Is there a program to diagnose and summarize why my mchine is slow? [closed]
Is there a program that will tell me why my server is slow? What is being over utilized? CPU / Memory / Disk / Network ... and what processes are driving that utilization? I want a program that can check and explain it to me. (I don't want to figure it out myself.) Maybe it could even suggest server tuning changes.
People are using top
or variations of it: htop
, atop
, glances
, bottom.
You can add to that: vmstat
, iostat
, mpstat
, dstat
Then you have proffillers (tracing and sampling), debuggers, tracers (kernel and userspace).
Core dump analyzers, CPU performance registers.
Network captures, USB and serial/UART debuggers/dumpers.
Add to this virtualization and containers, different architectures (CPU, NUMA...)
Don't expect one tool that will find all problems and solution and that tool to be maintained and evolve with technology. I expect in the future to be AI enhanced tools, that would help for 80% of issues, but new issues requires deep analysis that only a human would solve.
eBPF could fit the bill for a lot of issues. Seel also Brendan's page.