Getting things done with Elasticsearch - Devtalks 2015

Getting things done with Elasticsearch By Thijs Feryn

Explain in 1 slide

•Full-text search database •NoSQL database •Analytics engine •Written in Java
•Lucene based ( ~Solr) •Inverted indices •Easy to scale (~Elastic) •RESTFul interface (HTTP/JSON) •Schemaless •Real-time •ELK stack

Still with me?

Hi, I’m Thijs

I’m @ThijsFeryn on Twitter

I’m an at Evangelist

I’m a at board member

https://www.elastic.co/ downloads/elasticsearch

{ "status" : 200, "name" :
"War V", "cluster_name" : "elasticsearch", "version" : { "number" : "1.5.2", "build_hash" : "62ff9868b4c8a0c45860bebb259e21980778ab1c", "build_timestamp" : "2015-‐04-‐27T09:21:06Z", "build_snapshot" : false, "lucene_version" : "4.10.4" }, "tagline" : "You Know, for Search" } http://localhost:9200

RDBMS Elasticsearch Database Table Row Index Type Document

POST /my-‐index {"acknowledged":true} POST/my-‐index/my-‐type {
"key" : "value", "date" : "2015-‐05-‐10", "counter" : 1, "tags" : ["tag1","tag2","tag3"] } { "_index": "my-‐index", "_type": "my-‐type", "_id": "AU089olr9oI99a_rK9fi", "_version": 1, "created": true } CREATE DATABASE my-‐index INSERT INTO my-‐index.my-‐type (key,date,counter,tags) VALUES ("value","2015-‐05-‐10",1,"tag1,tag2,tag3") Confirmation

GET/my-‐index/my-‐type/AU089olr9oI99a_rK9fi?pretty { "_index": "my-‐index",
"_type": "my-‐type", "_id": "AU089olr9oI99a_rK9fi", "_version": 1, "found": true, "_source": { "key": "value", "date": "2015-‐05-‐10", "counter": 1, "tags": [ "tag1", "tag2", "tag3" ] } } Retrieve document by id Document & meta data SELECT * FROM my-‐index.my-‐type WHERE id="AU089olr9oI99a_rK9fi"

GET /my-‐index/_mapping?pretty { "my-‐index": {
"mappings": { "my-‐type": { "properties": { "counter": { "type": "long" }, "date": { "type": "date", "format": "dateOptionalTime" }, "key": { "type": "string" }, "tags": { "type": "string" } } } } } } Schemaless? Not really … “Guesses” mapping on insert

Explicit mapping … using actual data

Import data from MySQL to ElasticSearch

{ "category": [
"Default Category", "Men", "Shirts" ], "entity_id": "231", "created_at": "2013-‐03-‐05 06:48:12", "updated_at": "2013-‐03-‐05 09:27:15", "name": "French Cuff Cotton Twill Oxford", "price": "190.0000", "description": "Made with wrinkle resistant cotton twill, this French-‐cuffed luxury dress shirt is perfect for Business Class frequent flyers.", "sku": "msj000" } SELECT GROUP_CONCAT(distinct c.name) as category,p.entity_id, p.created_at, p.updated_at, p.name, p.price, p.short_description as description,p.sku FROM catalog_product_flat_1 p JOIN `catalog_category_product_index` AS `cat_index` ON cat_index.product_id=p.entity_id JOIN `catalog_category_flat_store_1` AS `c` ON cat_index.category_id=c.entity_id WHERE cat_index.store_id = 1 AND cat_index.visibility=1 GROUP BY p.name ORDER BY p.entity_id

POST /products { "mappings": {
"product" : { "_id" : { "path" : "entity_id" }, "properties" : { "entity_id" : {"type" : "integer"}, "name" : { "type" : "string", "index" : "not_analyzed", "fields" : { "raw" : { "type" : "string", "analyzer": "english" } } }, "description" : { "type" : "string", "index" : "not_analyzed", "fields" : { "raw" : { "type" : "string", "analyzer": "english" } } }, "price" : {"type" : "double"}, "sku" : {"type" : "string", "index" : "not_analyzed"}, "created_at" : {"type" : "date", "format" : "YYYY-‐MM-‐dd HH:mm:ss"}, "updated_at" : {"type" : "date", "format" : "YYYY-‐MM-‐dd HH:mm:ss"} , "category" : { "type" : "string", "index" : "not_analyzed" } } } } } Explicit mapping at index creation time

