How regionserver in hbase will read huge hfiles?

[#1]

Let say i have a region with region max store file size 10 GB and after major compaction i have a single hfile with size 9GB(stored in hadoop). Now my doubt is how regionserver will read and process that 9GB file from hadoop.

Will it initially load that file to regionserver inmemory? if yes how it is possible if my heap size is only 8GB?

[#2 · Accepted solution]

HBase does not have to read an entire HFile into memory to service reads.

[#3]

@Josh Elser

Tnx dude for your reply, can you please explain me how query matcher in hbase work or else if you have a good link regarding query matcher please share with me, it will helpful for me to understand.

Thank you