15.10 - Types of Resource Usage Table Columns - Teradata Database

Teradata Database Resource Usage Macros and Tables

Teradata Database
Release Number
April 2016
English (United States)
Last Update

This manual describes what each of the resource usage table columns report (that is, what each DBC.ResUsageXxxx.ColumnName reports) in a table format.

The actual table definitions are obtainable by executing the SHOW TABLE statement. For more information about SHOW TABLE, see SQL Data Definition Language.

All columns described in the following chapters and appendixes are type FLOAT unless otherwise specified in the description of that column. All nonexistent values are stored as NULL.

For each resource usage table column, this manual describes the:

  • Column Name
  • Mode

    For descriptions of the different modes of data reporting, see About the Mode Column.

  • Description
  • Data Type

    For descriptions of the data types described in this manual, see SQL Data Types and Literals.

The columns are grouped into either housekeeping columns or statistics columns. Statistic columns are further grouped by category and subcategory as shown in the following table.

Column Name Mode Description Data Type

Each table has:

  • Housekeeping columns which contain statistics on timestamp and current logging characteristics.
  • Statistics columns which can be further categorized into subcategories. Categories and subcategories may vary from table to table.

The following table shows the types of statistics subdivided into their respective subcategories.

Category Subcategories Description
File System
  • AutoCylPack
  • Block-level compression (BLC)
  • Cylinder Defragmentation Overhead
  • Cylinder Management Overhead Events
  • Cylinder MiniCylPack Overhead
  • Cylinder Split and Migrate Overhead
  • File Segment (FSG) Cache Wait
  • FSG I/O
  • Data Block Creations
  • Data Block Merge
  • Data Block Prefetches
  • Data Block Update Operations
  • Data Segment Lock Requests
  • Depot
  • Master Index (MI)
  • Multi-Row Requests
  • Segments Acquired
  • Segments Released
  • Single-Row Requests
  • Synchronized Full Table Scans
  • Transient Journal Requests
  • Transient Journal Overhead
  • Miscellaneous Write Ahead Logging (WAL)
Some of the columns can be viewed as a subset of memory columns by expanding on the operations performed on disk memory segments. Operations counted are logical memory and physical disk reads and writes (including aging) and locking control activities. Other columns identify the purpose of operations being performed on disk segments such as cylinder migration or data updates, the requests being made by database software on the file system, the number of cylinders that were selected or scanned by AutoCylPack, or the number of data blocks that have been compressed or uncompressed. The WAL columns identify the log-based file system recovery scheme in which modifications to permanent data are written to a log file, the WAL log.
General Concurrency Control
  • Database Locks
  • Monitor Management
These columns identify concurrency control activities. These activities provided are subdivided into control done for user level processing, system overhead processing, and database locks. It does not include control specific to disk, memory or net concurrency control, which are included in the disk, memory or net columns.
Host Controller (SHST)
  • Channel Traffic
  • Channel Management
  • Controller Overhead
These columns identify traffic on the host-to-node channels and TCP/IP networks. Some also give overhead and management information on the host channel and TCP/IP network.
  • Memory Allocations
  • Memory Availability Management
  • Memory Pages
  • Memory Resident
  • Paging
  • Task Context Segment Usage
These columns collect memory allocation and deallocation, logical memory and physical disk reads and writes (including paging), access, deaccess and memory control.

Memory management columns are also provided to identify events leading up to paging and aging activities. Finally, a detailed snapshot of the memory is provided by tracking the current states per memory types.

Logical Device
  • Input and Output Traffic
  • Outstanding Requests
  • Response Time
  • Seek Statistics
These columns identify individual logical device activities for external storage components connected through the buses.

The storage device statistics are calculated only on what can be derived from statistics collected by the operating system, since the disk array controllers do not provide us with any useful data for resource usage.

  • Broadcast Net Traffic
  • Group Coordination
  • Message Type
  • Message Delivery Times
  • Merge Services
  • Net Controller Status and Miscellaneous Management
  • Net Circuit Management
  • Net Miscellaneous Contention Management
  • Net Queues
  • Network Transport Data
  • Point-to-point Net Traffic
  • Work Mailbox Queue
These columns identify traffic over the BYNET through the number and direction of messages, subdivided into the type of transmission, as well as physical utilization of the BYNET. Logical messages and direction are identified through subdivisions of the message class. Controller overhead, channel utilization, and Teradata net contention are identified as well.
Process Scheduling
  • ChnSignal Status Tracking
  • CPU Utilization
  • Cylinder Read
  • Interrupted CPU Switching
  • PE and AMP User-Defined Function (UDF) CPU
  • Process Allocation
  • Process Pending Snapshot
  • Process Block Counts
  • Process Pending Wait Time
  • Scheduled CPU Switching
These columns provide a CPU-level snapshot of work started with current characteristics and states. Expanded detail is provided for work started but waiting on resources. This identifies the ability or inability of the system to effectively utilize resources. Time allotments are tracked by monitoring the time spent waiting for resources or processing code. Some of these columns also track the number of times processing was switched to another process for multitasking purposes or to perform interrupt services; and others can provide information about wether the UDFs were doing work for the AMP and PE vprocs.
Reserved None These columns are not used.
Spare None These columns are reserved for a future release or are for internal manipulation by Teradata developers.
  • AMP Worker Task
  • In use and Max Array Data
  • Priority Scheduler
  • Worktype Descriptions
  • Monitor WD
These columns report statistics on the AMP Worker Tasks (AWTs) and Priority Scheduler. These columns also provide RSS data to the System PMPC APIs (for example, the MONITOR WD request and MonitorWD function). For information, see Application Programming Reference.
Transient Journal Management Purge Overhead These columns identify background overhead associated with the occasional transient journal purge operation.
User Commands
  • User command
  • User command Arrival and Departure
These columns describe the types of commands given to Teradata Database by the user and the progress of those commands.
Teradata Virtual Storage (VS)
  • Allocation
  • I/O
  • Migration
  • Node Agent
These columns identify individual pdisk and vdisk device activities.