Skip to content

tmux plugin, Displays packet loss if at or above threshold

License

Notifications You must be signed in to change notification settings

jaclu/tmux-packet-loss

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Tmux-Packet-Loss

Tmux-Packet-Loss is a plugin for Tmux that displays the percentage of packet loss on your connection. It calculates the loss level as a weighted average by default, giving more emphasis to recent checks.

Recent changes

  • New opiton @packet-loss-run_disconnected
  • Losses are displayed, but no stats are saved for the first 30 seconds. This avoids getting initial errors before the network is re-established saved into the history during a laptop resume.
  • Fixed boolean parameter handling to allow for yes/no or true/false options.
  • Renamed variables and defaults to match the Tmux option names.
  • Refactored code into more task-isolated modules.

Screenshots

Partial status bar configuration: #{battery_smart} #{packet_loss}%a %h-%d %H:%M

Plugin output takes no space when under @packet-loss-level_disp level.

Loss levels

Display With hist avg Status
no_loss  under threshold
lvl_low  lvl_low_h low level losses
lvl_alert lvl_alert_h alert level losses
lvl_crit lvl_crit_h critical level losses

Trends

If @packet-loss-display_trend is yes, change since the previous check is indicated with a prefix character

Display Status
incr  Increasing
stable Stable
decr  Decreasing

Operation

This plugin runs a background process using repeated runs of ping to determine % package loss. The loss level is calculated as a weighted average of the stored data points by default, making the latest checks stand out.

Termination of background monitor

Unless @packet-loss-run_disconnected is yes, the background monitor terminates if no clients are connected. It will resume as soon as any client re-connects.

The background process monitors the tmux server pid, and terminates if the tmux server exits.

ping issues

If the monitor fails to calculate loss, packet loss above 100% is reported. So far I have created one special case ping parser, for iSH running Debian 10.

Result Explanation
101 Failed to find % loss in ping output. Temporary issue. Some pings don't report loss % if there is no connection to the host.
102 loss reported was < 0 or > 100, odd but hopefully temporary
103 ping reported error, on some distros this happens if there is no netork connection
201 Could not parse the output. This condition is unlikely to self-correct. If you file the output of ping -c 5 8.8.4.4 as an Issue and also mention what Operating System this is and any other factors you think are relevant, I will try to fix it by including parsing of that output format.

Dependencies

Ensure you have the following dependencies installed:

  • tmux 1.9
  • sqlite3
  • bash

Verified Environments

Tmux-Packet-Loss has been tested and verified to work in the following environments:

  • Linux
  • MacOS
  • iSH
  • Windows Subsystem for Linux (WSL)

Installation

Installation with Tmux Plugin Manager (tpm) (recommended)

Add the plugin to the list of TPM plugins in .tmux.conf:

set -g @plugin 'jaclu/tmux-packet-loss'

Hit prefix + I to fetch the plugin and source it. That's it!

Manual Installation

Clone the repository:

git clone https://github.com/jaclu/tmux-packet-loss.git ~/clone/path

Add this line to the bottom of .tmux.conf:

run-shell ~/clone/path/packet-loss.tmux

Reload the Tmux environment with $ tmux source-file ~/.tmux.conf - that's it!

Supported Format Strings

Code Action
#{packet_loss} Displays packet loss if at or above @packet-loss-level_disp
Otherwise nothing

