SAND CDBMS Administration Guide
SAND CDBMS File Types

 

This document lists the files created and/or used by components of SAND CDBMS. In most cases, only the file extension is listed because the full file name varies. The variable part of the file name is represented by the asterisk (*) wildcard. While the majority of the files listed below are proprietary to SAND CDBMS, some of the files are generic. In such cases, the description will clarify how the file or format is used by SAND CDBMS.


dbname_xxxx.tmp

The database temporary file (not to be confused with the temporary Delta File used in Virtual Mode operations) is used for storing temporary records produced during database operations. Since these temporary records do not take up space in the SAND database files, the database growth that occurs during some database operations can be substantially reduced.

The xxxx in the file name is the numerical process ID (PID) for the running database.

The default location for the database temporary file is the current database directory.

dbname_pvdb.chr This file is automatically created by SAND Enterprise Manager (SEM) for use with a Time Travel database. It contains an ASCII representation of the Time Travel tree file for internal use by SEM. This file must not be altered manually.
agent_username.idx An index file created by SAND Enterprise Manager in conjunction with the SAND Enterprise Agent log. This file is used internally.
agent_username.log

The log file maintained by SAND Enterprise Manager for SAND Enterprise Agent processes. The log is written to the LOG subdirectory of the SAND program directory.

*.BAD During an import, the SAND CDBMS Data Loader (ndlm) creates a rejected record file, to which are written any records that cannot be loaded into the target table due to some error condition. The reject file is called input-file.BAD, where input-file is the name of the ASCII file containing the data being imported. Data type mismatch, or violations of NOT NULL constraints are common examples of error conditions that might result in record rejection. The reject file is created in the current directory.
*.Cxx

The Virtual Mode Update File created with the SHUTDOWN...KEEP CHANGE command. The Update File is used to apply changes made to the master instance to the real database.

If the user does not specify a name and location for the Update File, it is given the same name as the database, and is saved in the directory specified by the DatabasePath parameter in the nucleus.ini section for the instance. The xx in the file extension corresponds to the number of the database checkpoint to which the changes will be applied.

*.CDF A core dump file (CDF) is produced by the nserv program if an error condition causes it to exit prematurely. The CDF is intended to be viewed by SAND developers for the purpose of debugging.
core

A core file is created by the UNIX operating system when a process terminates abnormally. This file contains information about what the process was doing when it terminated, which can be used by developers for debugging purposes. Usually this file is very large.

When a core file is produced by a SAND program that terminated abnormally, the file can be safely deleted in most cases. In some situations, SAND support personnel may require the core file to determine the cause of the abnormal termination.

*.DDL Data Definition Language script. This is a text file that contains a sequence of SQL commands that define (create) database objects. By default, the SAND Import/Export Wizard creates a DDL file for an Import or ODBC Import operation.
errmap.sql A resource file that contains the SQL error message strings. It is located in the SAND program directory. This file must never be edited manually.
hdrs.tmp A temporary file created by ndbchk while performing a database packing operation. The temporary file stores modified header frame records. This file should be deleted automatically by ndbchk when the operation finishes.
inter.out

The output file created or overwritten by the SAND CDBMS Interactive SQL client program (nisqlm) when session logging is enabled, either by invoking nisqlm with the -o1 flag, or by executing the .OUTPUT system command during a session.

The output file is saved in the current directory. If logging is turned off and then on again within a single nisqlm session, nisqlm continues to append to the output file without overwriting its contents. Turning on logging in a subsequent session, however, will overwrite any preexisting output file.

*.LOG A log file contains a list of diagnostic messages (including errors) returned by a program as it runs. By default, the SAND Import/Export Wizard maintains a running log file for each project run.
log.ndl The default SAND CDBMS Data Loader (ndlm) log file. The results of ndlm operations, including error messages, are recorded in the log.ndl file, which is saved in the current directory. The log file can be given a user-specified name by including the -e flag with the ndlm invocation.
*.MASTER

A file that stores information about the master engine. This file is created and maintained automatically by Octopus; it should not be edited manually.

*.nxx

A SAND database consists of one or more files with this extension. These files contain the database configuration information and all data represented as database tables (including system tables). These database files can be accessed only by nconfig and other SAND CDBMS tools.

The primary database file has a file name extension of .n00. The primary database file contains the database configuration information, in addition to data stored in tables.

The primary database file is also called the database root, because SAND CDBMS stores data exclusively in the primary database file until its maximum size is reached; only then does SAND CDBMS begin to store data in a second data file. Once the second data file reaches its maximum size, SAND CDBMS begins storing data in the third data file, and so on. In this sense, the primary database file is the root of a sequence of SAND database files.

Secondary data files are created by SAND CDBMS as needed, and have file name extensions that reflect their relative position in the sequence of data files for that particular database. The file name has the form database-name.nxx, where xx is a hexadecimal number from 00 to fff that reflects the position of the file in the sequence of secondary data files. For example, if testdb.n00 is the database root, testdb.n01 would be the second data file, testdb.n02 would be the third, testdb.n03 the fourth, and so on. When testdb.nff is reached, the next data file would be testdb.n100, followed by testdb.n101, and so on, until the last possible data file, testdb.nfff.

The maximum number of primary and secondary *.nxx files for a single database is 4096.

ndb.n00 The "seed" SAND database. This file is required for the creation of new databases. It is located in the SAND program directory.
*.NDL SAND CDBMS Data Loader script. This is a text file that contains the specification script for a load operation. By default, the SAND Import/Export Wizard creates an NDL file for an import or export Task.
nes.ini The SAND Enterprise Service/Manager configuration file, located in the SAND program directory. This file is generated automatically; it should not be edited manually.
nes.log The log file maintained by SAND Enterprise Manager for the SAND Enterprise Service process. The log is written to the LOG subdirectory of the SAND program directory.
nucleus.ini Every SAND CDBMS installation includes a configuration file named nucleus.ini. The nucleus.ini file resides in the same directory as the SAND CDBMS core executable files (this is the SUPPORT directory as specified in the NUCLEUS environment variable). Normally, nucleus.ini settings are made automatically, and should not be altered by users.

Server-side nucleus.ini files contain information related to SAND database operations and client connections on a particular host computer. Client-side nucleus.ini files contain information about how to connect as a client to a SAND database over a network.

odbc.ini The ODBC configuration file contains a list of Data Sources and Data Source properties. If Nucleus ODBC is used to connect to SAND CDBMS data sources, the odbc.ini file must contain appropriate Nucleus ODBC information.

*.OST

The cumulative Octopus statistics file.

Cumulative statistical information is compiled over a user-specified period. This period (measured in seconds) is set with the StatInterval parameter in the OCTOPUS section(s) of the nucleus.ini file. After the specified amount of time has elapsed, the values are written to a comma-separated value file (specified by the nucleus.ini CStatFile parameter, with the extension .ost) and the time counter is reset to 0. Cumulative values since startup of the Octopus system are also calculated.

The default detailed statistics file is database-name.ost, located in the directory specified by the DatabasePath parameter.

*.OTC The Octopus trace file.
*.OTL All Octopus error and warning messages are recorded in the Octopus log file. The name and location of this file (given the extension .otl) are specified with the LogFile parameter in the OCTOPUS section(s) of the nucleus.ini file. The default log file is database-name.otl, located in the directory specified by the DatabasePath parameter.
pesxxxxx.tmp A "pending errors storage" file, created by the Parallel Loader to accumulate error messages (if any) from the different threads in the operation. At the end of the load process, any messages stored in the file are displayed at the console.

Temporary files of this type are created and removed automatically by the loader. If such files still exist after the associated load operations have concluded, most likely the loader terminated abnormally, and therefore these leftover *.tmp files can be safely deleted.

The file name contains a numeric counter that starts at 00000, and is incremented by one for each subsequent file.

prcxxx.tmp A temporary file created by Octopus, used for storing STDOUT and STDERR messages. There is one prcxxx.tmp file per running OctoEng.

The xxx in the file name is a hexadecimal number that increases sequentially with each additional temporary file.

*.PRJ A SAND Import/Export Wizard project file. After one or more Tasks (load operations) are defined through the Import/Export Wizard, the Task information can be saved to a project file. Subsequently, the project can be edited further, run interactively through the Wizard, or run in batch mode from the command line.
*.RCV

A Time Travel recovery file. When a merge to the root node is performed, this file (named database-name.rcv) is created automatically in the same directory as the primary database file (database-name.n00). The recovery file contains information about the merge operation, such that if the merge fails for any reason, it can be completed subsequently from where it was interrupted.

The recovery file must not be manually deleted or moved. When the merge operation has successfully completed, the file will be removed automatically.

The size of the recovery file is the SAND CDBMS page size (1 MB or 64 KB) plus an additional 8 bytes.

sand.lic The SAND CDBMS license file. A valid license file is required to use the SAND server beyond the demo period. The license file is found in the SAND program directory, or in the directory specified by the LicenseFileLocation parameter.
*.SCT SAND Compacted Table (SCT) file. An SCT File is a storage archive containing highly compressed table data that, when used with SAND CDBMS Nearline, can be queried without the need to decompress it first. An SCT File is created using the SAND CDBMS Data Loader (ndlm). The contents of an SCT File can also be loaded into a SAND database using ndlm.
*.SKP

SAND CDBMS Data Loader (ndlm) skip files are created during an import or export only when the SKIPIF option is included in the load specification file. Records containing fields that satisfy a specified condition or conditions are placed in a "skip file" instead of being imported/exported.

If no name for this file is defined by the user, SAND CDBMS assigns it a default name consisting of the name of the load specification file, with the .skp extension. Note that when more than one SKIPIF clause is used, file names must be included for each (except for the last SKIPIF clause in a series, which will write by default to the spec-file.skp file if no other file name is specified).

*.SQL SQL script file. This file contains a series of SQL commands that can be executed sequentially in batch mode. By default, the SAND Import/Export Wizard creates an SQL file for an Export or ODBC Import operation.
*.STA When detailed statistics are compiled, the information is collected for each query executed in the Octopus system, and saved in the .sta file specified with the StatFile parameter in the relevant OCTOPUS section(s) of the nucleus.ini file. The default detailed statistics file is database-name.sta, located in the directory specified by the DatabasePath parameter.

*.Txx The temporary Delta File, created when the database instance is started in Virtual or Persistent Mode. It is transformed into an Update File upon execution of the SHUTDOWN...KEEP CHANGE command.

It is automatically deleted when the database instance is started again after being shut down. The location of temporary Delta Files is set with the DeltaPath parameter in the [DATABASE instance-name] nucleus.ini file; their maximum size is set with the MaxDeltaFileSize or MaxDeltaFileRatio parameter.

*.TxxIxx Generated by an Octopus system in Public Update mode, this index file points to the latest sub-entry point in the master Delta. It has the same root name as the database, with a .TxxIxx extension (where the Txx characters are a hexadecimal counter, and the Ixx digits represent the Snapshot version).

Note that this file is created and maintained automatically by Octopus; it should never be edited manually or deleted.

*.TRC

Trace files, which are text files that contain captured event data from specific areas of SAND software (for example, API, ODBC, or communication layer). The root name of the trace file indicates what kind of data it contains (for instance, api.trc).

These generated trace files are often used for troubleshooting by SAND support personnel, and are not enabled by default. Generally, the information contained in a trace file is not useful to end users. Trace files can be safely deleted when they are no longer needed.

*.TREE

Stores tree structure information for a Virtual or Persistent Mode database. The tree file is named database-name.tree, and is located in the same directory as the primary database file (database-name.n00).

*.Uxx The Persistent Mode Update File represents a Snapshot of the database. Changes made to the current database Snapshot are saved by executing the SHUTDOWN...KEEP CHANGE command from an SQL session, thereby creating an Update File/Snapshot. The Update File (named database-name.Uxx, where xx is a hexadecimal number from 00 to ff that increases incrementally with each update) is saved in the same directory as the primary database file (database-name.n00).
vfs[host.PID]xxxxx.tmp This Virtual File System (VFS) container file is used to store various temporary data during loader, ssau, or ssaserver operations. Characteristics of this temporary file are configured via the nucleus.ini [CLIENT] section parameters TEMPDRIVE, TEMPCACHE, and TEMPPAGE, as well as the ssa.ini [QUERY] section parameter VFS.

The file name contains, between square brackets, the current host name and the process ID of the process that created the VFS file. The file name also contains a numeric counter that starts at 00000, and is incremented by one for each subsequent file generated by the same host/process.

VFS files left by processes that terminated abnormally can be safely deleted.