Hi,
You seem to have quite a large number of shards (1180) for a single node
with only 7GB heap. As the total data volume is a bit over 600GB, the
average shard size is only a bit over 500MB, which is not very large. As
each shard is a separate Lucene index and carries some overhead, you would
I'd guess that you are hitting the capacity of the node.
Try closing/deleting indices or upgrading the instance.
On 19 April 2015 at 13:03, Dave Galbraith david92galbra...@gmail.com
wrote:
So I've got this Elasticsearch single-node instance running out on an EC2
m3.2xlarge in the cloud. I'm
So I've got this Elasticsearch single-node instance running out on an EC2
m3.2xlarge in the cloud. I'm sending about 2,000 documents per second to it
for writing. But it's dropping almost all of them on the floor, and any
search queries I try to make against it are taking an extraordinarily
