Unsupported features for DSE Search

Unsupported Apache Cassandra, Apache Solr, and other features.

Unsupported features include Apache Cassandra™ and Apache Solr™ features.

Apache Solr™ and Apache Lucene® limitations

Apache Solr and Lucene limitations apply to DSE Search. For example:
  • The 2 billion records per node limitation as described in Lucene limitations.
  • The 1024 maxBoolean clause limit in SOLR-4586.
  • Solr field name policy applies to the indexed field names:
    • Every field must have a name.
    • Field names must consist of alphanumeric or underscore characters only.
    • Fields cannot start with a digit.
    • Names with both leading and trailing underscores (for example, _version_) are reserved.
    Note: Non-compliant field names are not supported from all components. Backward compatibility is not guaranteed.
  • Limitations and known Apache Solr issues apply to DSE Search queries. For example: incorrect SORT results for tokenized text fields.

Unsupported Apache Cassandra features

These imitations apply to DSE Search:

  • Column aliases are not supported in solr_query queries.
  • Continuous paging.
  • Static columns
  • Compound primary keys for COMPACT STORAGE tables
  • Counter columns
  • Super columns
  • Thrift-compatible tables with column comparators other than UTF-8 or ASCII.
  • PER PARTITION clause is not supported for DSE Search solr_query queries.
  • Indexing frozen maps is not supported. However, indexing frozen sets and lists of native and user-defined (tuple/UDT) element types is supported.
  • Using DSE Search with newly created COMPACT STORAGE tables is deprecated.

Unsupported Apache Solr™ features

These imitations apply to DSE Search:
  • DSE Search does not support Solr Managed Resources.
  • Solr schema fields that are both dynamic and multiValued only for CQL-based search indexes.
  • The deprecated replaceFields request parameters on document updates for CQL-based search indexes. Instead, use the suggested procedure for inserting/updating data.
  • Block joins based on the Lucene BlockJoinQuery in search indexes and CQL tables.
  • Schemaless mode.
  • Partial schema updates through the REST API after search indexes are changed.

    For example, to update individual fields of a schema using the REST API to add a new field to a schema, you must change the schema.xml file, upload it again, and reload the core (same for copy fields).

  • org.apache.solr.spelling.IndexBasedSpellChecker and org.apache.solr.spelling.FileBasedSpellChecker

    Instead use org.apache.solr.spelling.DirectSolrSpellChecker for spell checking.

  • The commitWithin parameter.
  • The SolrCloud CloudSolrServer feature of SolrJ for endpoint discovery and round-robin load balancing.
  • The DSE Search configurable SolrFilterCache does not support auto-warming.
  • DSE Search does not support the duration Cassandra data type.
  • SELECT statements with DISTINCT are not supported with solr_query.
  • UnInvertedFieldRealtime field value cache.
  • GetReplicationHandler: Store & Restore.
  • useDocValuesAsStored in schema fields and as a query request parameter.
  • Graph queriesSolr SQLStreaming aggregations.
  • Data import handler.
  • Tuple/UDT subfield sorting and faceting.
  • The dataDir parameter in solrconfig.xml.

Deprecated Solr features

The Tika functionality that is bundled with Apache Solr is deprecated. Instead, use the stand-alone Apache Tika project.

Other unsupported features

  • JBOD mode.
  • The Solr updatelog is not supported in DSE Search.

    The commit log replaces the Solr updatelog. Consequently, features that require the updateLog are not supported. Instead of using atomic updates, partial document updates are available by running the update with CQL.

  • CQL Solr queries do not support native functions or column aliases as selectors.
  • RamDirectoryFactory or other non-persistent DirectoryFactory implementations.
  • Tuple and UDT limitations apply.