get logs from only the running NMO pods

Signed-off-by: oraz <[email protected]>

must-gather/README.md

@@ -5,10 +5,10 @@ You can use the `oc adm must-gather` command to collect information about your c
 With the node-maintenance-must-gather image you can collect manifests and logs related to node maintenance:
 - Node objects 
 - Custom Resource Definition
-- Node Maintenance Operator pod's logs
+- Node Maintenance Operator pod's logs (dismiss drained pods)
 - Custom Resources
 - Cluster's resource (CPU, and Memory) usage
-- Nodes' logs
+
 
 To collect this data, you must specify the extra image using the `--image` option.
 Example:

must-gather/collection-scripts/gather

@@ -1,7 +1,7 @@
 #!/bin/bash
 
-# # Create a directory for must-gather
-mkdir -p must-gather/operator-pod-logs/ must-gather/nodes-logs
+# Create a directory for must-gather
+mkdir -p must-gather/operator-pod-logs/
 
 # Generate /must-gather/version file
 DIR_NAME=$( cd -- "$( dirname -- "${BASH_SOURCE[0]}" )" &> /dev/null && pwd )
@@ -20,10 +20,10 @@ group_resources=()
 # Get namespace of node-maintenance-operator - where it is installed
 NMO_NAMESPACE=$(oc get subs -A --field-selector=metadata.name=node-maintenance-operator -o jsonpath='{.items[*].metadata.namespace}')
 
-# Get NMO Information - nodes' object, CRD, pod's logs, CRs, cluster's resource usage, and nodes' logs
+# Get NMO Information - nodes' object, CRD, pod's logs, CRs, cluster's resource usage.
 
 # Get Nodes' names
-NODES_NAMES=($(oc get nodes -o jsonpath='{.items[*].metadata.name}{"\n"}'))
+# NODES_NAMES=($(oc get nodes -o jsonpath='{.items[*].metadata.name}{"\n"}'))
 
 # Nodes' objects
 group_resources+=(nodes)
@@ -32,8 +32,8 @@ group_resources+=(nodes)
 NMO_CRD=$(oc get crds -o jsonpath='{range .items[*]}{"crd/"}{.metadata.name}{"\n"}{end}' | grep 'nodemaintenance.medik8s' | sed -z 's/\n/ /g')
 named_resources+=(${NMO_CRD})
 
-# NMO's POD name
-NMO_POD=$(oc get pods -o jsonpath='{range .items[*]}{"pod/"}{.metadata.name}{"\n"}{end}' -n ${NMO_NAMESPACE}  | grep 'node-maintenance')
+# NMO's running POD name
+NMO_PODS=($(oc get pods -o jsonpath='{range .items[*]}{.metadata.name}{"\n"}{end}' -n ${NMO_NAMESPACE}  | grep 'node-maintenance'))
 
 # Node Maintenance CRs
 group_resources+=(nm)
@@ -43,13 +43,15 @@ oc adm inspect --dest-dir must-gather --all-namespaces "${named_resources[@]}"
 group_resources_text=$(IFS=, ; echo "${group_resources[*]}")
 oc adm inspect --dest-dir must-gather --all-namespaces "${group_resources_text}"
 
-# Get pod's logs 
-echo "$(oc logs ${NMO_POD} -n ${NMO_NAMESPACE})" > must-gather/operator-pod-logs/${OPERATOR_NAME}
+# Get pod's logs for only the running pods. Dismiss drained NMO pods
+for NMO_POD_NAME in ${NMO_PODS[@]}; 
+do  if  [ $(oc get pod ${NMO_POD_NAME} -n ${NMO_NAMESPACE} -o jsonpath='{.status.phase}') == "Running" ];
+then oc logs ${NMO_POD_NAME} -n ${NMO_NAMESPACE} > must-gather/operator-pod-logs/${NMO_POD_NAME}; fi; done
 
 # Get cluster's resource (CPU and Memory) usage 
-echo "$(oc adm top node --use-protocol-buffers)" > must-gather/cluster_resoruce_statistics
+oc adm top node --use-protocol-buffers > must-gather/cluster_resoruce_statistics
 
 # Get all journal logs from each Node (TODO: select particluar journal logs, e.g., crio )
-for NODE in ${NODES_NAMES[@]}; do oc adm node-logs ${NODE} > must-gather/nodes-logs/${NODE}; done
+# for NODE in ${NODES_NAMES[@]}; do oc adm node-logs ${NODE} > must-gather/nodes-logs/${NODE}; done
 
 exit 0