Schema.xml for new collection in CloudEra search

[#1]

Hi,

I have installed cloudera search using RPMs on Cent OS and able to work on flume NRT as well. I had to add a  new fileld to the existing collection in order to support the same while indexing the data. I want to know, how and from where the schema.xml is picked up for a collection.? and how can we change the schema after creating the collection for the first time. I have tried changing schema.xml in various places, but nothing seems to be working..

[#2 · Accepted solution]

Hey Srini,

When you first created the collection, you ran a command similar to:

- solrctl instancedir --generate $HOME/solr_configs

This command was what created the configs that you imported for solr.  You can see the various configs by running:

- solrctl instancedir --list

Then pick the config that goes with your collection and run:

- solrctl instancedir --get <name_from_list> /path/to/local_fs

Then change the schema.xml in /path/to/local_fs/conf and run:

- solrctl instancedir --update <name_from_list> /path/to/local_fs

Then the schema will get updated.  Note, when you update the schema, you have to reindex all your documents or else they won't have indexes for the latest schema changes.

Hope this helps.

Thanks

Chris

[#3 · Accepted solution]

Hey Srini,

When you first created the collection, you ran a command similar to:

- solrctl instancedir --generate $HOME/solr_configs

This command was what created the configs that you imported for solr.  You can see the various configs by running:

- solrctl instancedir --list

Then pick the config that goes with your collection and run:

- solrctl instancedir --get <name_from_list> /path/to/local_fs

Then change the schema.xml in /path/to/local_fs/conf and run:

- solrctl instancedir --update <name_from_list> /path/to/local_fs

Then the schema will get updated.  Note, when you update the schema, you have to reindex all your documents or else they won't have indexes for the latest schema changes.

Hope this helps.

Thanks

Chris

[#4]

Thanks Chris.. I understood the process when i spent more time on it. I have tried the same and it did work.. Probably i might have asked question little too early.. 🙂 Thanks for your resonse...

[#5 · Accepted solution]

Wrt. morphline commands, the Solr schema is fetched from wherever the SolrLocator is pointing to: http://cloudera.github.io/cdk/docs/current/cdk-morphlines/morphlinesReferenceGuide.html#solrLocator

The Solr schema can be managed via the solrctrl CLI: http://www.cloudera.com/content/cloudera-content/cloudera-docs/Search/latest/Cloudera-Search-User-Gu...

For more on installing and managing Cloudera Search see http://www.cloudera.com/content/cloudera-content/cloudera-docs/Search/latest/Cloudera-Search-Install...