Description of problem:
Fluentd daemonset doesn't have a tolerate everything toleration which looks like this:
- operator: "Exists"
This means if a node is tainted fluentd won't be scheduled there.
Version-Release number of the following components:
All 3.11 branch
Steps to Reproduce:
1. oc taint node <node name> ataint=tainted:NoSchedule
fluentd won't be scheduled on tainted nodes
fluentd will be scheduled on tainted nodes
Created a pull request to fix it.
Please update the doctype and text.
features: Basic-Auth GSSAPI Kerberos SPNEGO
Steps to reproduce:
1. taint the node
# oc adm taint node $node NodeWithImpairedVolumes=true:NoExecute
#oc describe node $node | grep -i taint
2. delete the fluentd pods
# oc get ds logging-fluentd -o yaml > log-ds.yaml
# oc delete ds logging-fluentd
# oc create -f logging-fluentd
3. Note the logging pods
The logging pods get created with out any issue in spite of taint being present on the node.
oc get pods -n openshift-node -o wide | grep $node ; oc get pods -n openshift-sdn -o wide | grep $node ; oc get pods -n openshift-logging -o wide | grep $node
sync-6gbl8 1/1 Running 0 2h 172.31.15.203 ip-172-31-15-203.us-west-2.compute.internal <none>
ovs-xnxtq 1/1 Running 0 2h 172.31.15.203 ip-172-31-15-203.us-west-2.compute.internal <none>
sdn-7dflr 1/1 Running 0 2h 172.31.15.203 ip-172-31-15-203.us-west-2.compute.internal <none>
logging-fluentd-r8lc4 1/1 Running 0 1h 172.22.0.2 ip-172-31-15-203.us-west-2.compute.internal <none>
Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.
For information on the advisory, and where to find the updated
files, follow the link below.
If the solution does not work for you, open a new bug report.