Troubleshooting and Reference Guide : Using DRException Class
 
Share this page                  
Using DRException Class
 
DataFlow Exceptions
DataFlow Exceptions
All of the run time exceptions that DataFlow defines extend the DRException class found in the DataFlow commons package.
Exception
Description
Error indicating premature termination of a dataflow. A BrokenFlowError is raised when an output port is closed without EOD (End Of Data) having been sent. This is an irrecoverable event.
DaemonShutdownException
Exception indicating the requested action could not be performed due to cluster shutdown.
Unchecked exception issued to indicate that an input flow contains rows that are out of order.
Wrapper for errors occurring during the execution of a dataflow graph. Each error is recorded along with the path to the dataflow process which issued it.
Exception issued when an error relating to reaching the end of a flow is encountered. This may occur when:
You attempt to stop an input port which has already reached the EOD.
You attempt to push a token onto an output port to which you have already pushed EOD.
The application terminates but you have not read the EOD from all input ports.
The application terminates but you have not pushed EOD onto all output ports.
Exception indicating end-of-file has been unexpectedly reached on a stream.
Exception indicating a field already exists in a field namespace such as a record.
Unchecked exception issued to indicate a field in a record schema is of an invalid type for a given context. Many operators that accept record types place constraints on the schema, typically constraining them to be flat (that is, not nested) or have fields of homogeneous type. This exception will generally be seen as the cause of a TypeNotSupportedException raised by such an operator.
Unchecked exception issued when an attempt is made to compare two fields of incompatible types in a record schema. The fields and types are indicated as is the causing TypesNotComparableException. Note that comparison of record types is done by comparing schemas positionally. That is, fields are compared by type in the order in which they appear in the schema. This exception is issued during this comparison process and is generally seen as the cause of a TypesNotComparableException.
Exception indicating the cancellation of a running dataflow graph. If a running graph is canceled using LogicalGraphInstance.abort(), this exception will be raised.
Exception indicating that an aggregator was constructed with an invalid number of arguments.
Exception indicating reference to an invalid field (or fields) in a record structure.
Exception indicating the data type of a sub-expression is not valid for use in the containing expression. These errors are typically raised at one of two times:
During definition of a function instance, when it is determined that upper bounds on the expression cannot satisfy the requirements.
When constructing an evaluator in the context of a record schema and the sub-expression fails to satisfy the requirements.
Issued to indicate one of two situations:
The issuer attempted to parse a malformed or non-compliant PMML document, including one that used a different version of the PMML standard.
The issuer was required to build a PMML element that was non-compliant (for example, missing required attributes).
InvalidPropertyValueException
A standard exception for operators and processes to issue when the value of a property is not valid.
Exception that indicates a schema definition is invalid.
Exception issued if attempting to call TokenValuedIterator.stepNext() after it has already returned false.
Exception indicating an error while parsing a value expression.
Exception issued when the serialization of a path cannot be parsed.
Exception issued when the number of characters in the first row of a delimited text file exceeds the maximum allowed. This most often occurs when the file contains UNIX newlines (\n) but Windows newlines (\r\n) were specified, and the reader would otherwise buffer the entire file into memory.
Provides an unchecked wrapper for exceptions issued by scripts providing such information as line and column number of the error.
Exception that indicates a script attempted to assign a value of the wrong type to an output field.
Exception Indicating a parsing error on a data split. The exception always contains:
The path of the data source.
The offset into the data source; this information may be an approximation.
It may also contain additional information about the error, such as the index of record and field being parsed.
Unchecked exception issued to indicate a type is not valid in a given context. For record types, this is usually chained with a FieldTypeNotSupportedException indicating the field which caused this error.
Unchecked exception issued when an attempt is made to compare two typed objects of incompatible types. Exception chaining is used in two ways to provide additional details about the cause of a comparison error. In the case of a record type, the exception may contain a FieldTypesNotComparableException as the cause indicating which fields raised the error. Exceptions issued from operators may contain another TypesNotComparableException. The wrapping exception contains more contextual information about the comparison performed.
Exception issued to indicate that a PMML element was encountered that is not currently supported by our predictor.