How exactly Elastic search stores the data?


(stalin3443) #1

Hi All,

I'm newbie to Elastic search from RDBMS background.

In Relational databases, the indexing part stores only indexed column data,
and physical address of the exact block on the disk.

while coming to ES, if i have indexed MongoDB data using ES, What exactly
ES will store ?

does it store only few indexed columns ? or does it store entire data ?

please do the needful ..

Thanks,
Stalin

--
You received this message because you are subscribed to the Google Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.


(Ivan Brusic) #2

Everything is stored in a Lucene index, with each shard being its own
index. Lucene indices are append-only segments that are constantly created
and merged.

By default, each field is indexed in elasticsearch, but this behavior can
be changed via the mapping. You can index every field and you also can
index no fields (with only a few select fields being explicitly index such
as the id and type).

Ivan

On Wed, Oct 23, 2013 at 10:51 PM, stalin3443@gmail.com wrote:

Hi All,

I'm newbie to Elastic search from RDBMS background.

In Relational databases, the indexing part stores only indexed column
data, and physical address of the exact block on the disk.

while coming to ES, if i have indexed MongoDB data using ES, What exactly
ES will store ?

does it store only few indexed columns ? or does it store entire data ?

please do the needful ..

Thanks,
Stalin

--
You received this message because you are subscribed to the Google Groups
"elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an
email to elasticsearch+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

--
You received this message because you are subscribed to the Google Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.


(stalin3443) #3

Thanks Ivan for your quick response ..

that means i can customize in such a way that, i can have only index
columns in ES, rest of the data in another Data store, which will be
pointed from ES.

lets assume, if i have my data in MongoDB, i have indexed the Data using ES
with Default indexing, now i will have entire data in ES.

even if i drop all the data from MongoDb, still i can have all the data in
ES.

correct me if i am wrong ?

Thanks,
Stalin

On Thursday, October 24, 2013 11:43:50 AM UTC+5:30, Ivan Brusic wrote:

Everything is stored in a Lucene index, with each shard being its own
index. Lucene indices are append-only segments that are constantly created
and merged.

By default, each field is indexed in elasticsearch, but this behavior can
be changed via the mapping. You can index every field and you also can
index no fields (with only a few select fields being explicitly index such
as the id and type).

Ivan

On Wed, Oct 23, 2013 at 10:51 PM, <stali...@gmail.com <javascript:>>wrote:

Hi All,

I'm newbie to Elastic search from RDBMS background.

In Relational databases, the indexing part stores only indexed column
data, and physical address of the exact block on the disk.

while coming to ES, if i have indexed MongoDB data using ES, What
exactly ES will store ?

does it store only few indexed columns ? or does it store entire data ?

please do the needful ..

Thanks,
Stalin

--
You received this message because you are subscribed to the Google Groups
"elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an
email to elasticsearc...@googlegroups.com <javascript:>.
For more options, visit https://groups.google.com/groups/opt_out.

--
You received this message because you are subscribed to the Google Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.


(Ivan Brusic) #4

Lucene has two concepts: indexing and storing. An indexed field means that
is it is searchable using queries and filters. A stored field is one that
is returned in the data. A field can be either, both or neither (which does
not make much sense, but is possible). The term indexing can take on a few
meanings, but in general indexing a document means storing the document and
indexing a few/all fields.

Elasticsearch is very customizable. By default, each field is indexed, but
not stored (retrievable), however the entire document is stored as-is (as
compressed JSON). So you can store your data without actually "indexing"
any fields. Play around with the examples on the readme (
https://github.com/elasticsearch/elasticsearch) and try creating your own
mapping.

Ivan

On Wed, Oct 23, 2013 at 11:31 PM, stalin3443@gmail.com wrote:

Thanks Ivan for your quick response ..

that means i can customize in such a way that, i can have only index
columns in ES, rest of the data in another Data store, which will be
pointed from ES.

lets assume, if i have my data in MongoDB, i have indexed the Data using
ES with Default indexing, now i will have entire data in ES.

even if i drop all the data from MongoDb, still i can have all the data in
ES.

correct me if i am wrong ?

Thanks,
Stalin

On Thursday, October 24, 2013 11:43:50 AM UTC+5:30, Ivan Brusic wrote:

Everything is stored in a Lucene index, with each shard being its own
index. Lucene indices are append-only segments that are constantly created
and merged.

By default, each field is indexed in elasticsearch, but this behavior can
be changed via the mapping. You can index every field and you also can
index no fields (with only a few select fields being explicitly index such
as the id and type).

Ivan

On Wed, Oct 23, 2013 at 10:51 PM, stali...@gmail.com wrote:

Hi All,

I'm newbie to Elastic search from RDBMS background.

In Relational databases, the indexing part stores only indexed column
data, and physical address of the exact block on the disk.

while coming to ES, if i have indexed MongoDB data using ES, What
exactly ES will store ?

does it store only few indexed columns ? or does it store entire data ?

please do the needful ..

Thanks,
Stalin

--
You received this message because you are subscribed to the Google
Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send
an email to elasticsearc...@**googlegroups.com.

For more options, visit https://groups.google.com/**groups/opt_outhttps://groups.google.com/groups/opt_out
.

--
You received this message because you are subscribed to the Google Groups
"elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an
email to elasticsearch+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

--
You received this message because you are subscribed to the Google Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.


(system) #5