LinuxQuestions.org - Shell script to store high water mark for uptime.

- Programming (https://www.linuxquestions.org/questions/programming-9/)

- - Shell script to store high water mark for uptime. (https://www.linuxquestions.org/questions/programming-9/shell-script-to-store-high-water-mark-for-uptime-676544/)

Shell script to store high water mark for uptime.

I wrote a small shell script that reads /proc/uptime and prints it in human viewable time. I'd like to somehow store the longest uptime. Here's what I've got so far.

Code:

#!/bin/csh

#

# 

# Author D. McIntyre  07/31/08  mac@yestermac.com

#

# Modified 10/15/08 to work with Slackware linux. Original worked with netBSD.



# To avoid a "Badly Formed Number" error the following line converts uptime from float to integer



set systime = `cat /proc/uptime | awk 'BEGIN { FS = "." } ; { print $1 } '  `



# divide systime by 60, the remainder is minutes; divide systime by 3600 the remainder is hours

# divide systime by 86400 equals days.



@ minutes = ($systime / 60)  % 60

@ hours = ($systime / 3600) % 24  

@ days =  $systime / 86400 



#  If day, hour or minute equals 1 then print singular version.  



if ($days == 1  ) then

  set day="day, "

    else

    set day="days, "

endif



if ($hours == 1 ) then

  set hour="hour and "

    else

    set hour="hours and "

endif



if ($minutes == 1 ) then

  set minute=" minute. "

    else

    set minute=" minutes. "

endif



echo  "This system has been up" $days $day $hours $hour $minutes $minute

I seem to be suffering from a senior moment and can't figure out how to save the longest uptime and have it updated when the current uptime is greater then the longest uptime. I am going to write this to the MOTD when someone logs into the system.

I think the simplest method would be to store the current max value in a text file (I'd probably select a dotfile but that is a personal choice) and compare it to the current value, overwriting the data in the text file if the current value is greater.

If you're asking about the details of the comparison, I'd suggest using eval or bash/csh builtin arithmetic comparisons. I'm sure myself or others could give you examples of that sort of thing if that's what you're after.

hth

Wouldn't just converting both values to epoch and subtracting be easier?

This just stores the higher numeric value of what's already stored or the first field of proc uptime. If nothing is there when the script is ran it will store the contents of field 1 of /proc/uptime. Note the file location /home/ed/.myuptime will need to be altered for your system.

I attempt no manipulation of the values simply storing the larger of the two. Any manipulation can be done by a display routine - something like your script.

Code:

#!/bin/bash

FLE="/home/ed/.myuptime"

if [[ -f "$FLE" ]]

then        

        OLD_UP=$(cat "$FLE")

else        

        OLD_UP=0

fi

NEW_UP=$(awk '{ print $1 }' /proc/uptime)

if [[  $(echo "$NEW_UP > $OLD_UP" | bc) -eq 0 ]]

then

        NEW_UP=$OLD_UP

fi

echo "$NEW_UP" > "$FLE"

Hope this is useful to you.

Thanks! I was able to finish my script using your suggestions. I had to do a little reading to understand what was happening with the bash script and apply that logic it to csh.

Code:

if ( -e .maxuptime ) then

 if ( `cat .maxuptime ` < $systime ) then

    echo $systime > .maxuptime

 endif

  else

    echo $systime > .maxuptime

endif

set systime = `cat .maxuptime `

Quote:

I had to do a little reading to understand what was happening with the bash script and apply that logic it to csh.

Oops - I always assume everyone will be using bash - sorry ! Hope it wasn't too much work translating !