The application team complains that response times occasionally spike on an 8-vCPU CentOS virtual machine that hosts a PostgreSQL database. You collect the following data during a slowdown:
The load average is far above the VM's eight cores, yet only about 6 % of CPU time is spent on user and system work. More than 80 % of CPU time is in iowait, meaning the processors are idle while the kernel waits for storage operations to finish. The accompanying iostat output shows 99 % device utilization and a very high average wait time (~242 ms) on vda, confirming that the single virtual disk is saturated.
Because Linux counts processes stuck in uninterruptible I/O sleep (D-state) in the load average, many blocked database threads inflate the load number even though CPU is mostly idle. Resolving the issue requires relieving disk I/O pressure (for example, moving the data files to faster or additional storage), not tuning CPU, memory, or network resources.
The other options do not match the evidence: CPU saturation would show high %usr/%sys, an interrupt storm would raise %sys and context-switch rates, and memory thrashing would appear as high swap-in/out rather than dominant %iowait and disk latency.
Ask Bash
Bash is our AI bot, trained to help you pass your exam. AI Generated Content may display inaccurate information, always double-check anything important.
What does 'iowait' mean in Linux performance metrics?
Open an interactive chat with Bash
What does the load average represent, and why is it high even if the CPU is idle?
Open an interactive chat with Bash
How can disk I/O bottlenecks like this be resolved?
Open an interactive chat with Bash
CompTIA Linux+ XK0-006 (V8)
Troubleshooting
Your Score:
Report Issue
Bash, the Crucial Exams Chat Bot
AI Bot
Loading...
Loading...
Loading...
IT & Cybersecurity Package Join Premium for Full Access