Closed wkruse closed 3 years ago
We also tried setting validate_after_inactivity
to 0
without success.
output {
opensearch {
hosts => ["opensearch:9450"]
validate_after_inactivity => 0
index => "logstash-%{[system]}-%{+YYYY.MM.dd}"
}
}
@wkruse Thanks for reporting this bug. I see that you are running opensearch in backward compatible mode from version number. It is known bug and fix is merged in 1.1. we will release it on Sept 21.
@wkruse we release 1.1.0 with fix. Can you please uninstall and install plugin and let us know did that solve your problem?
bin/logstash-plugin uninstall logstash-output-opensearch
bin/logstash-plugin install logstash-output-opensearch
@VijayanB 1.1.0 solved it for us. Thank you!
Describe the bug
We are getting
Could not connect to cluster
errors andRestored connection to OpenSearch instance
warnings every 5 seconds.Running
tcpdump
we see that the health checks succeed.OpenSearch output plugin config
To Reproduce Steps to reproduce the behavior:
Expected behavior No
Could not connect to cluster
errors.Host/Environment (please complete the following information):
34.20210821.3.0
1.22.2
logstash-oss-with-opensearch-output-plugin:7.13.2
opensearchproject/opensearch:1.0.1
Additional context
opensearch:9450
is a service in Kubernetes in front of 3 OpenSearch nodes. We also tried the plugin with a list of our 3 nodes (both internal Kubernetes StatefulSet names and external host names) and were still getting the errors.Logstash OSS 7.10.2 with Elasticsearch output plugin works without problems.