Database: td_metric_svc
If you need detailed statistics for each virtual disk (vdsk) in the system, use this view to:
- Compare how each AMP is using its allocated virtual disks.
- Evaluate the amount of time consumed by outstanding requests, per AMP.
- Identify non-parallel activity for various metrics such as Read/Write count, Read/Write volume, and so on.
| View Column | Data Type | Format | Comment |
|---|---|---|---|
| Location | VARCHAR(2048) UNICODE CASESPECIFIC | X(2048) | File path in object storage for metric data objects. |
| path_component_id | VARCHAR(8000) UNICODE CASESPECIFIC | X(8000) | Operational group (primary cluster or compute cluster) ID. |
| path_collecttimestamp | VARCHAR(8000) UNICODE CASESPECIFIC | X(8000) | Timestamp, in EPOCH timestamp format, this metric data was collected from your database and written to object storage. |
| path_year | VARCHAR(8000) UNICODE CASESPECIFIC | X(8000) | The year in which the data was logged. |
| path_month | VARCHAR(8000) UNICODE CASESPECIFIC | X(8000) | The month in which the data was recorded in the table. |
| path_day | VARCHAR(8000) UNICODE CASESPECIFIC | X(8000) | The day on which the data was recorded in the table. |
| path_hour | VARCHAR(8000) UNICODE CASESPECIFIC | X(8000) | The hour at which the data was recorded in the table. |
| path_minute | VARCHAR(8000) UNICODE CASESPECIFIC | X(8000) | The minute at which the data was recorded in the table. |
| path_ver | VARCHAR(8000) UNICODE CASESPECIFIC | X(8000) | The version of Vantage software that was running on the operational group. |
| TheDate | DATE NOT NULL | YYYY/MM/DD | Local Date of the log entry. |
| NodeID | INTEGER NOT NULL | ZZZ9-9999 | Numeric value that uniquely identifies a node (cloud compute instance) within a compute cluster. For example, '10001','10002'. |
| TheTime | FLOAT NOT NULL | 99:99:99.99 | Nominal local time of the log entry. |
| GmtTime | FLOAT NOT NULL | 99:99:99.99 | Nominal GMT time of the log entry. Greenwich Mean Time is not affected by the Daylight Saving Time adjustments that occur twice a year. |
| CabinetID Not applicable. Do not use.
|
Not applicable. | ||
| ModuleID Not applicable. Do not
use.
|
Not applicable. | ||
| NodeType | CHAR(8) LATIN NOT CASESPECIFIC NOT NULL | X(8) | Indicates the node (cloud compute instance) family. For example, '67_R5'. |
| TheTimestamp | BIGINT NOT NULL | -(20)9 | Number of seconds since midnight, January 1, 1970. |
| CentiSecs | INTEGER NOT NULL | -(11)9 | Actual number of centiseconds in the logging period. |
| Secs | SMALLINT NOT NULL | -(6)9 | Actual number of seconds in the logging period. |
| NominalSecs | SMALLINT NOT NULL | -(6)9 | Nominal number of seconds in the logging period. |
| SummaryFlag | CHAR(1) LATIN NOT CASESPECIFIC NOT NULL | X(1) | Summarization status of this row. Possible values are N if the row is a non-summary row, and S if the row is a summary row. |
| ReservedS0 | CHAR(1) LATIN NOT CASESPECIFIC NOT NULL | X(1) | Reserved for future use. |
| PM_COD_CPU | SMALLINT NOT NULL | -(6)9 | Platform Metering CPU COD value in one tenths of a percent. For example, a value of 500 represents a COD value of 50.0%. |
| PM_COD_IO | SMALLINT NOT NULL | -(6)9 | Platform Metering IO COD value in whole percent. For example, a value of 50 represents a COD value of 50.0%. |
| WM_COD_CPU | SMALLINT NOT NULL | -(6)9 | Workload Management CPU COD value in one tenths of a percent. For example, a value of 500 represents a COD value of 50.0%. |
| WM_COD_IO | SMALLINT NOT NULL | -(6)9 | Workload Management IO COD value in whole percent. For example, a value of 50 represents a COD value of 50.0%. |
| TIER_FACTOR | SMALLINT NOT NULL | -(6)9 | 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. |
| VprId | INTEGER NOT NULL | -(11)9 | Identifies the AMP vproc |
| ReadKB | FLOAT NOT NULL | -(20)9 | Number of KB read from the logical device. |
| WriteKB | FLOAT NOT NULL | -(20)9 | Number of KB written to the logical device |
| ReadRespTot | FLOAT NOT NULL | -(20)9 | Total of individual read response time in centiseconds. |
