This page documents system variables related to the CONNECT storage engine. See Server System Variables for instructions on setting them.
See also the Full list of MariaDB options, system and status variables.
Description: Java class path
Commandline: --connect-class-path=value
Description: Enable condition pushdown
Commandline: --connect-cond-push={0|1}
Commandline: --connect-conv-size=#
Description: Default depth used by Json, XML and Mongo discovery.
Commandline: --connect-default-depth=#
Description: Default precision used for doubles.
Commandline: --connect-default-prec=#
Commandline: --connect-enable-mongo={0|1}
Description: Whether the CONNECT engine should return an exact record number value to information queries. It is OFF by default because this information can take a very long time for large variable record length tables or for remote tables, especially if the remote server is not available. It can be set to ON when exact values are desired, for instance when querying the repartition of rows in a partition table.
Commandline: --connect-exact-info={0|1}
Description: Force using BSON for JSON tables. Starting with these releases, the internal way JSON was parsed and handled was changed. The main advantage of the new way is to reduce the memory required to parse JSON (from 6 to 10 times the size of the JSON source to now only 2 to 4 times). However, this is in Beta mode and JSON tables are still handled using the old mode. To use the new mode, tables should be created with TABLE_TYPE=BSON, or by setting this session variable to 1 or ON. Then, all JSON tables will be handled as BSON. This is temporary until the new way replaces the old way by default.
Commandline: --connect-force-bson={0|1}
Description: Enable file mapping for index files. To accelerate the indexing process, CONNECT makes an index structure in memory from the index file. This can be done by reading the index file or using it as if it was in memory by “file mapping”. Set to 0 (file read, the default) or 1 (file mapping).
Commandline: --connect-indx-map=#
Description: Java wrapper.
Commandline: --connect-java-wrapper=val
Default Value: wrappers/JdbcInterface
Description: Discovery to generate json path for all columns if ON (the default) or do not when the path is the column name.
Commandline: --connect-json-all-path={0|1}
Description: Max number of rows for JSON aggregate functions.
Commandline: --connect-json-grp-size=#
Default Value: 50 (>= Connect 1.7.0003), 10 (<= Connect 1.7.0002)
Description: Representation of JSON null values.
Commandline: --connect-json-null=value
Description: Path to JVM library.
Commandline: --connect-jvm_path=value
Description: Determines the handling of TEXT columns.
NO: The default until Connect 1.06.005, no conversion takes place, and a TYPE_ERROR is returned, resulting in a “not supported” message.
YES: The default from Connect 1.06.006. The column is internally converted to a column declared as VARCHAR(n), n being the value of connect_conv_size.
FORCE (>= Connect 1.06.006): Also convert ODBC blob columns to TYPE_STRING.
SKIP: No conversion. When the column declaration is provided via Discovery (meaning the CONNECT table is created without a column description), this column is not generated. Also applies to ODBC tables.
Commandline: --connect-type-conv=#
Valid Values: NO, YES, FORCE or SKIP
Description:
NO: The first algorithm is always used. Because it can cause errors when updating variable record length tables, this value should be set only for testing.
AUTO: This is the default value. It leaves CONNECT to choose the algorithm to use. Currently it is equivalent to NO, except when updating variable record length tables (DOS, CSV or FMT) with file mapping forced to OFF.
YES: Using a temporary file is chosen with some exceptions. These are when file mapping is ON, for VEC tables and when deleting from DBF tables (soft delete). For variable record length tables, file mapping is forced to OFF.
FORCE: Like YES but forces file mapping to be OFF for all table types.
TEST: Reserved for CONNECT development.
Commandline: --connect-use-tempfile=#
Description: Size of the CONNECT work area used for memory allocation. Permits allocating a larger memory sub-allocation space when dealing with very large if sub-allocation fails. If the specified value is too big and memory allocation fails, the size of the work area remains but the variable value is not modified and should be reset.
Commandline: --connect-work-size=#
Scope: Global, Session (Session-only from CONNECT 1.03.005)
Range: 4194304 upwards, depending on the physical memory size
Description: Console trace value. Set to 0 (no trace), or to other values if a console tracing is desired. Note that to test this handler, MariaDB should be executed with the --console parameter because CONNECT prints some error and trace messages on the console. In some Linux versions, this is re-routed into the error log file. Console tracing can be set on the command line or later by names or values. Valid values (from Connect 1.06.006) include:
INDEX or 4: Index construction
MEMORY or 8: Allocating and freeing memory
SUBALLOC or 16: Sub-allocating in work area
QUERY or 32: Constructed query sent to external server
STMT or 64: Currently executing statement
HANDLER or 128: Creating and dropping CONNECT handlers
BLOCK or 256: Creating and dropping CONNECT objects
For example:
set global connect_xtrace=0; No trace
set global connect_xtrace='YES'; By name
set global connect_xtrace=1; By value
set global connect_xtrace='QUERY,STMT'; By name
set global connect_xtrace=96; By value
set global connect_xtrace=1023; Trace all
Commandline: --connect-xtrace=#
Valid Values: See description
This page is licensed: CC BY-SA / Gnu FDL
RetroSearch is an open source project built by @garambo | Open a GitHub Issue
Search and Browse the WWW like it's 1997 | Search results from DuckDuckGo
HTML:
3.2
| Encoding:
UTF-8
| Version:
0.7.4