We are using ElasticSearch version 1.2.1 in a two node cluster. We have documents with two different nested properties: 'addresses' and 'employeeList'. Nested queries on those nested properties were working as expected until we restarted both nodes recently.
Queries on 'addresses' work as expected, but any query involving 'employeeList' is returning 0 results now. We have encountered the same issue once before when our hosted VMs were stopped, and we had to restart them (and ElasticSearch running on them). Our solution before was to reindex the data to a new index. I was hoping there would be another solution as it takes us about 24 hours to reindex all of our data.
We have tried reindexing a few documents to the same index, but there are still zero results. I am confused as to why we couldn't query one property vs the other. The 'employeeList' property will almost always have an order of magnitude more documents than 'addresses'. The 'addresses' property list is usually a few documents at most while 'employeeList' will can have up to 1000, and frequently has near 100. We have issues with circuit breaking for these queries in the past, but seemed to reach a stable point. There are no errors in our application logs or the elasticsearch logs that we can find that are relevant.
Any help would be appreciated! Below is an example query and our index settings/mappings.
Here's an example query:
{
"from": 0,
"size": 10,
"query": {
"bool": {
"must": [
{
"bool": {
"should": [],
"must_not": []
}
},
{
"nested": {
"path": "employeeList",
"query": {
"bool": {
"should": [
{
"wildcard": {
"title": "ceo"
}
}
],
"must_not": []
}
}
}
}
],
"should": [],
"must_not": [
{
"term": {
"calculatedStatus": "INACTIVE"
}
}
],
"minimum_should_match": 1
}
},
"min_score": 0,
"sort": [
{
"contactInfoScore": {
"order": "desc"
}
},
{
"_score": {
"order": "desc"
}
},
{
"id": {
"order": "asc"
}
}
]
}
Here's the mappings:
{
"settings": {
"index": {
"number_of_shards": 10,
"number_of_replicas": 1,
"analysis": {
"text": [
"Lynch & Robbins, P.A.",
"99Bottles of Beer, LLC"
],
"filter": {
"english_stop": {
"type": "stop",
"stopwords": "english"
}
},
"analyzer": {
"just_lower": {
"type": "custom",
"filter": [
"lowercase"
],
"tokenizer": "keyword"
},
"processed": {
"type": "custom",
"filter": [
"word_delimiter",
"lowercase",
"stemmer"
],
"tokenizer": "whitespace"
},
"startswith": {
"tokenizer": "keyword",
"filter": [
"lowercase",
"english_stop"
]
}
}
}
}
},
"mappings": {
"company": {
"_id": {
"index": "not_analyzed",
"store": true,
"path": "id",
"doc_values": true
},
"properties": {
"id": {
"type": "string",
"index": "not_analyzed",
"doc_values": true
},
"addresses": {
"type": "nested",
"properties": {
"pretty": {
"type": "string",
"analyzer": "just_lower"
},
"raw": {
"type": "string",
"analyzer": "just_lower"
},
"city": {
"type": "string",
"analyzer": "just_lower"
},
"zip": {
"type": "string",
"analyzer": "just_lower"
},
"state": {
"type": "string",
"analyzer": "just_lower"
},
"county": {
"type": "string",
"analyzer": "just_lower"
},
"latitude": {
"null_value": -1000,
"index": "not_analyzed",
"type": "double"
},
"longitude": {
"null_value": -1000,
"index": "not_analyzed",
"type": "double"
}
}
},
"employeeList": {
"type": "nested",
"properties": {
"title": {
"type": "string",
"analyzer": "just_lower"
}
}
}
}
}
}
}