If a search is performed with an index specified:
curl -XPOST http://localhost:9200/my_specific_index/_search/ -d '{
"fields": [
"upc"
],
"query": {
"match_all": {}
}
}'
Then I receive the correct # of total hits and 10 upc fields are
returned in the hits array as results.
BUT if the index is not specified then 3 out of the 10 entries that
are returned in the hits array ... seem to have some kind of control
information specified in them and I'm not sure what the significance/
motivation behind that is:
If a search is performed with an index specified:
curl -XPOST http://localhost:9200/my_specific_index/_search/ -d '{
"fields": [
"upc"
],
"query": {
"match_all": {}
}
}'
Then I receive the correct # of total hits and 10 upc fields are
returned in the hits array as results.
BUT if the index is not specified then 3 out of the 10 entries that
are returned in the hits array ... seem to have some kind of control
information specified in them and I'm not sure what the significance/
motivation behind that is:
If a search is performed with an index specified:
curl -XPOST http://localhost:9200/my_specific_index/_search/ -d '{
"fields": [
"upc"
],
"query": {
"match_all": {}
}
}'
Then I receive the correct # of total hits and 10 upc fields are
returned in the hits array as results.
BUT if the index is not specified then 3 out of the 10 entries that
are returned in the hits array ... seem to have some kind of control
information specified in them and I'm not sure what the significance/
motivation behind that is:
A river needs to store some internals values. For couchDb, it's the last sequence number provided by the _change API.
More, a river need to store its settings : couchDb URL, port, ...
Es store this as Json documents in _river special index.
This way, river settings are "duplicated" in the cluster. If a node failed, a second one will continue the job.
If a search is performed with an index specified:
curl -XPOST http://localhost:9200/my_specific_index/_search/ -d '{
"fields": [
"upc"
],
"query": {
"match_all": {}
}
}'
Then I receive the correct # of total hits and 10 upc fields are
returned in the hits array as results.
BUT if the index is not specified then 3 out of the 10 entries that
are returned in the hits array ... seem to have some kind of control
information specified in them and I'm not sure what the significance/
motivation behind that is:
Le 6 mars 2012 à 19:26, David Pilato david@pilato.fr a écrit :
A river needs to store some internals values. For couchDb, it's the last sequence number provided by the _change API.
More, a river need to store its settings : couchDb URL, port, ...
Es store this as Json documents in _river special index.
This way, river settings are "duplicated" in the cluster. If a node failed, a second one will continue the job.
If a search is performed with an index specified:
curl -XPOST http://localhost:9200/my_specific_index/_search/ -d '{
"fields": [
"upc"
],
"query": {
"match_all": {}
}
}'
Then I receive the correct # of total hits and 10 upc fields are
returned in the hits array as results.
BUT if the index is not specified then 3 out of the 10 entries that
are returned in the hits array ... seem to have some kind of control
information specified in them and I'm not sure what the significance/
motivation behind that is:
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.