Elasticsearch index size is less than it takes on disk


I noticed strange issue with disk space consumption. I checked how much space my indexes take:

GET _cat/indices/*?v&s=store.size

Than I calculated total size (ofcoure store.size) it's about 1250 GB. Then I checked system information:

GET _cat/allocation?v

shards disk.indices disk.used disk.avail disk.total disk.percent
2030 1.5tb 1.6tb 272.1gb 1.9tb 86
2030 1.5tb 1.6tb 271.7gb 1.9tb 86

Also I cheked disk size with "du" command and get the same result 1.6 TB used per node.

This is dedicated drive for elasticsearch only. There is only elasticsearch information, I mean only /nodes directory.

So my qestion is - why Lucene indexes take 3 times more space than Elasticsearch indexes.

I hope I provided clear explanation of my issue.

How did you arrive at this conclusion based on the data you provided? I am afraid I do not follow the calculation.

OK, I can explain.

I think its clear why elasticsearch indexes take 1250 GB.

Also I have 2 nodes - 2 TB per nodes.

So, since I see in elastic that total index size devided between 2 nodes is 1250 GB.
And disk space consumed is about 3.2 TB (1.6 +1.6TB). So disk space usage is almost 3 times larger than elasticsearch indexes.

So, nobody knows why i have this behavior.

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.