Bug 1571478

Created attachment 1426308 [details]
kibana screenshot

Description of problem:

In OpenShift 3.10 the master-api, master-controllers and etcd components run as static pods in the kube-system project.

After installing logging 3.10.0-0.27.0 and implementing the workaround to get the fluentd daemonset running (see https://bugzilla.redhat.com/show_bug.cgi?id=1569106), logs are indexed for the pods in kube-system but they are not available in kibana.

# oc exec -n openshift-logging -c elasticsearch $POD -- curl --connect-timeout 2 -s -k --cert /etc/elasticsearch/secret/admin-cert --key /etc/elasticsearch/secret/admin-key https://logging-es:9200/_cat/indices?v                            
health status index                                                               pri rep docs.count docs.deleted store.size pri.store.size                                                                        
green  open   .operations.2018.04.24                                                1   0     177631            0      155mb          155mb                                                                        
green  open   project.svt0.c500fdf2-47ee-11e8-ae99-022e0d5a511e.2018.04.24          1   0       1956            0      582kb          582kb                                                                        
green  open   project.svt4.c5a84f2c-47ee-11e8-ae99-022e0d5a511e.2018.04.24          1   0       1957            0    589.1kb        589.1kb                                                                        green  open   .kibana                                                               1   0          1            0      3.1kb          3.1kb                                                                        green  open   project.svt1.c5283390-47ee-11e8-ae99-022e0d5a511e.2018.04.24          1   0       1957            0    566.2kb        566.2kb                                                                        green  open   .kibana.3c767c41afb12ada140190ed82db3fd930e2efa3                      1   0          8            0     77.6kb         77.6kb                                                                        green  open   project.kube-system.7dce2381-47d8-11e8-9763-022e0d5a511e.2018.04.24   1   0      38351            0     14.9mb         14.9mb                                                                        
green  open   project.svt3.c5804343-47ee-11e8-ae99-022e0d5a511e.2018.04.24          1   0       1957            0    676.3kb        676.3kb                                                                        
green  open   project.svt2.c54d857d-47ee-11e8-ae99-022e0d5a511e.2018.04.24          1   0       1956            0    602.2kb        602.2kb                                                                        
green  open   .searchguard.logging-es-data-master-sxvdpq1q                          1   0          5            0     33.8kb         33.8kb               


Running REST API searches against the index for known strings in the master logs returns good results.  

However, after logging in to kibana as a cluster-admin, the kibana index dropdown (see attached screenshot)  does not show the kube-system project and searching .all returns no known good strings.  Selecting the kubernetes.namespace_name stat widget in the menu shows other namespaces, but not kube-system.

I believe cluster-admins should be able to see the master and etcd pod logs.


Version-Release number of selected component (if applicable): logging 3.10.0-0.27.0


How reproducible: Always (2 for 2 for my installs, anyways)


Steps to Reproduce:
1. Install an OpenShift 3.10 cluster
2. Deploy logging 3.10.0-0.27.0 (inventory below, adjust as appropriate)
3. Implement the fluentd workaround in bug 1569106
4. Use the ES REST API to verifiy kube-system index exists and is queryable
5. Log in to kibana as cluster-admin

Actual results:

kube-system project logs not available

Expected results:

cluster-admin can view pod logs for control plane pods

Additional info:

[OSEv3:children]                                                      
masters                                                               
etcd                                                                  
                                                                

[masters]                                                             
ip-172-31-2-228                         

[etcd]                                                                
ip-172-31-2-228                          

                      

[OSEv3:vars]                                                          
deployment_type=openshift-enterprise                                  

openshift_deployment_type=openshift-enterprise                        
openshift_release=v3.10                                                
openshift_docker_additional_registries=registry.reg-aws.openshift.com 


openshift_logging_install_logging=true                                
openshift_logging_master_url=https://ec2-54-191-21-206.us-west-2.compute.amazonaws.com:8443
openshift_logging_master_public_url=https://ec2-54-191-21-206.us-west-2.compute.amazonaws.com:8443
openshift_logging_kibana_hostname=kibana.apps.0424-rtb.qe.rhcloud.com                                        
openshift_logging_image_prefix=registry.reg-aws.openshift.com:443/openshift3/                                                                
openshift_logging_image_version=v3.10                               
openshift_logging_es_cluster_size=1                                   
openshift_logging_es_pvc_dynamic=true                                 
openshift_logging_es_pvc_size=20Gi                                    
openshift_logging_es_pvc_storage_class_name=gp2                       
openshift_logging_fluentd_read_from_head=false                                                           
openshift_logging_curator_nodeselector={"region": "infra"}            
openshift_logging_kibana_nodeselector={"region": "infra"}             
openshift_logging_es_nodeselector={"region": "infra"}