"entity_id" : {"type" : "integer"},
"name" : { "type" : "string", "index" : "not_analyzed", "fields" : { "raw" : { "type" : "string", "analyzer": "english" } } }, "description" : { "type" : "string", "index" : "not_analyzed", "fields" : { "raw" : { "type" : "string", "analyzer": "english" } } }, "price" : {"type" : "double"}, "sku" : {"type" : "string", "index" : "not_analyzed"}, "created_at" : {"type" : "date", "format" : "YYYY-‐MM-‐dd HH:mm:ss"}, "updated_at" : {"type" : "date", "format" : "YYYY-‐MM-‐dd HH:mm:ss"} , "category" : { "type" : "string", "index" : "not_analyzed" } Not analyzed? English analyzer?

Analyzed vs non-analyzed

Full-text vs exact value

Analyzer •Character filters •Tokenizers •Token filters Replaces characters for analyzed
text Break text down into terms Add/modify/ delete tokens

Built-in analyzers •Standard •Simple •Whitespace •Stop •Keyword •Pattern •Language •Snowball
•Custom Standard tokenizer Lowercase token filter English stop word token filter

Set the shape to semi-transparent by calling set_trans(5) set, the,
shape, to, semi, transparent, by, calling, set_trans, 5 set, the, shape, to, semi, transparent, by, calling, set, trans Set, the, shape, to, semi-transparent, by, calling, set_trans(5) set, shape, semi, transpar, call, set_tran, 5 Standard Simple Whitespace English

Building a custom analyzer PUT /my_index {
"settings": { "analysis": { "char_filter": { "&_to_and": { "type": "mapping", "mappings": [ "&=> and "] }}, "filter": { "my_stopwords": { "type": "stop", "stopwords": [ "the", "a" ] }}, "analyzer": { "my_analyzer": { "type": "custom", "char_filter": [ "html_strip", "&_to_and" ], "tokenizer": "standard", "filter": [ "lowercase", "my_stopwords" ] }} }}}

Search

GET /products/product/_search?pretty { "took": 2,
"timed_out": false, "_shards": { "total": 1, "successful": 1, "failed": 0 }, "hits": { "total": 192, "max_score": 1, "hits": [ { "_index": "products", "_type": "product", "_id": "231", "_score": 1, "_source": { "category": [ "Default Category", "Men", "Shirts" ], "entity_id": "231", "created_at": "2013-‐03-‐05 06:48:12", "updated_at": "2013-‐03-‐05 09:27:15", "name": "French Cuff Cotton Twill Oxford", "price": "190.0000", "description": "Made with wrinkle resistant cotton twill, this French-‐cuffed luxury dress shirt is perfect for Business Class frequent flyers.", "sku": "msj000" } }, … SELECT * FROM products.product

GET /products/product/_search?pretty POST /products/product/_search?pretty {
"query": { "match_all": {} } } Search “lite” vs full query DSL SELECT * FROM products.product

GET /products/product/_search? pretty&q=name.raw:Blazer POST /products/product/_search?pretty {
"query": { "match": { "name.raw": "Blazer" } } } SELECT * FROM products.product WHERE name LIKE "%Blazer%" Search “lite” vs full query DSL

SELECT * FROM products.product WHERE
name = "Linen Blazer" POST /products/product/_search? pretty { "query": { "match": { "name.raw": "Linen Blazer" } } } POST /products/product/_search? pretty { "query": { "filtered": { "query": { "match_all": {} }, "filter": { "term": { "name": "Linen Blazer" } } } } } Matches 2 products Matches 1 product SELECT * FROM products.product WHERE name LIKE "%Linen%" OR name LIKE "%Blazer%"

Filter vs Query

Filter •Does it match? Yes or no •When relevance doesn’t
matter •Faster & cacheable •For non-analyzed data Query •How well does it match? •For full-text search •On analyzed/tokenized data

Match Query Multi Match Query Bool Query Boosting Query Common
Terms Query Constant Score Query Dis Max Query Filtered Query Fuzzy Like This Query Fuzzy Like This Field Query Function Score Query Fuzzy Query GeoShape Query Has Child Query Has Parent Query Ids Query Indices Query Match All Query More Like This Query Nested Query Prefix Query Query String Query Simple Query String Query Range Query Regexp Query Span First Query Span Multi Term Query Span Near Query Span Not Query Span Or Query Span Term Query Term Query Terms Query Top Children Query Wildcard Query Minimum Should Match Multi Term Query Rewrite Template Query

