A typical request to a software system is handled by multiple subsystems, many of which may be distinct servers residing on distinct hosts and locations. Tracing the request flow on distributed systems can be challenging, as log messages are scattered across various systems and intermingled with messages for other requests. To make this easier, a correlation ID can be assigned to a request, which is then added to every associated operation as the request flows through the larger system. The correlation ID allows related log messages to be easily located and grouped. The server supports correlation IDs for all HTTP requests received through its HTTP(S) Connection Handler.
When a HTTP request is received, it is automatically assigned a correlation ID. This ID
can be used to correlate HTTP responses with messages recorded to the HTTP Detailed
Operation log and the trace log. For specific web APIs, the correlation ID may also be
passed to the LDAP subsystem. For the SCIM 1, Delegated Admin, Consent, and Directory
REST APIs, the correlation ID will also appear with associated requests in LDAP logs in
correlationID key. The correlation ID is also used as the default
client request ID value in Intermediate Client Request Controls used by the SCIM 2,
Consent, and Directory REST APIs. Values related to the Intermediate Client Request
Control appear in the LDAP logs in the
via key, and are forwarded to
downstream LDAP servers when received by the PingDirectoryProxy server. The correlation
ID header is also added to requests forwarded by the PingDataGovernance gateway.
For Server SDK extensions that have access to the current HttpServletRequest, the current
correlation ID can be retrieved as a string through the
com.pingidentity.pingdata.correlation_id attribute. For example: