The common denominator is that when I start a [*node.master: false] [node.data: false] *node fist, the cluster is not created properly, no matter how the discovery is configured, EC2 or unicast localhost.
On Tuesday, March 18, 2014 10:41:39 AM UTC-4, Bastien Chong wrote: > > Hi, > > I have a server with 2 ES instances. The first one is the master one, used > to store documents. The second one is just there to receive requests from > Kibana (I call it ES read-only), it has : > > node.master: false >> node.data: false >> > > Both are configured with cloud-aws plugin, and the http/java port are left > by default for automatic assignment. > > When master start first, it's binded to 9200/9300, then the second one to > 9201/9300. When I do : *curl -XGET > 'http://localhost:9200/_cluster/health?pretty=true > <http://localhost:9200/_cluster/health?pretty=true>'* everything is > working as expected and both nodes are in the same cluster. > > *But*, if I start the read-only instance first, discovery stop working. I > have enabled DEBUG and TRACE but I didn't found what's the issue. > > I also tried to hardcode the port allocation : > > transport.tcp.port: 930(0/1) >> http.port: 920(0-1) >> > > And after that, It's actually worse, whatever the order I start the > instances, EC2 discovery is broken. So it's sort of a race-condition that's > happening. > -- You received this message because you are subscribed to the Google Groups "elasticsearch" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/elasticsearch/d18b9838-188a-4efe-bb0f-dccba07fbd79%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