And Filter Bool Filter Exists Filter Geo Bounding Box Filter
Geo Distance Filter Geo Distance Range Filter Geo Polygon Filter GeoShape Filter Geohash Cell Filter Has Child Filter Has Parent Filter Ids Filter Indices Filter Limit Filter Match All Filter Missing Filter Nested Filter Not Filter Or Filter Prefix Filter Query Filter Range Filter Regexp Filter Script Filter Term Filter Terms Filter Type Filter

Filter examples

POST /products/product/_search?pretty { "query": {
"filtered": { "filter": { "ids": { "values": [231,234,258] } } } } } SELECT * FROM products.product WHERE entity_id IN (231,234,258)

"filtered": { "filter": { "bool": { "must": [ { "range": { "price": { "gte": 100, "lte": 400 } } } ], "must_not": [ { "term": { "name": "Convertible Dress" } } ], "should": [ { "term": { "category": "Women" } }, { "term": { "category": "New Arrivals" } } ] } } } } } SELECT * FROM products.product WHERE price BETWEEN 100 AND 400 AND name != "Convertible Dress" AND (category = "Women" OR category = "New Arrivals")

"filtered": { "filter": { "prefix": { "name": "Slim" } } } } } SELECT * FROM products.product WHERE name LIKE "Slim%"

{ "filtered" : {
"filter" : { "geo_distance_range" : { "from" : "200km", "to" : "400km", "pin.location" : { "lat" : 40, "lon" : -‐70 } } } } } Requires “geo point” typed field

Relevance

{ "took": 1, "timed_out":
false, "_shards": { "total": 1, "successful": 1, "failed": 0 }, "hits": { "total": 2, "max_score": 3.3350093, "hits": [ { "_index": "products", "_type": "product", "_id": "243", "_score": 3.3350093, "fields": { "name": [ "Linen Blazer" ] } }, { "_index": "products", "_type": "product", "_id": "246", "_score": 0.5954286, "fields": { "name": [ "Stretch Cotton Blazer" ] } } ] } } POST /products/product/_search?pretty { "fields": ["name"], "query": { "match": { "name.raw": "Linen Blazer" } } } Hits both terms. More relevant

{ "took": 1, "timed_out":
false, "_shards": { "total": 1, "successful": 1, "failed": 0 }, "hits": { "total": 1, "max_score": 1, "hits": [ { "_index": "products", "_type": "product", "_id": "243", "_score": 1, "fields": { "name": [ "Linen Blazer" ] } } ] } } POST /products/product/_search?pretty { "query": { "filtered": { "query": { "match_all": {} }, "filter": { "term": { "name": "Linen Blazer" } } } } } No relevance on filters Score is always 1

POST /products/product/_search?pretty { "fields": ["name","description"],
"query": { "bool": { "must": [ { "match": { "description.raw": "suit" } } ], "should": [ { "match": { "name.raw": "coat" } } ] } } } Only search for “suit” Increase relevance if name contains “coat”

POST /products/product/_search?pretty { "fields": ["price","name","description","category"],
"query": { "bool": { "must": [ { "filtered": { "filter": { "bool": { "must": [ { "range": { "price": { "gte": 100, "lte": 400 } } }, { "term": { "category": "Men" } } ] } } } } ], "should": [ { "match": { "category": "New Arrivals" } } ] } } } Increase relevance Combined filters

"bool": { "must": [ { "match_all": {} } ], "should": [ { "match": { "description.raw": { "query": "comfort", "boost": 2 } } }, { "match": { "name.raw": { "query": "tee", "boost": 3 } } } ] } } } Increase relevance Query- time boosting

Multi index multi type

SELECT * FROM *.tbl_a* WHERE
key LIKE ‘%val%’ •Cross-database queries •Wildcard database queries •Wildcard table queries Not supported in SQL

/_search /products/_search /products/product/_search /products,clients/_search /pro*/_search
/pro*,cli*/_search /products/product,invoice/_search /products/pro*/_search /_all/product/_search /_all/product,invoice/_search /_all/pro*/_search

Multi “all the things”

Aggregations

Group by on steroids

SELECT COUNT(category) FROM products.product GROUP BY category
Aggregations in SQL Metric Bucket

