This will probably work for about half of a year's input assuming that all lines are in exactly the same format and that the CST is referring to US Central Standard timezone. It won't work when dayligth savings time is in effect. And, of course, it will only work for dates in 2015.

The following modifications should work year around for the sample input provided in 2015:

Code:

string1='a.sh start time is Fri Jan  9 17:17:33 CST 2015';
string1=${string1%* C?T 2015}
string1=${string1##* }

string2='a.sh start time is Fri Jun 12 17:17:33 CDT 2015';
string2=${string2%* C[DS]T 2015}
string2=${string2##* }

clear && echo $string1 && echo "" && echo $string2

The ? will match any character; the [DS] will match an uppercase D or an uppercase S. I will leave the modifications needed to recognize additional years as an exercise for the reader. If some of the dates being processed come from other timezones, I would suggest using a different approach.

You seem to be mixing the awk match() and split() functions. But, I get a syntax error when I try this code. The match() function takes two arguments (string to be matched and an ERE to match) and sets various awk variables to indicate whether or not a match was found and, if there was a match, the starting position of the match and the length of the match. The awk script I suggested in post #8 in this thread uses match() this way.

The awk split() function creates an array of values corresponding to values separated by field separators specify by an ERE. With the ERE you use, the desired date would not appear in the array at all since the ERE selects the date as the field separator.