如何从Lucidworks Fusion迁移配置/集合至纯Solr？许可证过期替代方案

Absolutely—pure Apache Solr is fully capable of handling the CSV parsing and indexing workflow you were doing with Lucidworks Fusion, and it’s straightforward to set up. Here’s a step-by-step breakdown to get you up and running:

1. Install & Start Pure Solr #

Grab the latest stable Apache Solr release, extract it, and fire it up with the default command:

bin/solr start

This will launch Solr on the default port (8983) with all core services ready to go.

2. Create a New Collection #

First, make a collection to store your documents (equivalent to Fusion’s collection):

bin/solr create -c your_target_collection

Replace your_target_collection with a name that makes sense for your dataset.

3. Configure the Schema for _s & _t Fields #

Fusion automatically creates _s (string, exact-match) and _t (text, analyzed) field variants, but in pure Solr we’ll need to set these up explicitly. You have two flexible options:

Option A: Dynamic Fields (Recommended for Flexibility) #

If you want Solr to automatically handle any field ending in _s or _t, add dynamic field definitions via the Solr Schema API (no file editing required):

curl -X POST -H 'Content-type:application/json' --data-binary '{
  "add-dynamic-field": [
    {"name":"*_s", "type":"string", "indexed":true, "stored":true},
    {"name":"*_t", "type":"text_general", "indexed":true, "stored":true}
  ]
}' http://localhost:8983/solr/your_target_collection/schema

This way, any CSV column named like product_name_s or description_t will be mapped correctly without defining each field individually.

Option B: Copy Fields (If Your CSV Uses Plain Column Names) #

If your CSV has columns without the _s/_t suffix (e.g., product_name instead of product_name_s), define base fields and use copy fields to populate both variants:

curl -X POST -H 'Content-type:application/json' --data-binary '{
  "add-field": {"name":"product_name", "type":"text_general", "indexed":true, "stored":true},
  "add-copy-field": [
    {"source":"product_name", "dest":"product_name_s"},
    {"source":"product_name", "dest":"product_name_t"}
  ]
}' http://localhost:8983/solr/your_target_collection/schema

Repeat this for each column in your CSV, or combine with dynamic fields for the copied variants to save time.

4. Index Your CSV File #

Solr has a built-in post tool that handles CSV files seamlessly. Use this command to index your data:

bin/post -c your_target_collection your_data.csv

If your CSV uses a non-standard separator (like ; instead of ,), add the separator parameter:

bin/post -c your_target_collection -params "separator=;" your_data.csv

If you used copy fields (Option B), the post tool will automatically populate both _s and _t variants from your base columns.

5. Verify Your Indexed Documents #

Head to the Solr Admin UI at http://localhost:8983/solr, select your collection, and use the Query tab to run a simple search (e.g., *:*). You’ll see that each row from your CSV is now a document, with both _s and _t fields populated exactly as they were in Fusion.

Pure Solr has all the core indexing capabilities Fusion leverages under the hood, so this workflow will work just as well for your basic needs. If you ever need the advanced orchestration or UI tools Fusion offers later, you can always circle back—but for now, pure Solr is a perfect, free solution.

内容的提问来源于stack exchange，提问作者Mark Miller