It is a playbook bug, the PLAY [Update vm.max_map_count for ES 5.x] was skipped.
 
The elasticsearch pod became ready after I configured the  vm.max_map_count manually.

PLAY [Update vm.max_map_count for ES 5.x] **************************************
META: ran handlers

TASK [Checking vm max_map_count value] *****************************************
task path: /usr/share/ansible/openshift-ansible/playbooks/openshift-logging/private/config.yml:85
Friday 27 July 2018  08:53:10 +0000 (0:00:00.053)       0:00:37.868 *********** 
skipping: [qe-qitang-311-gcenode-1] => {"changed": false, "skip_reason": "Conditional result was False"}

TASK [Updating vm.max_map_count value] *****************************************
task path: /usr/share/ansible/openshift-ansible/playbooks/openshift-logging/private/config.yml:90
Friday 27 July 2018  08:53:10 +0000 (0:00:00.023)       0:00:37.892 *********** 
skipping: [qe-qitang-311-gcenode-1] => {"changed": false, "skip_reason": "Conditional result was False"}
META: ran handlers
META: ran handlers

PLAY [Remove created 99-elasticsearch sysctl] **********************************
META: ran handlers

*** This bug has been marked as a duplicate of bug 1609138 ***