Previous Topic: DYN Schema ExtensionsNext Topic: DYN Class Name Selection


DYN Attribute Name Selection

When manually assigning attributes for a DYN endpoint type (rather than using Connector Xpress), eTDYN-str-multi- is usually the best choice as you are less likely to run out of attributes.

If caching is required, we recommend that you use eTDYN-str-multi-ca, as the Provisioning Server does not typically cache the attribute values in the DYN namespace (except for a set of well-known attributes, for example, eTDYNConnectionURL, eTDYNHost, and such). Use a –ca- variant for cases requiring caching, for example, extra connection-related attributes.

For attributes in classes other than accounts and account templates, and non-capability attributes on these classes, the fact that these underlying LDAP attributes are multi-valued and strings is not important, as the metadata specified for them controls whether they accept multiple values and their real type.

For more information about available classes and attributes you may choose to mention in your metadata see DYN Schema Extensions.

For capability attributes on accounts and account templates where policy merging by the provisioning server comes into play, the distinction between single verses multi-valued attributes, real type, and case-sensitivity become important at the LDAP level. We recommend that you consider:

Note: Where an attribute id matches one of the regexes found in sensitiveAttrIdRegexes property in server_jcs.xml, the attributes are automatically treated as a sensitive attribute and are obscured in logging output (even when logging is turned on at the lowest levels of the ApacheDS code). The substrings password, pwd, and cred are defined to trigger this behavior by default. A good practice is to use attribute names with higher numbered suffixes for such sensitive attributes, allowing them to be excluded from logs across all connectors without negative impact.