Table of contents
TOC
Collapse the table of content
Expand the table of content

Known Issues with Microsoft R Server

j-martens|Last Updated: 1/11/2017
|
1 Contributor

Known Issues with Microsoft R Server 9.0.1

Package: RevoScaleR

Distributed Computing

  • On SLES 11 systems, there have been reports of threading interference between the Boost and MKL libraries.
  • The value of consoleOutput that is set in the RxHadoopMR compute context when wait=FALSE determines whether or not consoleOutput will be displayed when rxGetJobResults is called; the value of consoleOutput in the latter function is ignored.
  • When using RxInTeradata, if you encounter an intermittent failure, simply try resubmitting your R command.
  • The rxDataStep function does not support the varsToKeep and varsToDrop arguments in RxInTeradata.
  • The dataPath and outDataPath arguments for the RxHadoopMR compute context are not yet supported.
  • The rxSetVarInfo function is not supported when accessing xdf files with the RxHadoopMR compute context.
  • When specifying a non-default RNGkind as an argument to rxExec, identical random number streams can be generated unless the RNGseed is also specified.
  • When using small test data sets on a Teradata appliance, some test failures may occur due to insufficient data on each AMP.

Data Import and Manipulation

  • Appending to an existing table is not supported when writing to a Teradata database.
  • When reading VARCHAR columns from a database, white space will be trimmed. To prevent this, enclose strings in non-white-space characters.
  • When using functions such as rxDataStep to create database tables with VARCHAR columns, the column width is estimated based on a sample of the data. If the width can vary, it may be necessary to pad all strings to a common length.
  • Using a transform to change a variable's data type is not supported when repeated calls to rxImport or rxTextToXdf are used to import and append rows, combining multiple input files into a single .xdf file.
  • When importing data from the Hadoop Distributed File System, attempting to interrupt the computation may result in exiting the software.

Analysis Functions

  • Composite xdf data set columns are removed when running rxPredict(.) with rxDForest(.) in Hadoop and writing to the input file.
  • The rxDTree function does not currently support in-formula transformations; in particular, using the F() syntax for creating factors on the fly is not supported. However, numeric data will be automatically binned.
  • Ordered factors are treated the same as factors in all RevoScaleR analysis functions except rxDTree.

Package: RevoIOQ

  • If the RevoIOQ function is run concurrently in separate processes, some tests may fail.

Package: RevoMods

  • The RevoMods timestamp() function, which masks the standard version from the utils package, is unable to find the C_addhistory object when running in an Rgui, Rscript, etc. session. If you are calling timestamp(), call the utils version directly as utils::timestamp().
  • In the nls function, use of the port algorithm occasionally causes the R front-end to stop unexpectedly. The nls help file advises caution when using this algorithm. We recommend avoiding it altogether and using either the default Gauss-Newton or plinear algorithms.

Operationalization (formerly DeployR)

  • When Azure active directory authentication is the only form of authentication enabled, it is not possible to run diagnostics.

Known Issues with Microsoft R Server 8.0.5

RevoScaleR

Distributed Computing

  • On SLES 11 systems, there have been reports of threading interference between the Boost and MKL libraries.
  • The value of consoleOutput that is set in the RxHadoopMR compute context when wait=FALSE determines whether or not consoleOutput will be displayed when rxGetJobResults is called; the value of consoleOutput in the latter function is ignored.
  • When using RxInTeradata, if you encounter an intermittent failure, simply try resubmitting your R command.
  • The rxDataStep function does not support the varsToKeep and varsToDrop arguments in RxInTeradata.
  • The dataPath and outDataPath arguments for the RxHadoopMR compute context are not yet supported.
  • The rxSetVarInfo function is not supported when accessing xdf files with the RxHadoopMR compute context.

Data Import and Manipulation

  • Appending to an existing table is not supported when writing to a Teradata database.
  • When reading VARCHAR columns from a database, white space will be trimmed. To prevent this, enclose strings in non-white-space characters.
  • When using functions such as rxDataStep to create database tables with VARCHAR columns, the column width is estimated based on a sample of the data. If the width can vary, it may be necessary to pad all strings to a common length.
  • Using a transform to change a variable's data type is not supported when repeated calls to rxImport or rxTextToXdf are used to import and append rows, combining multiple input files into a single .xdf file.
  • When importing data from the Hadoop Distributed File System, attempting to interrupt the computation may result in exiting the software.

Analysis Functions

  • Composite xdf data set columns are removed when running rxPredict(.) with rxDForest(.) in Hadoop and writing to the input file.
  • The rxDTree function does not currently support in-formula transformations; in particular, using the F() syntax for creating factors on the fly is not supported. However, numeric data will be automatically binned.
  • Ordered factors are treated the same as factors in all RevoScaleR analysis functions except rxDTree.

DeployR

  • On Linux, if you attempt to change the DeployR RServe port using the adminUtilities.sh, the script incorrectly updates Tomcat's server.xml file, which prevents Tomcat from starting, and does not update the necessary the Rserv.conf file. You must revert back to an earlier version of server.xml to restore service.

  • Using deployrExternal() on the DeployR Server to reference a file that in a specified folder produces a ‘Connection Error’ due to an improperly defined environment variable. For this reason, you must log into the Administration Console and go to The Grid tab. In that tab, edit Storage Context value for each and every node and specify the full path to the external data directory on that node’s machine, such as <DEPLOYR_INSTALLATION_DIRECTORY>/deployr/external/data.

RevoIOQ Package

  • If the RevoIOQ function is run concurrently in separate processes, some tests may fail.

RevoMods Package

  • The RevoMods timestamp() function, which masks the standard version from the utils package, is unable to find the C_addhistory object when running in an Rgui, Rscript, etc. session. If you are calling timestamp(), call the utils version directly as utils::timestamp().
  • In the nls function, use of the port algorithm occasionally causes the R front-end to stop unexpectedly. The nls help file advises caution when using this algorithm. We recommend avoiding it altogether and using either the default Gauss-Newton or plinear algorithms.
© 2017 Microsoft