 : This problem I already solved before.
Step 2 : Explained him the heap based approach and coded in Java using PriorityQueue
Step 3 : He understood the solution and asked me to code
Step 4 : After running working solution, he asked me to test with different test cases
Step 5 : He then asked time complexity

Step 1 : I first used Hashmap to count the occurence of each letter in target string
Step 2 : Then Using sliding window concept I tried to find the subsequence
Step 3 : On each finding, I was capurting results for smallest window size
Step 4 : I explained the approach to inteviewer
Step 5 : After that I did coding and ran the code
Step 6 : At the end, interviewer asked time complexity which was O(N)

Tip 1 : First I asked the scale of logs like how many logs are getting stored every minute to findout scale
Tip 2 : Then I asked How these logs are currently stored in datacenters currently. I tried to understand the system in a more clear way.
Tip 3 : I used Messaging queue to parallely fetch logs and send to the target data center.
Tip 4 : Then the further question was how we will be storing in data center. For this, we discussed the concept of DB sharding and consistent hashing

Tip 1 : I tried to understand the system more like how requests are being sending to different servers
Tip 2 : Then I gave one approach to add load balancer to distribute requests equally
Tip 3 : Interviewer asked to discuss pros/cons of this.
Tip 4 : We discussed the security concern and to handle that we can use Network load balancer instead of Application load balancer
Tip 5 : Then we discussed the increase in latency because the requests are going through one extra system
Tip 6 : To overcome this we then discussed the consistent hashing at producer level to equally send request among multiple servers.
Tip 7 : This round was all about discussing different approaches and finding pros/cons of each system