Tutorial 25-28

Q1 Advanced Boot Interventions & Resource Limiting

You are troubleshooting a kernel module issue that only manifests under low-memory conditions. You need to temporarily restrict the system's available memory for the next boot sequence only.

Task

Reboot the system and interrupt the boot loader menu.
Modify the kernel parameters at boot time to restrict the total usable physical memory to exactly 1 GiB.
Ensure the system boots successfully into the default target with this memory restriction.
Verify that the change is active, but ensure this restriction does not persist across subsequent reboots.

Solution

Access Linux boot menu
- sudo systemctl reboot --boot-loader-menu=60
Restrict the memory to 1 GiB
- Press e on the Linux boot menu.
- Find the line start with linux then move the cursor to the end of line
- Press Space then add/edit this line : mem=1G
- Press Ctrl+X to save and reboot
Verify
- You should explicitly see mem=1G at the end of the printed command line string.
- cat /proc/cmdline
The mem=1G will not persist in next reboot

Q2 Process Management and System Load Analysis

An administrator left a simulation process running, causing potential CPU bottlenecks. You are tasked with analyzing the system capacity, generating a temporary synthetic load for testing, and then clearing the background tasks.

Task

Determine the total number of logical CPU cores available on the system and note the current 1-minute load average.
Launch a command using dd that reads from /dev/zero and writes to /dev/null as a background job to simulate CPU load.
Launch a sleep 3000 command in the foreground, pause it, and move it to run in the background.
View the active job list to ensure both tasks are running in the background.
Terminate the dd process using a method that kills all instances of it by name, leaving the sleep job intact.

Solution

Find out the total CPU units available on your system.
- Look for the line labelled CPU(s):
- If CPU(s): 2, your system can handle 2 fully active processes at any given instant.
- lscpu or lscpu | grep "CPU(s):"
Current 1-minute load average
- uptime
Simulate CPU load
- dd if=/dev/zero of=/dev/null &
Run sleep command
- sleep 3000
- Press Ctrl+Z to pause the job
- List all jobs: jobs
- Move job to background: bg %[Job Index]. Example: bg %2
- Verify: jobs
Terminate dd process
- killall dd
- Verify: jobs

Q3 Real-Time Performance Diagnostics

The system is experiencing intermittent performance degradation. You need to use real-time monitoring tools to diagnose resource allocation and specific CPU metrics.

Task

Launch the interactive system activity dashboard.
Configure the view to display individual CPU core metrics rather than a combined average.
Identify which specific CPU metric field indicates the percentage of time the CPU is idling while waiting for disk I/O operations to complete.
Determine the amount of currently available physical memory. Identify which memory metric indicates that the system is dangerously low on physical RAM and has begun swapping pages to the hard drive.

Solution

Launch dashboard
- top
Press the key 1 see how many CPU cores you have.
wa (iowait): Time the CPU is sitting idle waiting for the hard drive or network to respond. If wa is high (above 5%), your storage drive is bottlenecking your entire system.
Look for the line labelled MiB Mem --> free (available memory).
- If this drops close to zero, your system will start using the MiB Swap space (simulated memory on the hard drive), which will severely degrade performance.
- So try to avoid using MiB Swap space

Page updated

Google Sites

Report abuse