I am having a client node and a data node. i am using the client node to communicate to data node. My question is when i perform an insertion from client node does the program control returns to client node once the information is sent to data node over the network or is it blocked till the data is inserted and indexed on the data node.
If you are using the java API the answer is maybe, it depends on how you use it. In addition, even if you use the synchronous API you have to wait for a refresh for you data to be searchable. By default they happen every second.
I am having a client node and a data node. i am using the client node to
communicate to data node. My question is when i perform an insertion from
client node does the program control returns to client node once the
information is sent to data node over the network or is it blocked till the
data is inserted and indexed on the data node.
Good question and it's worth a few links This link might be helpful,
look at async replication and refresh
Also look at translog
At this point you'll be confused between refresh and flush, then read this
On Sun, Feb 2, 2014 at 1:05 PM, Nik Everett nik9000@gmail.com wrote:
If you are using the java API the answer is maybe, it depends on how you
use it. In addition, even if you use the synchronous API you have to wait
for a refresh for you data to be searchable. By default they happen every
second.
I am having a client node and a data node. i am using the client node to
communicate to data node. My question is when i perform an insertion from
client node does the program control returns to client node once the
information is sent to data node over the network or is it blocked till
the
data is inserted and indexed on the data node.
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.