ttyplot1
Realtime terminal plotting utility
Usage
ttyplot [options]
ttyplot -v
ttyplot -h
Overview
ttyplot takes data from standard input, most commonly some tool like ping(1), snmpget(1), netstat(8), ifconfig(8), sar(1), vmstat(8), etc., and plots in text mode on a terminal in real time.
Supports rate calculation for counters and up to two graphs on a single display using reverse video for second line.
The following options are supported:
-
-2 -
Read two values and draw two plots, the second in reverse video.
-
-r -
Calculate counter rate and divide by measured sample interval.
-
-c plotchar -
Use plotchar for the plot line, e.g. ‘@ # % .’ etc.
-
-e errcharmax -
Use errcharmax for plot error line when value exceeds hardmax. Default: ‘e’.
-
-E errcharmin -
Use errcharmin for plot error symbol, displayed when plot value is less than hardmin. Default: ‘v’.
-
-s softmax -
Use softmax as the initial maximum value but allow it to grow with input.
-
-m hardmax -
Use hardmax as a hard value limit after which an error line will be drawn( see
-e). Should be greater than hardmin, if set. -
-M hardmin -
Use hardmin as a definite minimum limit of the plot range. If a plot value is less than this, error symbol will be drawn( see
-E). -
-t title -
Use title as the plot title.
-
-u unit -
Label the vertical axis unit.
-
-v -
Print the current version and exit.
-
-h -
Print this help message and exit.
Key bindings
When reading data from a pipe, ttyplot accepts the following commands typed at the terminal:
q-
Quit.
r-
Toggle "rate mode" on and off.
Examples
CPU usage from vmstat(8) using awk(1) to pick the right column:
vmstat -n 1 \
| gawk '{ print 100-int($(NF-2)); fflush(); }' \
| ttyplot
CPU usage from sar(1) with title and fixed scale to 100%:
sar 1 \
| gawk '{ print 100-int($NF); fflush(); }' \
| ttyplot -s 100 -t "cpu usage" -u "%"
Memory usage from sar(1), using perl(1), to pick the right column:
sar -r 1 \
| perl -lane 'BEGIN{$|=1} print "@F[5]"' \
| ttyplot -s 100 -t "memory used %" -u "%"
Number of processes in running and io blocked state:
vmstat -n 1 \
| perl -lane 'BEGIN{$|=1} print "@F[0,1]"' \
| ttyplot -2 -t "procs in R and D state"
Load average via uptime(1) and awk(1):
{ while true; do
uptime | gawk '{ gsub(/,/, ""); print $(NF-2) }'
sleep 1
done } | ttyplot -t "load average" -s load
Ping plot with sed(1):
ping 8.8.8.8 \ | sed -u 's/^.*time=//g; s/ ms//g' \ | ttyplot -t "ping to 8.8.8.8" -u ms
WiFi signal level in -dBM (higher is worse) using iwconfig(8):
{ while true; do
iwconfig 2>/dev/null \
| grep "Signal level" \
| sed -u 's/^.*Signal level=-//g; s/dBm//g'
sleep 1
done } | ttyplot -t "wifi signal" -u "-dBm" -s 90
CPU temperature from proc;
{ while true; do
awk '{ printf("%.1f0, $1/1000) }' \
/sys/class/thermal/thermal_zone0/temp
sleep 1
done } | ttyplot -t "cpu temp" -u C
Fan speed from sensors(1) using grep(1), tr(1) and cut(1):
{ while true; do
sensors | grep fan1: | tr -s " " | cut -d" " -f2
sleep 1
done } | ttyplot -t "fan speed" -u RPM
Bitcoin price chart using curl(1) and jq(1):
{ while true; do
curl -sL https://api.coindesk.com/v1/bpi/currentprice.json \
| jq .bpi.USD.rate_float
sleep 600
done } | ttyplot -t "bitcoin price" -u usd
Stock quote chart:
{ while true; do
curl -sL https://api.iextrading.com/1.0/stock/googl/price
echo
sleep 600
done } | ttyplot -t "google stock price" -u usd
Prometheus load average via node_exporter:
{ while true; do
curl -s http://10.4.7.180:9100/metrics \
| grep "^node_load1 " \
| cut -d" " -f2; sleep 1
done } | ttyplot
Network/disk throughput examples
ttyplot supports two-line plots for in/out or read/write.
Local network throughput for all interfaces combined from sar(1):
sar -n DEV 1 | gawk '{
if($6 ~ /rxkB/) {
print iin/1000;
print out/1000;
iin=0;
out=0;
fflush();
}
iin=iin+$6;
out=out+$7;
}' | ttyplot -2 -u "MB/s"
SNMP network throughput for an interface using ‘ttg’:
ttg -i 10 -u Mb 10.23.73.254 public 9 \
| gawk '{ print $5,$8; fflush(); }' \
| ttyplot -2 -u Mb/s
SNMP network throughput for an interface using snmpdelta(1):
snmpdelta -v 2c -c public -Cp 10 \
10.23.73.254 1.3.6.1.2.1.2.2.1.{10,16}.9 \
| gawk '{ print $NF/1000/1000/10; fflush(); }' \
| ttyplot -2 -t "interface 9 throughput" -u Mb/s
Disk throughput from iostat(1):
iostat -xmy 1 nvme0n1 \ | stdbuf -o0 tr -s " " \ | stdbuf -o0 cut -d " " -f 4,5 \ | ttyplot -2 -t "nvme0n1 throughput" -u MB/s
Rate calculator for counters
ttyplot also supports counter style metrics, calculating a rate by measuring time difference between samples.
SNMP network throughput for an interface using snmpget(1):
{ while true; do
snmpget -v 2c -c public \
10.23.73.254 1.3.6.1.2.1.2.2.1.{10,16}.9 \
| awk '{ print $NF/1000/1000; }'
sleep 10
done } | ttyplot -2 -r -u "MB/s"
Local interface throughput using ip(8) and jq(1):
{ while true; do
ip -s -j link show enp0s31f6 \
| jq '.[].stats64.rx.bytes/1024/1024, \
.[].stats64.tx.bytes/1024/1024'
sleep 1
done } | ttyplot -r -2 -u "MB/s"
Prometheus node exporter disk throughput for /dev/sda:
{ while true; do
curl -s http://10.11.0.173:9100/metrics \
| awk '/^node_disk_.+_bytes_total{device="sda"}/ {
printf("%f0, $2/1024/1024);
}'
sleep 1
done } | ttyplot -r -2 -u MB/s -t "10.11.0.173 sda writes"
History
- Debian 13 (Trixie): the text shown here
- Debian 12 (Bookworm)
-h since Debian 13 (Trixie);
-v since Debian 13 (Trixie).