Hi,
I am trying to query Elasticsearch to show for each week, the average count of documents per day. I understand I must use the pipeline aggregation to achieve this, however I am having an issue with it...
This is my request (that I am firing from UNIX command line):
curl -XGET '<host>:<port>/<indices>/_search' -d '
{
"size" : 0,
"aggs" : {
"weekly_agg" : {
"date_histogram" : {
"field" : "@timestamp",
"interval" : "week"
},
"aggs" : {
"daily_agg" : {
"date_histogram" : {
"field" : "@timestamp",
"interval" : "day"
}
}
}
},
"avg_monthly_daily_count" : {
"avg_bucket" : {
"buckets_path" : "weekly_agg>daily_agg"
}
}
}
}'
Host, port, and index names are masked as above for security reasons. From this request the response boils down to:
Parse Failure [Could not find aggregator type [avg_bucket] in [avg_monthly_daily_count]]]; }]","status":400}
Unfortunately I cannot show the entire response as it reveals too much about our set up. Please let me know if you need more information.
Cheers,
Ben