| WriteRespTot | FLOAT NOT NULL | -(20)9 | Total of individual write response time in centiseconds. |
| WriteRespSq | FLOAT NOT NULL | -(20)9 | Total of squares of the individual write response time in centiseconds. |
| ExtMigrateReadRespTot | FLOAT NOT NULL | -(20)9 | Migration read I/O response time |
| ExtMigrateWriteRespTot | FLOAT NOT NULL | -(20)9 | Migration write I/O response time |
| ExtMigrateIOTimeImprove | FLOAT NOT NULL | -(20)9.99 | Estimates the percent improvement in average I/O response time due to migrations completing in the log interval. If the average I/O response time was 10 ms and the ExtMigrateIOTimeImprove is 10%, average I/O response time is (100%-10%)*10ms = 9ms. |
| Active | FLOAT NOT NULL | -(20)9 | Controls whether the rows are logged to the resource usage tables if Active Row Filter Mode is enabled. |
| ReadCnt | FLOAT NOT NULL | -(20)9 | Number of logical device reads. |
| WriteCnt | FLOAT NOT NULL | -(20)9 | Number of logical device writes. |
| ReadRespMax | FLOAT NOT NULL | -(20)9 | Maximum of individual read response time in centiseconds. |
| WriteRespMax | FLOAT NOT NULL | -(20)9 | Maximum of individual write response time in centiseconds. |
| ConcurrentMax | FLOAT NOT NULL | -(20)9 | Maximum number of concurrent I/O requests. |
| ConcurrentReadMax | FLOAT NOT NULL | -(20)9 | Maximum number of concurrent read I/O requests. |
| ConcurrentWriteMax | FLOAT NOT NULL | -(20)9 | Maximum number of concurrent write I/O requests. |
| OutReqTime | FLOAT NOT NULL | -(20)9 | Time with outstanding requests (busy time), in centiseconds. |
| ExtAllocHot | FLOAT NOT NULL | -(20)9 | Number of hot allocations made in the current log period. An allocation whose estimated temperature falls within the pre-defined hot temperature range. Temperature is the frequency of access to the data by I/O independent of where the data resides. |
| ExtAllocWarm | FLOAT NOT NULL | -(20)9 | Number of warm allocations made in the current log period. An allocation whose estimated temperature falls within the pre-defined warm temperature range. Temperature is the frequency of access to the data by I/O independent of where the data resides. |
| ExtAllocTotal | FLOAT NOT NULL | -(20)9 | Total number of allocations. Cold Allocation = ExtAllocTotal - ExtAllocHot - ExtAllocWarm, QueryPacing Allocation = ExtAllocTotal - ExtAllocNonPacing - ExtAllocSystemPacing, Dynamic Allocation = ExtAllocTotal - ExtAllocStatic. |
| ExtAllocNonPacing | FLOAT NOT NULL | -(20)9 | Number of non-pacing allocations made in the current log period. A non-pacing allocation is an allocation whose data access affects neither system performance nor individual query performance. |
| ExtAllocSystemPacing | FLOAT NOT NULL | -(20)9 | Number of system pacing allocations made in the current log period. A system pacing allocation is an allocation whose data access affects system performance. |
| ExtAllocStatic | FLOAT NOT NULL | -(20)9 | Number of static allocations made in the current log period. A static allocation is an allocation whose requested temperature is used and the measured temperature is ignored during migration. |
| ExtMigrateFaster | FLOAT NOT NULL | -(20)9 | Number of cylinders migrated to faster locations. SlowerMigration = ExtMigrateTotal - ExtMigrateFaster, Cylinders are migrated to slower locations to make room for hotter cylinders to replace them. |
| ExtMigrateTotal | FLOAT NOT NULL | -(20)9 | Total number of cylinders migrated to a different physical location. For more information, see the ExtMigrateFaster column. |
| ExtMigrateIOTimeCost | FLOAT NOT NULL | -(20)9 | Estimates the total cost, in centiseconds, incurred by migration I/Os completing during the log period, where cost is the extra time waited by all non-migration I/Os as a result of the migration I/O. |
| ExtMigrateIOTimeBenefit | FLOAT NOT NULL | -(20)9 | Estimates the total I/O time savings achieved by migrations completing in the log period. The I/O time savings include the improvement in response time caused by the new data arrangement up to the time horizon. |