I am building an elk cluster to index and search lots of http access log,
about more than 7000Event per second and also there will be more than 100
cocurrent searchs.
I have 2 machines. One of them has 24 cpu cores, 64G memory and 2T sata
disk(no raid). The other one is much powerful, which has 24 core cpu, 384G
memory and 300G sas disk*8.
My plan is to build a 3-node elasticsearch, one running on small server, the
other two running on the big one. Can I route all index request to one node
while all search request to the other two nodes? Is this a good idea to do
like this? Any comments?
I am building an elk cluster to index and search lots of http access log,
about more than 7000Event per second and also there will be more than 100
cocurrent searchs.
I have 2 machines. One of them has 24 cpu cores, 64G memory and 2T sata
disk(no raid). The other one is much powerful, which has 24 core cpu, 384G
memory and 300G sas disk*8.
My plan is to build a 3-node elasticsearch, one running on small server,
the other two running on the big one. Can I route all index request to one
node while all search request to the other two nodes? Is this a good idea
to do like this? Any comments?
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.