Monitoring and Debugging in the Linux Environment

Monitoring memory

• top, ps and more
• How the OS allocates RAM
• Being not lazy in RAM allocation
• Swap (adding, removing, monitoring)
• To swap or not to swap

Monitoring applications

• ps and it's many options
• how does /proc work
• the /proc folder per app
• What are zombies? How do you identify them? How do you handle them?
• Controlling affinity of apps
• Limiting apps resources
• strace (standalone and attaching)
• top and it's many options
• What is a software watchdog?
• Writing a simple Linux watchdog
• Using inittab for simple watchdogging
• How does the Linux scheduler work? nice(1), chrt(1)

Monitoring disk

• The mount paradigm
• using du and df correctly
• Difference between disk utilization and logical size of files
• Seeing all open files of applications
• Controlling disk scheduling priority and class (ioprio).
• iotop(1) and iostat(1)
• sar(1) and other monitoring tools
• The different file systems Linux supports
• Implications of file system selection on application performance.
• Controlling file system features.

Monitoring network

• netstat and it's many options
• Monitoring and controlling the OS network stack (/proc).
• ifconfig and it's many options
• Monitoring the network interface (ethtool).
• ip(1) and it's many options
• Load balancing and failing over ethernet interfaces
• Bridging network interfaces.
• vlans and tunnels.
• lo and what it's for?
• Name resolution and it's problems.
• tap devices.

Monitoring the kernel

• Seeing kernel threads in ps
• Examining interrupts
• Controlling affinity of interrupt handles
• Interrupt load balancing
• Controlling affinity of kernel threads
• Modules: loading, unloading, monitoring

The boot process

• The BIOS
• The bootloader
• Configuring grub
• Using other bootloaders (lilo, syslinux, uboot)
• Advantages and disadvantages of certain boot loaders
• The init process
• SysV init
• Upstart
• inittab and how it works