Product Selector

Fusion 5.11
    Fusion 5.11

    HDFS V1 Connector Configuration Reference

    Table of Contents

    Hadoop Distributed File System (HDFS). It traverses the Hadoop file system as it would a regular Unix filesystem.

    See also the Hadoop connectors, which use MapReduce to distribute the crawl processes.

    When there is a lot of content to process, these MapReduce-enabled connectors will be significantly faster.

    Configuration

    When entering configuration values in the UI, use unescaped characters, such as \t for the tab character. When entering configuration values in the API, use escaped characters, such as \\t for the tab character.

    Connector for HDFS or similar Hadoop filesystems.

    id - stringrequired

    Unique name for this datasource.

    >= 1 characters

    Match pattern: ^[a-zA-Z0-9_-]+$

    pipeline - stringrequired

    Name of an existing index pipeline for processing documents.

    >= 1 characters

    description - string

    Optional description for this datasource.

    parserId - string

    Parser used when parsing raw content. For some connectors, a configuration to 'retry' parsing if an error occurs is available as an advanced setting

    properties - Properties

    Datasource configuration properties

    db - Connector DB

    Type and properties for a ConnectorDB implementation to use with this datasource.

    type - string

    Fully qualified class name of ConnectorDb implementation.

    >= 1 characters

    Default: com.lucidworks.connectors.db.impl.MapDbConnectorDb

    inlinks - boolean

    Keep track of incoming links. This negatively impacts performance and size of DB.

    Default: false

    aliases - boolean

    Keep track of original URI-s that resolved to the current URI. This negatively impacts performance and size of DB.

    Default: false

    inv_aliases - boolean

    Keep track of target URI-s that the current URI resolves to. This negatively impacts performance and size of DB.

    Default: false

    url - string

    The start link is the URI for the HDFS connector. The `namenode` is 8020 by default, e.g., hdfs://hmbrnasnfs2.highmark.com:8020/datalake/hm/phi/refined/UnstructuredData/NEPA

    >= 1 characters

    Match pattern: .*:.*

    max_docs - integer

    Maximum number of documents to fetch. The default (-1) means no limit.

    >= -1

    exclusiveMinimum: false

    Default: -1

    max_bytes - integer

    Maximum size (in bytes) of documents to fetch or -1 for unlimited file size.

    >= -1

    exclusiveMinimum: false

    Default: 10485760

    index_directories - boolean

    Set to true to add directories to the index as documents. If set to false, directories will not be added to the index, but they will still be traversed for documents.

    Default: false

    max_threads - integer

    The maximum number of threads to use for fetching data. Note: Each thread will create a new connection to the repository, which may make overall throughput faster, but this also requires more system resources, including CPU and memory.

    Default: 1

    add_failed_docs - boolean

    Set to true to add documents even if they partially fail processing. Failed documents will be added with as much metadata as available, but may not include all expected fields.

    Default: false

    crawl_item_timeout - integer

    Time in milliseconds to fetch any individual document.

    exclusiveMinimum: true

    Default: 600000

    maximum_connections - integer

    Maximum number of concurrent connections to the filesystem. A large number of documents could cause a large number of simultaneous connections to the repository and lead to errors or degraded performance. In some cases, reducing this number may help performance issues.

    Default: 1000

    initial_mapping - Initial field mapping

    Provides mapping of fields before documents are sent to an index pipeline.

    skip - boolean

    Set to true to skip this stage.

    Default: false

    label - string

    A unique label for this stage.

    <= 255 characters

    condition - string

    Define a conditional script that must result in true or false. This can be used to determine if the stage should process or not.

    reservedFieldsMappingAllowed - boolean

    Default: false

    retentionMappings - array[object]

    Fields that should be kept or deleted

    Default:

    object attributes:{field required : {
     display name: Field
     type: string
    }
    operation : {
     display name: Operation
     type: string
    }
    }

    updateMappings - array[object]

    Values that should be added to or set on a field. When a value is added, any values previously on the field will be retained. When a value is set, any values previously on the field will be overwritten.

    Default:

    object attributes:{field required : {
     display name: Field
     type: string
    }
    value required : {
     display name: Value
     type: string
    }
    operation : {
     display name: Operation
     type: string
    }
    }

    translationMappings - array[object]

    Fields that should be moved or copied to another field. When a field is moved, the values from the source field are moved over to the target field and the source field is removed. When a field is copied, the values from the source field are copied over to the target field and the source field is retained.

    Default: {"source":"fetch_time","target":"fetch_time_dt","operation":"move"}{"source":"ds:description","target":"description","operation":"move"}

    object attributes:{source required : {
     display name: Source Field
     type: string
    }
    target required : {
     display name: Target Field
     type: string
    }
    operation : {
     display name: Operation
     type: string
    }
    }

    unmappedRule - Unmapped Fields

    Fields not mapped by the above rules. By default, any remaining fields will be kept on the document.

    keep - boolean

    Keep all unmapped fields

    Default: true

    delete - boolean

    Delete all unmapped fields

    Default: false

    fieldToMoveValuesTo - string

    Move all unmapped field values to this field

    fieldToCopyValuesTo - string

    Copy all unmapped field values to this field

    valueToAddToUnmappedFields - string

    Add this value to all unmapped fields

    valueToSetOnUnmappedFields - string

    Set this value on all unmapped fields

    converter - string

    Fully-qualified classname for a custom converter to produce valid SolrInputDocuments extracted from Hadoop Sequence or MapReduce files.

    with_kerberos - boolean

    Set to true to use Kerberos to authenticate to HDFS for access to the content.

    Default: false

    kerberos_user - string

    Kerberos principal name, i.e., 'username@YOUR-REALM.COM'.

    kerberos_keytab - string

    Full path to the Kerberos keytab file.

    crawl_depth - integer

    Number of levels in a directory or site tree to descend for documents.

    >= -1

    exclusiveMinimum: false

    Default: -1

    bounds - string

    Limits the crawl to a specific directory sub-tree, hostname or domain.

    Default: tree

    Allowed values: treehostdomainnone

    include_paths - array[string]

    Regular expressions for URI patterns to include. This will limit this datasource to only URIs that match the regular expression.

    exclude_paths - array[string]

    Regular expressions for URI patterns to exclude. This will limit this datasource to only URIs that do not match the regular expression.

    include_extensions - array[string]

    List the file extensions to be fetched. Note: Files with possible matching MIME types but non-matching file extensions will be skipped. Extensions should be listed without periods, using whitespace to separate items (e.g., 'pdf zip').

    commit_on_finish - boolean

    Set to true for a request to be sent to Solr after the last batch has been fetched to commit the documents to the index.

    Default: true

    verify_access - boolean

    Set to true to require successful connection to the filesystem before saving this datasource.

    Default: true