Amazon Software Development Engineer II interview questions

Email test: *The problem:* A logfile contains newline-terminated, space-separated text formatted like: <unix_timestamp> <hostname> <hostname> Example: 1366815793 quark garak 1366815795 brunt quark 1366815811 lilac garak Each line represents connection from a host (left) to another host (right) at a given time. The lines are *roughly* sorted by timestamp. They might be out of order by a maximum of 5 minutes. *Your task:* Implement a tool that parses such log files. The tool should both parse previously written log files by filename, or read from the standard input a stream of lines in the previously defined format. The script will output: * a list of hostnames that connected to a given (configurable) host during the last hour * a list of hostnames that received connections from a given (configurable) host during the last hour * the hostname that generated most connections in the last hour * the hostname that received most connections in the last hour Expect the number of loglines and hostnames to be very high. Consider implementing a CPU and memory-efficient solution. On-site: - How would you implement T9 ? - You have a e-shop and you must implement the shop card. The e-shop has some bonuses that if you buy for example 3 same stuff one is free, if you buy this stuff there is 25% free etc.