The issue looks similar to the one fixed in BZ 1523142 , however it is observed while mounting ConfigMap & secrets volumes. This is resulting in very slow deployment. This issue is generally observed when the node is running around 180-200 pods, however CPU, memory resource are not exhausted on node. This issue started impacting deployment after upgrade to 3.7
*** Bug 1601398 has been marked as a duplicate of this bug. ***
*** Bug 1506275 has been marked as a duplicate of this bug. ***
Updating priority and severity since a work-around exists.
*** Bug 1689364 has been marked as a duplicate of this bug. ***
In 3.10 externalKubernetesClientConnectionOverrides was removed: https://github.com/openshift/origin/commit/704f432db592630455b5ceece908b84614dccccc#diff-29dc1a6e13ab4dfdfcbf8bdbbfe0fb6e Based on the PR here we never supported this client, changes here likely are not doing any thing for other OCP versions either. https://github.com/openshift/origin/pull/19191
Created bug https://bugzilla.redhat.com/show_bug.cgi?id=1691046 around masterClientConnectionOverrides being ignored when set.
This likely can be closed as a duplicate to this bug: https://bugzilla.redhat.com/show_bug.cgi?id=1691046
As stated in other comments, we have addressed this with documentation. If increases to node-side API rate limits do not fix a customer's issues, please open a new bug, as this bug is related to client-side api rate limiting on the node (which is fixed by increasing the limits as described in the new docs). (See https://bugzilla.redhat.com/show_bug.cgi?id=1691046#c15 for links to the docs.) *** This bug has been marked as a duplicate of bug 1691046 ***