I use ES to store files, and the size of each file is within 100kb. The following two problems encountered in the use process have not been solved:
1.We set the requests cache: false, and the memory will still slowly reach 99%. Even if we only use -XHEAD to determine whether there is no other operations, the memory will be 99%. Is there any way to prevent the memory usage from being so high?
2.When we use the 7.4 version, although the memory is 99%, the system can still run normally for a long time, but the updated version after 7.11 will cause the system to fail to run because the memory is 100% every 2 days. The high version cannot roll back the low version, so we are very painful and have to restart the service every two days.
Cluster：No
CPU：X5690，3.47GHz
Memory：64G
System：WinServer 2008 R2 Enterprise
elasticsearch.yml：Default configuration
JAVA：-Xmx30g -Xms30g
The amount of data：600W
Disk usage：400G
Index number：7
mapping：
{
"order": 0,
"index_patterns": [
"db_f_*",
],
"settings": {
"index": {
"max_result_window": "50000",
"refresh_interval": "1m",
"requests": {
"cache": {
"enable": "false"
}
},
"number_of_shards": "4",
"translog": {
"flush_threshold_size": "500m",
"sync_interval": "120s",
"durability": "async"
},
"number_of_replicas": "0"
}
},
"mappings": {
"properties": {
"iD": {
"type": "keyword"
},
"data": {
"type": "binary"
},
"date": {
"type": "long"
}
}
},
"aliases": {}
}