Analysis Metrics for Teradata Database

Teradata Viewpoint User Guide

brand
Analytical Ecosystem
prodname
Teradata Viewpoint
vrm_release
16.00
category
User Guide
featnum
B035-2206-106K
The following metrics are available to analyze resource usage for a Teradata Database system.
Metric Description Type
Active Sessions Sessions with active queries Number
AMP CPU Skew Comparison of CPU use on the busiest AMP to the average AMP Percent
AMP I/O Skew Comparison of disk use on the busiest AMP to the average AMP Percent
Average Memory Available Average approximate free memory, as reported by the Linux “free” command, across all nodes Number
Average Outstanding Disk Reqs Outstanding disk requests (disk queue size) Number
AWT Average AMP worker tasks in use on each AMP Number
Components Components, such as BYNETs or AMPs that are down Number
CPU Average node CPU use. CPU is calculated as the sum of the user CPU and system CPU usage percentages. Percent
CPU AWT Processing CPU resources spent processing an AMP worker task. Normalized by multiplying the value returned from Teradata Database by the number of physical CPUs on the associated node of the AWT. Percent
CPU/Disk Ratio Ratio of CPU use to disk use. Calculated as the node CPU time divided by the physical disk usage. Percent
CPU Dispatcher Processing CPU resources spent in PE Dispatcher processing Percent
DBC Disk Space Available DBC disk space in use Percent
Disk I/O Disk I/Os Number
Disk Reads Total physical disk reads per system during the sample period Number
Disk Usage Disk use on the system Percent
Disk Writes Total physical disk writes per system during the sample period Number
FSG Cache Miss Percentage of FSG cache misses. Calculated as physical I/Os divided by logical I/Os. Percent
Host Block Reads Message blocks (one or more messages sent in one physical group) received from all clients Number
Host Block Writes Message blocks sent to all hosts Number
Index Ratio Index I/Os divided by the data block I/Os Percent
Logical MB/Sec Logical I/O of the system in megabytes per second Number
Max Disk Space By AMP Available disk space currently in use Percent
Max Spool Space By AMP Available spool space in use Percent
Minimum Memory Available Minimum approximate free memory, as reported by the Linux “free” command, across all nodes. The value is from the node with the least free memory. Number
Net A Usage BYNET A usage (BYNET receiver usage) Percent
Net Reads Reads from the BYNET during the sample period Number
Net Writes Messages written to the BYNET during the sample period Number
Node CPU Skew Comparison of CPU use on the busiest node to the average node Percent
Node CPU Usage Average CPU usage of all online nodes in the configuration Percent
Node I/O Skew Comparison of disk use on the busiest node to the average node Percent
Nodes Down System nodes down Number
Parallelism Percentage of parallelism, calculated as the average CPU usage divided by the maximum CPU usage x 100 Percent
PE CPU Usage Average CPU usage of the PEs, calculated as the average usage of the parser and the dispatcher (both normalized values). Normalized by multiplying the value by the number of physical CPUs on the associated node of the PE. Percent
Read I/O Percent of I/O that are reads Percent
Session Login Count Sessions currently logged on to the system Number
Swap Drops Pages or segments dropped from memory during the sample period due to swapping Number
Swap Reads Pages or segments read into memory from the disk after a prior write or drop during the sample period Number
Swap Writes Pages or segments written into swap area from memory during the sample period Number
System CPU Average CPU time spent in System mode Percent
Total AMP CPU Total AMP CPU use Percent
Total Disk Space Total disk space in use Percent
Total Node CPU Total node CPU use Percent
User CPU Average CPU time spent in User mode Percent
Vproc CPU Usage Average CPU usage of all online vprocs in the configuration Percent
Wait I/O CPU Average CPU time spent waiting for I/O Percent