I have a small three node Elasticsearch cluster running in AWS with an Elastic Load Balancer in place pointing to all three nodes in the cluster. I set the HistoryStorageURL parameter in my zabbix_server.conf file to point to the ELB, however, when I restart the zabbix server all the History syncher processes appear to be connecting to the same node in cluster and are overwhelming it with search requests resulting in multiple HTTP 503 errors and a VERY slow history sync.
Is there are way to get the HistoryStorageURL setting to load balance against multiple nodes in a cluster? Does the parameter support multiple URLs?
e.g. HostiryStorageURL=http://es1.lan:9200,http://es2.lan:9200,http://es3.lan:9200 . ???
Is there are way to get the HistoryStorageURL setting to load balance against multiple nodes in a cluster? Does the parameter support multiple URLs?
e.g. HostiryStorageURL=http://es1.lan:9200,http://es2.lan:9200,http://es3.lan:9200 . ???
Comment