I'm in a bit of a crunch. We have some servers we added to our cluster
whose processors were not as good as the older servers. I'd rather not get
into how we got swindled into accepting them, but this leaves me in a place
with imbalanced hardware between my nodes.
The crunch bit comes in that during our peak load we begin to have slow
queries which lead to rejected queries once we max out the CPUs on the new
machines.
I'm trying to find a way to tell ES to put more shards on the more powerful
nodes.
The simplest idea, would be if I could give a weight to servers or
something else rather than try to move shards around by hand, create zones
and force some indexes into them, etc.
I'm in a bit of a crunch. We have some servers we added to our cluster
whose processors were not as good as the older servers. I'd rather not get
into how we got swindled into accepting them, but this leaves me in a place
with imbalanced hardware between my nodes.
The crunch bit comes in that during our peak load we begin to have slow
queries which lead to rejected queries once we max out the CPUs on the new
machines.
I'm trying to find a way to tell ES to put more shards on the more
powerful nodes.
The simplest idea, would be if I could give a weight to servers or
something else rather than try to move shards around by hand, create zones
and force some indexes into them, etc.
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.