New to ec2/S3 and ES and working through some growing pains. I have
followed the instructions via http://www.elasticsearch.org/tutorials/elasticsearch-on-ec2/ and can't seem
to find where to set my env. var's for the likes of ES_MIN_MEM, etc.
Do you know of any URLs or videos that outline this process step by step?
The EC site seems a little short on details and there seems to be a lot of
conflicting issues with different users when setting this up.
If you're on a debian-based system, you can declare vars inline like so:
ES_MAX_MEM=2gb /usr/local/bin/elasticsearch -f
And when you're in production, and using the service wrapper, declare those vars in /etc/init.d/elasticsearch, If you just want it to work, be fast, and reliable check us out at qbox.io ;).
If you're on a debian-based system, you can declare vars inline like so:
ES_MAX_MEM=2gb /usr/local/bin/elasticsearch -f
And when you're in production, and using the service wrapper, declare those
vars in /etc/init.d/elasticsearch, If you just want it to work, be fast, and
reliable check us out at qbox.io http://qbox.io ;).
You need to set the ES_HEAP_SIZE environment variable to some reasonable value (~300MB?), and disable the bootstrap.mlockall setting, micros and weak VMs have trouble with it.
Also, install ES from the official deb/rpm packages, so you don't wrestle with the settings being applied.
Lastly, note that a micro instance is useful mainly for testing/evaluation, or a system with very low demands.
I played around with microinstances in the past. I would use a 32-bit
instance/JVM whenever possible since there is no need larger object
references, although the JVM will compress the object references for heaps
under 32GB.
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.