SELECT COUNT(category) FROM products.product GROUP BY category
POST /products/product/_search? pretty&search_type=count { "aggs": { "number_of_categories" : { "cardinality": { "field": "category" } } } } Metric Bucket Only aggs, no docs

POST /products/product/_search?pretty { "fields": ["category","price","name"],
"query": { "match": { "name.raw": "blazer" } }, "aggs": { "avg_price": { "avg": { "field": "price" } }, "min_price" : { "min": { "field": "price" } }, "max_price" : { "max": { "field": "price" } }, "number_of_products_per_category" : { "terms": { "field": "category", "size": 10 } } } } Multi-group by besides query

"aggregations": {
"min_price": { "value": 455 }, "number_of_products_per_category": { "doc_count_error_upper_bound": 0, "sum_other_doc_count": 0, "buckets": [ { "key": "Blazers", "doc_count": 2 }, { "key": "Default Category", "doc_count": 2 }, { "key": "Men", "doc_count": 2 } ] }, "max_price": { "value": 490 }, "avg_price": { "value": 472.5 } } Aggregation output Drill down search

Min Aggregation Max Aggregation Sum Aggregation Avg Aggregation Stats Aggregation
Extended Stats Aggregation Value Count Aggregation Percentiles Aggregation Percentile Ranks Aggregation Cardinality Aggregation Geo Bounds Aggregation Top hits Aggregation Scripted Metric Aggregation Global Aggregation Filter Aggregation Filters Aggregation Missing Aggregation Nested Aggregation Reverse nested Aggregation Children Aggregation Terms Aggregation Significant Terms Aggregation Range Aggregation Date Range Aggregation IPv4 Range Aggregation Histogram Aggregation Date Histogram Aggregation Geo Distance Aggregation GeoHash grid Aggregation

Managing Elasticsearch

Plenty of ways … for which we don’t have enough
time

Clustering

Single node 2 node cluster 3 node cluster

Example config settings node.rack: my-‐location node.master: true
node.data: true http.enabled: true cluster.name: my-‐cluster node.name: my-‐node index.number_of_shards: 5 index.number_of_replicas: 1 discovery.zen.minimum_master_nodes: 2

GET /_cat

GET /_cat =^.^= /_cat/allocation /_cat/shards /_cat/shards/{index}
/_cat/master /_cat/nodes /_cat/indices /_cat/indices/{index} /_cat/segments /_cat/segments/{index} /_cat/count /_cat/count/{index} /_cat/recovery /_cat/recovery/{index} /_cat/health /_cat/pending_tasks /_cat/aliases /_cat/aliases/{alias} /_cat/thread_pool /_cat/plugins /_cat/fielddata /_cat/fielddata/{fields} Non-JSON output

GET /_cat/shards?v index shard prirep state
docs store ip node my-‐index 2 r STARTED 6 7.2kb 192.168.10.142 node3 my-‐index 2 p STARTED 6 9.5kb 192.168.10.142 node2 my-‐index 0 p STARTED 4 7.1kb 192.168.10.142 node3 my-‐index 0 r STARTED 4 4.8kb 192.168.10.142 node2 my-‐index 3 r STARTED 5 7.1kb 192.168.10.142 node1 my-‐index 3 p STARTED 5 7.2kb 192.168.10.142 node3 my-‐index 1 p STARTED 1 2.4kb 192.168.10.142 node1 my-‐index 1 r STARTED 1 2.4kb 192.168.10.142 node2 my-‐index 4 p STARTED 5 9.5kb 192.168.10.142 node1 my-‐index 4 r STARTED 5 9.4kb 192.168.10.142 node3 5 shards & a single replica by default

GET /_cat/health? v&h=cluster,status,node.total,shards,pri,unassign,init cluster status node.total shards
pri unassign init mycluster green 3 12 6 0 0 Cluster health

The ELK stack

Logs Parse & ship Store Visualize

Integrating Elasticsearch

It’s REST, deal with it! Directly from Javascript? Don’t forget
to “restrict” access AJAX #FTW Behind a proxy No backend code

Or just use a backend API PHP Java Perl Python
Ruby .NET

Getting things done with Elasticsearch - Devtal...

Getting things done with Elasticsearch - Devtalks 2015

More Decks by Thijs Feryn

Other Decks in Technology

Featured

Transcript