Current Logging Characteristic Columns | ResUsageSps Table | Teradata Vantage - 17.05 - Miscellaneous Housekeeping Columns - Teradata Database

Teradata Vantage™ - Resource Usage Macros and Tables

prodname
Advanced SQL Engine
Teradata Database
vrm_release
17.00
17.05
created_date
June 2020
category
Administration
User Guide
featnum
B035-1099-170K

These columns provide statistics on current logging characteristics.

Column Name Mode Description Data Type
GmtTime n/a Greenwich Mean Time is not affected by the Daylight Saving Time adjustments that occur twice a year. FLOAT
CabinetID n/a The physical cabinet number of the node. INTEGER
ModuleID n/a The physical module number of the node. INTEGER
NodeType n/a Type of node, representing the per node system family type. CHAR(8)
PPId n/a This column is not used and returns a zero value. BYTEINT
VprType n/a Type of vproc (for example, AMP, PE, and MISC). Rows reported as vproc type of MISC contain data for all vproc types other than the AMP and PE vproc types. CHAR(4)
NumAMPs n/a Number of AMPs on the Node. NumAMPs is used to divide columns that are reporting data from all the AMPs. This allows the ResSpsView view to report the data columns on a per AMP basis. SMALLINT
RowIndex1 n/a This column contains the pWDid. The pWDid value ranges from 0 to 255. SMALLINT
WDId track WD ID.

The default PGs (System, L, M, H, R) have no associated WD and will have a WDID of zero in the table. Only PGs with a name that has “TDWMWD” (SLES 11 or higher) in it have a nonzero WDID.

Resource usages of the workload definition (WD) to which a job instance was assigned can be obtained by joining the DBC.DBQLUtilityTbl table and DBC.ResSPSView on the WDID column for the duration of the job instance.

The example below calculates the sums of IOTA (I/O token allocation) columns.

SELECT SUM(FullPotentialIota), SUM(CodPotentialIota), SUM(UsedIota)
FROM DBC.ResSpsView S,
     DBC.DBQLUtilityTbl U
WHERE S.WDID = U.WDID
  AND U.UtilityName = 'MULTLOAD'
  AND U.UserName = 'USER1'
  AND U.JobEndTime = TIMESTAMP '2013-04-19 16:48:14.98'
  AND S.TheDate >= CAST(Phase1StartTime AS DATE)
  AND S.TheTime >= ((EXTRACT(HOUR FROM Phase1StartTime) * 10000) +
                    (EXTRACT(MINUTE FROM Phase1StartTime) * 100) +
                    (EXTRACT(SECOND FROM Phase1StartTime)))
  AND S.TheDate <= CAST(Phase2EndTime AS DATE)
  AND S.TheTime <= ((EXTRACT(HOUR FROM Phase2EndTime) * 10000) +
                    (EXTRACT(MINUTE FROM Phase2EndTime) * 100) +
                    (EXTRACT(SECOND FROM Phase2EndTime)));
FLOAT
Secs n/a Actual number of seconds in the log period represented by this row. Normally, this value is the same as NominalSecs, but can be different in three cases:
  • The first interval after a log rate change
  • A sample logged late because of load on the system
  • System clock adjustments affect reported Secs

Useful for normalizing the count statistics contained in this row, for example, to a per-second measurement.

SMALLINT
CentiSecs n/a Number of centiseconds in the logging period. This column is useful when performing data calculations with small elapsed times where the difference between centisecond-based data and whole seconds results in a percentage error. INTEGER
NominalSecs n/a Specified or nominal number of seconds in the logging period. SMALLINT
NCPUs n/a Number of online CPUs on this node.

This column is useful for normalizing the CPU utilization column values for the number of CPUs on the node. This is especially important in:

  • Coexistence systems where the number of CPUs can vary across system nodes.
  • Elastic TCore systems where the number of online CPUs can change without database restart.
SMALLINT
Active max Controls whether or not the rows will be logged to the resource usage tables when Active Row Filter Mode is enabled.

If Active is set to a non-zero value, the row contains data columns.

If Active is set to a zero value, none of the data columns in the row have been updated during the logging period.

FLOAT
TheTimestamp n/a Number of seconds since midnight, January 1, 1970.

This column is useful for aligning data with the DBQL log.

BIGINT
PM_COD_CPU n/a Platform Metering (PM) CPU COD value in one tenths of percent. For example, a value of 500 represents a PM CPU COD value of 50%.

The value is set to 1000 if the PM CPU COD is disabled.

SMALLINT
PM_COD_IO n/a Platform Metering (PM) I/O Capacity On Demand (COD) value in whole percent values for the entire system. For example, a value of 50 represents a PM I/O COD value of 50%.

The value is set to 100 if the PM I/O COD is disabled.

SMALLINT
WM_COD_CPU n/a Workload Management (WM) CPU Capacity On Demand (COD) value in one tenths of a percent. For example, a value of 500 represents a WM CPU COD value of 50.0%.

The value is set to 1000 if the WM CPU COD is disabled.

SMALLINT
WM_COD_IO n/a Workload Management (WM) I/O Capacity On Demand (COD) value in whole percent. For example, a value of 50 represents a WM I/O COD value of 50.0%.

The value is set to 100 if the WM I/O COD is disabled.

SMALLINT
TIER_FACTOR n/a I/O performance limit placed on a core-reduced node. For example, a value of 75 represents an I/O limit of 75.0% placed before other COD values

The field is sourced from tpanodep->tierfactor which is derived from the /proc/tdmeter/tier_performance file. A value of 100 indicates that there is no I/O performance limit applied to the node.

.
SMALLINT
VPId track Virtual partition ID.

Only one VpId is associated with a pWDid and VprType row at any point in time. There can be multiple pWDid values associated with a VPId.

FLOAT