An application running on Linux used to process 1000 requests per second. Now it is only processing 2 requests/second.
How will you troubleshoot this issue?
This is a Linux troubleshooting question about a sudden drop in application throughput, not a coding algorithm problem. A strong answer should start by narrowing the scope—whether the issue affects one host, one process, or one endpoint—and then check CPU, memory, disk I/O, network, logs, and process state. Good candidates would mention tools such as top, ps, vmstat, iostat, ss/netstat, dmesg, logs, and, when needed, strace or perf to determine whether the bottleneck is CPU saturation, lock contention, I/O blocking, exhausted connections, or a slow downstream dependency.
END
