macOS-Monitoring

Simple monitoring using native macOS commands. These should work on any macOS by downloading, making executable, and running.

Usage (CSV data):

./metrics.sh > metrics.log

Termination:
./term.sh

Or you can kill the specific processes:

./met-term.sh
./net-stat-term.sh

lightweightmetrics.awk is needed for normal script operation.

Both metrics.awk formats the general metrics output for a curl.

cpu-loop.sh is the most often used for "right-sizing" a VM, and is sperated out so users can easily change the frequency of CPU monitoring by modifying the sleep command. (cpu.awk formats that data for a curl if desired.)

Data Fields:

• FirstField: Date
• NetworkIn: From netstat - Bytes inbound
• NetworkOut: From netstat - Bytes transmitted out
• CPU-USED: From ps - Percent by core, total percentage is 100% x Number of Cores
• FreeMem: From top - Free memory with "M" or "G" on the end
• RAMPlot: FreeMem converted to Megs of RAM, more useful for plotting
• CPUIdle: From top - Percent of total CPU cores (min 0%, max 100%)
• DiskRead: From top - Bytes attempted to be read. This is a sumed value for VMs - always increasing.
• DiskReadCap: From top - The system reported read cap (usually in G)
• DiskWrite: From top - Bytes attempted to be written. This is a sumed value for VMs - always increasing.
• DiskWriteCap: From top - The system reported write cap (usually in G)
• RWCURRENT: From {time dd} - The actual written bytes divided by time to write (bytes/s)
• RWTime: From {time dd} - The time to write the specified data

Troubleshooting:

The command time dd command can create Xcode to hang if a double write occurs. Should a hung Xcode be noticed, it is recommended this line be commented out, or a seperate partition be created for the metrics.sh script.