Configuration Variables

Variable Default Purpose
@packet-loss-ping_host 8.8.8.8 The host to ping. Choosing a well-connected & replicated host like 8.8.8.8 or 1.1.1.1 gives a good idea of your general link quality.
@packet-loss-ping_count 6 Number of pings per statistics update.
@packet-loss-history_size 6 Number of results to keep when displaying loss statistics.
Keeping this value low is recommended since it's more useful to see current status over long-term averages.
For a historical overview, use @packet-loss-hist_avg_display.
@packet-loss-weighted_average yes yes Use weighted average focusing on the latest data points.
no Average over all data points.
@packet-loss-display_trend no yes Display trend with + prefix for higher levels and - prefix for lower levels.
no Do not indicate change since the previous loss level.
@packet-loss-hist_avg_display no yes Show historical average when displaying current losses.
no Do not show historical average.
@packet-loss-run_disconnected no yes monitor only exits when tmux is shut down.
no monitor exits if no clients are connected and restarts when a client re-connects.
@packet-loss-level_disp 1 Display loss if at or higher than this level.
@packet-loss-level_alert 17 Color loss with color_alert if at or above this level.
Suggestion: set it one higher than the percentage representing one loss in one update to avoid single packet loss triggering an alert initially.
@packet-loss-level_crit 40 Color loss with color_crit if at or above this level.
@packet-loss-hist_avg_minutes 30 Minutes to keep the historical average.
@packet-loss-hist_separator '~' Separator for current/historical losses.
@packet-loss-color_alert colour226 Use this color if the loss is at or above @packet-loss-level_alert.
@packet-loss-color_crit colour196 Use this color if the loss is at or above @packet-loss-level_crit.
@packet-loss-color_bg black Background color when alert/crit colors are used in the display.
@packet-loss-prefix '|' Prefix for status when displayed.
@packet-loss-suffix '|' Suffix for status when displayed.
@packet-loss-log_file If defined this file will be used for logging.

My config

set -g @packet-loss-display_trend     yes
set -g @packet-loss-hist_avg_display  yes

#
# In combination with weighted_average, ping_count and history_size,
# This makes a single ping loss disappear from being displayed in 15s
#
set -g @packet-loss-level_disp   5

set -g @packet-loss-color_alert  colour21
set -g @packet-loss-color_bg     colour226

Balancing reporting

To obtain a clearer picture of the current situation, consider adjusting the ping count. A higher ping count results in more nuanced data per check. For instance, if only 2 packets are checked per round, the results may only be 0%, 50%, or 100%, lacking granularity. Increasing the ping count enhances the accuracy of each check.

However, be cautious not to exceed a certain limit, as a higher ping count prolongs the time taken for each test. This delay may render the reported data irrelevant to the current link status, particularly if your focus is on real-time monitoring.

For longer-term averages, it is better to use @packet-loss-hist_avg_display

Additionally, it's advisable to review and potentially adjust the status-interval setting to align with your reporting needs. Ensuring that the update rate for this plugin in the status bar remains relevant enhances the effectiveness of your monitoring system.

Given that ping is instantaneous, consider setting the status-interval to one lower than @packet-loss-ping_count. This adjustment synchronizes the sampling and reporting processes more effectively, providing timely and accurate updates.

Nerdy stuff

If the data folder, where the database and status files are kept disappears, the database and statuses will be recreated and a new monitor process will be started. So this is a simple and quick way to clear historical data, without having to bother with SQL!

If @packet-loss-weighted_average is set to yes (the default) losses are displayed as the largest of:

  • last value
  • avg of last 2
  • avg of last 3
  • avg of last 4
  • avg of last 5
  • avg of last 6
  • avg of last 7
  • avg of all

If set to no, the average of all samples is always displayed.

Suggested Alert Levels

Depending on the ping count, it is suggested to set an alert, so that a single lost packet won't show up as an alert.

pings one higher than
a single loss %
history size
for aprox 30s
10 11 3 (27) 4 (36)
9 12 4 (32)
8 13 5 (35)
7 15 5 (30)
6 (default) 17 6 (30)
5 21 8 (32)
3 34 15 (30)

Database

There are three tables

table Description
t_loss Contains the current loss statuses
t_1_min Keeps all samples from the last minute, to feed one-minute averages to the t_stats table
t_stats Keeps one-minute averages for the last @packet-loss-hist_avg_minutes minutes

Each table contains two fields, time_stamp, and value. The time_stamp field is used to purge old records.

Simulating losses

To examine the plugin displaying losses in the status bar, there is an included test script. Run it without params to get a help summary

./scripts/test_data.sh

The monitor will be automatically restarted two minutes after the last run of test_data.

Contributing

Contributions are welcome, and they're appreciated. Every little bit helps, and credit is always given.

The best way to send feedback is to file an issue at tmux-packet-loss/issues

License

MIT

About

tmux plugin, Displays packet loss if at or above threshold

Topics

Resources

License

Stars

Watchers

Forks

Languages