INDIAN INSTITUTE OF INFORMATION TECHNOLOGY, NAGPUR

Each PostgreSQL backend connection is an independent operating system process allocating ~10MB of RAM and competing for CPU scheduler time. Using connection poolers like PgBouncer or Neon connection pooling caps concurrent connections at ~50 while queuing burst requests. Better throughput, zero server crashes.

When a process references memory, the CPU's MMU uses page tables to translate the virtual address to a physical RAM frame. If the present bit is 0, the CPU triggers a page fault interrupt. The OS pauses the thread, fetches the 4KB block from swap disk into RAM, updates the page table, and resumes the instruction. That is why excessive swapping grinds your computer to a halt.

I have 47 open tabs on Chrome right now: 3 are for my project, 12 are LeetCode problems I looked at and closed, and the rest are Reddit threads from 2018 solving obscure bugs.

My favorite campus moment is when everyone in the examination hall looks up at the ceiling at the exact same second because Question 4 is completely unhinged.

I know which lecture hall chairs have the squeaky springs and which ones have working power outlets. That knowledge is my true university degree.

I once spent 6 hours fixing an SQL query that was returning duplicate rows, only to realize I had written `UNION ALL` instead of `UNION`. I didn't tell my project teammates; I told them it was a 'complex database concurrency conflict'.

Accidentally opened a YouTube video with full volume in the pin-drop silent reading room. It was a Bollywood song remix. I closed my laptop with such force I think I cracked the trackpad.

In India, network connectivity drops every time you enter a basement elevator, train tunnel, or hostel bathroom. Building apps that immediately crash with 'No Internet' is terrible UX. Store state locally with SQLite / WatermelonDB and sync background mutations with exponential backoff when connectivity restores.

Relying on remote cloud databases like Supabase or Neon over spotty hostel Wi-Fi during local development introduces 200ms latency to every reload. Run Postgres locally via Docker or brew: instant queries, zero network dependency, and full offline freedom.

We ran a campus ticket booking system for our college fest. Everything was fine in testing with 50 users. As soon as 2,000 students refreshed at 10 AM, the server froze. PostgreSQL was doing sequential table scans on `orders.user_id` for every check. One simple `CREATE INDEX` reduced query latency from 12,000ms to 4ms. Always index your foreign keys!

For students wanting to build apps quickly for hackathons and startup MVPs, Flutter's cross-platform speed is unbeatable. But if your goal is Big Tech mobile engineering roles, Kotlin and Jetpack Compose are what recruiters search for. Don't let influencers confuse you — pick based on your immediate objective.