INdex not rolling over correctly

Elastic Stack Elasticsearch
1

I have a ILM policy
template, policy, alias

{
  "prod3_template" : {
    "order" : 1,
    "index_patterns" : [
      "prod3-cloudwatch*"
    ],
    "settings" : {
      "index" : {
        "lifecycle" : {
          "name" : "prod3_policy",
          "rollover_alias" : "prod3-cloudwatch"
        },
        "number_of_shards" : "1",
        "number_of_replicas" : "1"
      }
    },
    "mappings" : { },
    "aliases" : { }
  }
}

policy

{
  "prod3_policy" : {
    "version" : 3,
    "modified_date" : "2020-04-19T17:32:20.186Z",
    "policy" : {
      "phases" : {
        "hot" : {
          "min_age" : "0ms",
          "actions" : {
            "rollover" : {
              "max_size" : "50gb",
              "max_age" : "5d"
            }
          }
        },
        "delete" : {
          "min_age" : "30d",
          "actions" : {
            "delete" : { }
          }
        }
      }
    }
  }
}

My index is now 75G but has not rolled over

Here is the explain

{
  "indices" : {
    "prod3-cloudwatch-2020.04.18-1" : {
      "index" : "prod3-cloudwatch-2020.04.18-1",
      "managed" : true,
      "policy" : "prod3_policy",
      "lifecycle_date_millis" : 1587229446320,
      "phase" : "hot",
      "phase_time_millis" : 1587229447358,
      "action" : "rollover",
      "action_time_millis" : 1587229655594,
      "step" : "check-rollover-ready",
      "step_time_millis" : 1587229655594,
      "phase_execution" : {
        "policy" : "prod3_policy",
        "phase_definition" : {
          "min_age" : "0ms",
          "actions" : {
            "rollover" : {
              "max_size" : "50gb"
            }
          }
        },
        "version" : 2,
        "modified_date_in_millis" : 1587229330842
      }
    }
  }
}

Can u please help as to why its not rolling over

2

Rollover is based on the size of the primary shards and as you have a replica you probably have not reached the limit yet.

3

Hi Christian
Not sure what does that mean, isnt replica replicates the shards, if the index size is 75G how is shared between primary and replica. Can i tell the size of the primary and replica, plz advise.

thx

4

Look at the cat shards API.

5

Thanks so the index shown in kibana is total with replica, so when primary goes to 50G then wiil rollover?

thx for the quick turn around

prod3-cloudwatch-2020.04.18-1 0 p STARTED 46833737 39.4gb 172.20.224.142 elasticsearch-data-0
prod3-cloudwatch-2020.04.18-1 0 r STARTED 46833737 39.3gb 172.20.224.2 elasticsearch-data-5

6

Thank you Christian, that was it, it rolled over at 100GB(50GB prim+ 50GBcopy). Appreciate your help

7

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