Hello,
we have a ES cluster 1.7 (it's production so we couldn' t upgrade it to a more recent version) and we are storing network device syslogs in daily indexes.
Because the volume of data collected is huge, we have scripts to manage the log retention policy: we keep security logs for 90 days but traffic logs (99% of the data) are deleted after 8 days.
To delete we use this kind of query:
2017-03-06 15:36:03: curl -XDELETE '127.0.0.1:9200/prodlogs-2017-02-27/logs/_query' -d '{
"query": {
"bool": {
"must": [{
"query_string": {
"query": "type:traffic",
"default_operator": "AND"
}
}, {
"range": {
"date": {
"lt": "now-8d"
}
}
}]
}
}
}'
so every day we delete 98-99% of the data in the index and the kopf plugin show that the number of document is reduced as expected.
The problem we have is that the size of the index is no decreased.
I have tried to optimize the index but the size stays the same.
This is a big problem for us because we are running out of disk space (4Tb).
Any idea here would be more that welcome
Antoine