Filter container logs and output to file

filter-logs.sh

#!/bin/bash

label=$1
namespace=$2
filter=$3
fileout=$4
container=$5

if [ $# -eq 0 ]
then
        echo "no arguments supplied, [label] [namespace] [filter] [filename.output] [container]"
        echo "example: app=updatedb spuprod ds2 test.txt"
        exit 1
fi

while IFS= read -r pod
do
        echo processing logs of $pod
        kubectl logs -n spuprod -c $container $pod|grep -B 5 fail|grep sellerId>> $fileout
done < <(kubectl get pods --no-headers=true -o custom-columns=:metadata.name -n $namespace -l $label |grep $filter)

Can be further post-processed (e.g. for grouping) with
awk '{a[$8]+=1;}END{for(i in a)print i", "a[i];}' test.txt