Hello, everyone!
This post describes Health Status of Solr Instances is Bad.
[Applicable Versions]
6.5.x
[Symptom]
After Solr is started, the health status of some instances Concerning, and then changes to Bad.
[Fault Locating]
The health check of Solr instances is abnormal, which has multiple causes:
The curl installation package version is incorrect. You can run the curl -V command to check whether the curl version contains GSS-Negotiate to confirm whether it is provided by FusionInsight. If the curl version is not provided by FusionInsight, you need to uninstall the existing curl package.
The port configured for the instance is occupied. You need to change the instance port and then restart the instance process.
The node hostname is configured for several times. In this case, the curl command of the health check fails to be executed. You need to configure the hostname only once.
The solr_SolrServer*.tgt file under the /tmp directory of the node where the instance resides is not updated. Using the old tgt file leads to authentication failures. As a result, the health check script fails to be executed. You need to delete the solr_SolrServer*.tgt file manually.
[Solution]
Run the curl command on the node where the instance resides and access https://$hostName:$SOLR_SEC_PORT/solr/admin/collections?action=LIST. Then determine the cause of the issue and take the corresponding measure based on the returned information of the request.
Hope you can learn from it, thank you!
