We have an operator that is deployed after creating a catalog and a subscription.
When a subscription is created while the catalog pod is still being created, it may take more than 10 minutes for the catalog-operator to create an install plan that creates the operator.
Wondering if we could shorten this delay.
This delay doesn't occur if we wait for the catalog pod to be in ready state, before creating the subscription.
Log from catalog-operator:
E0610 18:30:54.401951 1 queueinformer_operator.go:186] Sync "kubevirt-hyperconverged" failed: {kubevirt-hyperconverged alpha {hco-catalogsource openshift-operator-lifecycle-manager}} not found: CatalogSource {hco-catalogsource openshift-operator-lifecycle-manager} not found
time="2019-06-10T18:30:54Z" level=info msg="building connection to registry" currentSource="{hco-catalogsource openshift-operator-lifecycle-manager}" id=WT+5A source=hco-catalogsource
time="2019-06-10T18:30:54Z" level=info msg="client hasn't yet become healthy, attempt a health check" currentSource="{hco-catalogsource openshift-operator-lifecycle-manager}" id=WT+5A source=hco-catalogsource
time="2019-06-10T18:30:57Z" level=info msg="building connection to registry" currentSource="{hco-catalogsource openshift-operator-lifecycle-manager}" id=rEOVW source=hco-catalogsource
time="2019-06-10T18:30:57Z" level=info msg="client hasn't yet become healthy, attempt a health check" currentSource="{hco-catalogsource openshift-operator-lifecycle-manager}" id=rEOVW source=hco-catalogsource
time="2019-06-10T18:31:00Z" level=info msg="building connection to registry" currentSource="{hco-catalogsource openshift-operator-lifecycle-manager}" id=T7Rk9 source=hco-catalogsource
time="2019-06-10T18:31:00Z" level=info msg="client hasn't yet become healthy, attempt a health check" currentSource="{hco-catalogsource openshift-operator-lifecycle-manager}" id=T7Rk9 source=hco-catalogsource
time="2019-06-10T18:31:03Z" level=info msg="building connection to registry" currentSource="{hco-catalogsource openshift-operator-lifecycle-manager}" id=ucqKD source=hco-catalogsource
time="2019-06-10T18:31:03Z" level=info msg="client hasn't yet become healthy, attempt a health check" currentSource="{hco-catalogsource openshift-operator-lifecycle-manager}" id=ucqKD source=hco-catalogsource
time="2019-06-10T18:43:59Z" level=warning msg="no installplan found with matching manifests, creating new one" id=F1v3Y namespace=kubevirt-hyperconverged
time="2019-06-10T18:43:59Z" level=info msg=syncing id=xN+BQ ip=install-rjfz6 namespace=kubevirt-hyperconverged phase=
time="2019-06-10T18:43:59Z" level=info msg="skip processing installplan without status - subscription sync responsible for initial status" id=xN+BQ ip=install-rjfz6 namespace=kubevirt-hyperconverged phase=
time="2019-06-10T18:43:59Z" level=info msg=syncing id=Pttev ip=install-rjfz6 namespace=kubevirt-hyperconverged phase=Installing
time="2019-06-10T18:44:00Z" level=info msg=syncing id=Lc4qi ip=install-mlp5v namespace=openshift-operator-lifecycle-manager phase=Complete
time="2019-06-10T18:44:08Z" level=info msg=syncing id=BJEkS ip=install-rjfz6 namespace=kubevirt-hyperconverged phase=Complete
We have an operator that is deployed after creating a catalog and a subscription. When a subscription is created while the catalog pod is still being created, it may take more than 10 minutes for the catalog-operator to create an install plan that creates the operator.
Wondering if we could shorten this delay.
This delay doesn't occur if we wait for the catalog pod to be in ready state, before creating the subscription.
Log from catalog-operator: