I was able to index 230K documents overnight, doing a curl on health and indices I can see the index and the health returns a yellow status - everything just like on the previous version of ES - except I can't seem to connect to instance.
I've tried 3rd party tools - that all work on the last version of ES - and none connect (ElasticHQ for example). When I try to "search" my site returns a blank, white, page - no errors.
The installation process seemed very different for this version - as does the start process - and I can't get ES to run as a service with this version. Is it supposed to be installed in a specific location? Right now it's in /opt/elasticsearch-1.6.0
Is it possible to downgrade the installation or copy the index into an installation of a previous version? I just don't want to spend another 8 hours indexing today
Also, when it started, the process said "recovered [0] indices…" and then hung, I had to cntr-c out of it, but I posted it as an issue on Github last night and was told that was okay.
That's the first I've seen Tomcat mentioned on this server. I'm running Apache2, php5.9,Mariadb10.x and not much else. No idea how Tomcat got into this mix - like I said, I just followed the instructions as I have with previous versions, on the same image I've been using for months.
Hmm ok, might just be an environment variable.
Why not just use a deb/rpm to install ES, that way you get the service to handle starting of the service?
Kibana 3 against ElasticSearch 1.4 throws an Connection Failed screen. The error text says to set http.cors.allow-origin, but it misses out the important http.cors.enabled: true
Working config:
$ grep cors elasticsearch-1.4.0.Beta1/config/elasticsearch.yml
http.cors.allow-origin: "/.*/"
http.cors.enabled: true
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.