Elasticsearch reindex tool

Elasticsearch reindex tool provides easy way to rebuild indexes, it is also possible to move indexes between clusters. Multiple threads are used in order to query(read) and index(write) data. In order to do that efficiently, scan and scroll is used to retrieve batches of documents from the old index and then the bulk API to push them into new one.

Elasticsearch version compability

Master branch is compatible with version 2.x

If you wish to use it with version 1.x please checkout branch 1.x

Why another reindex tool?

Our idea was to speed up index rebuilding. To decrease time of reindexing, our tool reads data from old index and writes it to the new one in parallel using multiple threads. To make it possible, each thread reads piece of data from the index based on a chosen field and its values.

Currently tool supports double type and string type fields. For double field type queries are spread into segments with given list of thresholds, for string type fields with given prefixes list.

In the future we plan provide more segmentation strategies.

There are similar tools:

Requirements

Usage

First create a package:

./gradlew jar

Example of reindex:

REMEMBER: use elasticsearch binary transport port (by default 9300), not the one used for rest communication (by default 9200)

Without segmentation:

./run.sh -s http://host:9300/index/type -t http://host1:9300/index1/type1 -sc cluster_name -tc cluster_name1

With segmentation by double field:

./run.sh -s http://host:9300/index/type -t http://host1:9300/index1/type1 -sc cluster_name -tc cluster_name1

./run.sh -s http://host:9300/index/type -t http://host1:9300/index1/type1 -sc cluster_name -tc cluster_name1 -segmentationField rate.newCoolness -segmentationThresholds 0.0,0.5,0.59,0.6,0.7,0.9,1.0

Index querying will divide data into segments based on rate.newCoolness field: (0.0-0.5] (0.5-0.59] (0.59-0.6] (0 .6-0.7] (0.7-0.9],(0.9-1.0]

With segmentation by prefix on string field:

./run.sh -s http://host:9300/index/type -t http://host1:9300/index1/type1 -sc cluster_name -tc cluster_name1 -segmentationField userId -segmentationPrefixes 1,2,3,4,5,6,7

In this example index querying will divide data into segments based on the first character of the userId field: 1,2,3,4,5,6,7

With query option:

./run.sh -s http://host:9300/index/type -t http://host1:9300/index1/type1 -sc cluster_name -tc cluster_name1 -query "{\"range\": {\"timestamp\" : {\"gte\" : \"2015-12-08T07:03:00.216Z\"}}}" -sort _timestamp -sortOrder DESC

In this example index querying will be filtered with query and reindex will take place ordered by sort field and sortOrder

With target index name pattern

Use the actual source-index of the document (useful when you use wildcards for the source-index): ./run.sh -s http://host:9300/*/type -t http://host1:9300/${_index}/type1 -sc cluster_name -tc cluster_name1

Use a field from the document: ./run.sh -s http://host:9300/index/type -t http://host1:9300/${customer}/type1 -sc cluster_name -tc cluster_name1

Use a time-field from the document and apply a format: ./run.sh -s http://host:9300/index/type -t http://host1:9300/${startTime:yyyy-MM-dd}/type1 -sc cluster_name -tc cluster_name1

Options:

-s, source
   Source f.e. http://localhost:9300/source_index/type
-sc, source-cluster
   Source cluster name
-t, target
   Target f.e. http://localhost:9300/target_index/type
-tc, target-cluster
   Target cluster name
-disable-cluster-sniffing
   Don't try to determine additional cluster nodes (e.g. when your network
   only allows access to one of the nodes)
   Default: false
-segmentationField
   Segmentation field
-segmentationPrefixes
   Segmentation prefixes (comma-separated)
-segmentationThresholds
   Segmentation thresholds (only double type)
-query
   Give a query to filter data 
-sort
   Give field to sort on (if query option in use)
-sortOrder
   Give sortOrder  (if query option in use)    

segmentationField, segmentationThreshold and segmentationPrefixes are optional parameters, allowing to spread querying for field with double values or prefix for string field

disable-cluster-sniffing allows to work in cases where the network-setup makes it impossible to connect to all nodes of the source or target cluster. Note that it may lead to slightly reduced reindexing rates as data can only be sent via one node then.

During reindex process progress message is prompted after each scroll query.

Example of progress message with the time how long it lasts, number of items queried and indexed, occupancy of queue, number of concurrent reader threads and number of failed document indexing:

11:24:59.567 [pool-1-thread-1] INFO c.y.e.t.r.s.ProcessStatistics - PT11M43.346S items: 3572086 / 3580842 (10 1) failed=0

License

Elasticsearch reindex tool is published under Apache License 2.0.