I've hit several instances of FLP not able to connect to Kafka as documented in NETOBSERV-672 which also has workaround to delete FLP pods however this time it was in unrecoverable state and those workarounds did not help.
Confirmed that ebpf was sending logs successfully.
See discussion: https://coreos.slack.com/archives/C02939DP5L5/p1668739325026059
It was unclear what was causing it. Attaching flowcollector Kafka_issue_1118.txt manifest and logs from netobserv NS inspect.local.4758633599300681843.zip
- is related to
-
NETOBSERV-672 FlowLogs-Pipeline not reconnecting to Kafka under undetermined circumstances
-
- Closed
-