Changing configuration settings

There are some configuration settings that you can change yourself, without involving your support representative. These settings are referred to in documented Netezza Performance Server procedures elsewhere in this publication or elsewhere in the Netezza Performance Server product library.

Attention: Never change a configuration setting unless directed to do so explicitly by either:

Your support representative
A documented Netezza Performance Server procedure

Incorrect configuration settings might lead to a significant loss of performance and other problems.

There are two ways to change configuration settings:

Temporarily

For you to be able to change a configuration setting temporarily, you must either be the admin user or your user account must have the Manage System privilege. To change a configuration setting temporarily, pause the system, issue the appropriate nzsystem set command, then resume the system. This is done by issuing the following commands:

nzsystem pause
nzsystem set -arg setting=value
nzsystem resume

A change that is made in this way remains effective only until the system is restarted; at system startup, all configuration settings are read from the system configuration file and loaded into the registry.

Permanently

To change a configuration setting permanently, edit the corresponding line in the configuration file, system.cfg. Configuration settings are loaded from this file to the registry during system startup.

The following tables describe the configuration settings that you can change yourself, without involving your support representative.

Table 1. Configuration settings for short query bias (SQB)
Setting	Type	Default	Description
host.schedSQBEnabled	bool	true	Whether SQB is enabled (true) or disabled (false).
host.schedSQBNominalSecs	int	2	The threshold, in seconds, below which a query is to be regarded as being short.
host.schedSQBReservedGraSlots	int	10	The number of GRA scheduler slots that are to be reserved for short queries.
host.schedSQBReservedSnSlots	int	6	The number of snippet scheduler slots that are to be reserved for short queries.
host.schedSQBReservedSnMb	int	50	The amount of memory, in MB, that each SPU is to reserve for short query execution.
host.schedSQBReservedHostMb	int	64	The amount of memory, in MB, that the host is to reserve for short query execution.

Table 2. Configuration settings for updating virtual tables
Setting	Type	Default	Description
host.graVtUpdateInterval	int	600	The number of seconds between updates to the _vt_sched_gra table, which contains resource usage statistics for completed jobs.
host.snVtUpdateInterval	int	60	The number of seconds between updates to the _vt_sched_sn table, which contains resource usage statistics for completed snippets.
host.sysVtUpdateInterval	int	600	The number of seconds between updates of the _vt_sched_sys virtual table.
host.sysUtilVtUpdateInterval	int	60	The number of seconds between updates of the _vt_system_util virtual table.

Table 3. Configuration settings for plan history
Setting	Type	Default	Description
host.planHistArchive	bool	true	The system stores the plan files and C++ source files that it creates for each snippet. This setting determines where these files are stored: true The files are stored in a compressed archive (tar) file. This option was introduced in Release 7.0. false The files are stored in plan directories. This was the only option that was available in Release 6.0.8 and earlier.
host.planHistFiles	int	2000	If host.planHistArchive=true, the maximum number of plans that can be stored in a single compressed archive file. If host.planHistArchive=false, the maximum number of plan directories.
host.planHistKeepArchives	int	10	The number of compressed archive files to keep.

Table 4. Configuration settings for downtime event logging
Setting	Type	Default	Description
sysmgr.numberOfDownPortToRiseEvent	int	5	The number of ports on the same switch that must be down for the amount of time specified by sysmgr.portDownTime1ToRiseEvent before the system logs a HW_NEEDS_ATTENTION event. Specify 0 to deactivate logging HW_NEEDS_ATTENTION events based on sysmgr.portDownTime1ToRiseEvent.
sysmgr.portDownTime1ToRiseEvent	int	300	The number of seconds that the system waits after the number of ports specified by sysmgr.numberOfDownPortToRiseEvent go down before logging a HW_NEEDS_ATTENTION event. For example, if sysmgr.numberOfDownPortToRiseEvent=5 and sysmgr.portDownTime1ToRiseEvent=300, when at least 5 ports on the same switch are down for at least 300 seconds the system logs a HW_NEEDS_ATTENTION event. Under normal operating conditions, ports sometimes go down for a few seconds at a time. The delay introduced by this setting prevents events from being logged for such transient state changes.
sysmgr.portDownTime2ToRiseEvent	int	600	The number of seconds that any one port must be down before the system logs a HW_NEEDS_ATTENTION event for that port. This happens independently of the settings for sysmgr.numberOfDownPortToRiseEvent and sysmgr.portDownTime1ToRiseEvent. Under normal operating conditions, ports sometimes go down for a few seconds at a time. The delay introduced by this setting prevents events from being logged for such transient state changes.

Table 5. Configuration settings for backup
Setting	Type	Default	Description
host.bnrNumStreamsDefault	int	0	The default number of streams to use for a backup operation as described in Multi-stream backup. A value of 0 or 1 causes the system to use to one stream per backup location. The maximum number of streams depends on the number of data slices on the system. For example, if you have 120 data slices, the maximum number of streams is 120.

Table 6. Configuration settings for restore
Setting	Type	Default	Description
system.disableMtxDataMgrLocks	int	false	Duplicate records may be added to a table if not set to false.

Table 7. Configuration settings for loading UTF-8 codepoint into a table
Setting	Default	Description
host.disableNfcGlobally	yes	Starting from version 11.0.1.0, any valid UTF-8 codepoint can be easily loaded or inserted into a table.

Table 8. Configuration settings for including queries in tables
Setting	Default	Description
host.queryHistShowInternal	yes	Starting from version 11.0.1.0, the `JIT-Stats Sample Scan` queries are included in the `_T_QRYSTAT / _T_QRYHIST` tables.