I'm using the following script to deploy new image to a kubernetes cluster:

#!/usr/bin/env bash

set -euo pipefail
set -x

rootPath=".."
gcpProjectId="example123"
gcpRepoHostname="eu.gcr.io"
imageName="wordpress"
containerName="wordpress"
tagNameSuffix=$(date +%s)
tagNameSuffix="deploy-${tagNameSuffix}"
tagName="$gcpRepoHostname"/"$gcpProjectId"/"$imageName":$tagNameSuffix

echo "$GCLOUD_API_KEYFILE" | base64 --decode --ignore-garbage > ./gcloud-api-key.json
gcloud auth activate-service-account --key-file gcloud-api-key.json
gcloud config set project "$gcpProjectId"

sudo docker build "$rootPath"/wordpress/ -t "$tagName"
sudo gcloud docker -- push "$tagName"
kubectl set image deployment/wordpress "$containerName"="$tagName"

Strangely this worked when I modified the wordpress/Dockerfile to be:

FROM rlesouef/alpine-nginx

If I change it to:

FROM wordpress:php7.1-fpm-alpine

COPY custom-entrypoint.sh /usr/local/bin/

I get a non responsive service - fails to connect to the external ip.

Heres is some debug information relating to the last change (wordpress image):

kubectl get pods shows the following:

NAME                        READY     STATUS    RESTARTS   AGE
wordpress-702001878-xb0b1   2/2       Running   0          8m

If I check the logs via kubectl logs wordpress-702001878-xb0b1 wordpress:

[07-Aug-2017 15:13:14] NOTICE: fpm is running, pid 1
[07-Aug-2017 15:13:14] NOTICE: ready to handle connections

How can I diagnose this issue? I suppose I need to determine the exact docker image being currently run on the cluster / pods.