2.5 KiB
Level 1 Questions
How would you prove the code is correct? To prove the code is correct I suppose I could set up a test environment where I could account for all connections and run tests like ssh connections, dhclient or dig (dns) connections, nmap port scans, and then ensure the connections I expect from the tests appear in the program's output.
How would you make this solution better? I'd improve the solution by making the functions more focused on specific tasks, improving how I present the interface of each function to each other to more specifically eliminate type or value errors. I'd also allot more time to complete the task since I wasn't able to find enough before the deadline. Is it possible for this program to miss a connection? Yes, if the connection is brief enough so as not to last more than 10 seconds and thereby be missed by the most recent read of the /proc/net/tcp file If you weren't following these requirements, how would you solve the problem of logging every new connection? I would likely find a solution that leverages inotify to detect any change to the file and then read it, rather than polling it every 10 seconds.
Level 2 Questions
Why did you choose 'x' to write the build automation? I would simply choose bash, as it's what I'm most familiar with and it can reliably use the go build tools as well as invoke testing systems like kvm/qemu, docker/podman and the like. Is there anything else you would test if you had more time? If I had more time I'd spend it fleshing out my testing as outlined above and doing a more thorough job of structuring my data in the way that I've modeled it in my head - the choice of go definitely consigned me to reading more than writing with the little time I had. Should have gone with bash I think. What is the most important tool, script, or technique you have for solving problems in production? Explain why this tool/script/technique is the most important.
Level 3 Questions
If you had to deploy this program to hundreds of servers, what would be your preferred method and why?
I would use ansible and use git to clone, build, and install the script.
This method is my preferred one because it's simple to
troubleshoot in the event that I run into issues and
ansible doesn't require a central orchestrator node. It's
also the Infrastructure-as-code tool I'm most familiar with.
What is the hardest technical problem or outage you've had to solve in your career? Explain what made it so difficult?