Just like using auto-increment primary keys in a traditional database, sometimes we would like to uniquely identify a document in Solr. In this case, we can use something called
UUID. Although it is not a perfect solution, giving the extremely low probability of a duplicated UUID we can reasonably believe that it is safe. Here is how to configure Solr to use UUID:
1. Change the field for id in conf/schema.xml to the following line
2. Add the uuid field type
3. Add the following lines to solrconfig.xml. (We can put it below a comment block started with Update Processors )
id
4. We can then put a referece to that processor in the handlers that upload a document. Like this:
dedupe
uuid
