ElasticSearch report a strange os memory and search very slow,

Elastic Stack Elasticsearch
1

Running ES in docker on a physical box with VMware ESXi6 CentOS 7 OS

Queries are currently taking more than 10 seconds to complete

_cluster/stats reports a strange os memory:

total_in_bytes: 134778593280
free_in_bytes: 2296119296
used_in_bytes: 132482473984
free_percent: 2

Heap usage looks good.

29 AM
Screen Shot 2017-12-12 at 11.37.29 AM.png1606×824 92.5 KB

However, when I try free command:

Mem total: 65809860
Mem used: 36892596
Mem available: 24905108

37 AM
Screen Shot 2017-12-12 at 11.38.37 AM.png1588×132 6.04 KB

the number reported by ES is strange to me.

Any idea? Thanks.

2

You're looking at cluster stats which aggregates across your two nodes. Instead, look at node-level stats.

3

Thanks, yes figured out, the two nodes are on same machine, so it doubles the os memory.

4

However, still the search is super slow. About 10X slower than other direct physical boxes with same deployment configurations.

5

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