closing, was mistake on my part, one es node service didnt get restarted correctly post upgrade (somehow!!??) resulting in a cluster with 2 es nodes on 7.10.2 and one still on 6.8. restarting the es service via
sudo service elasticsearch restart
resolved the issue
I used
curl -X GET "graylog01.test.internal:9200/"
curl -X GET "graylog02.test.internal:9200/"
curl -X GET "graylog03.test.internal:9200/"
to find the running es version of each host