Chat now with support
Chat with Support

Foglight for Storage Management Shared 4.9 - User Guide

Getting Started Configuring Agents to Monitor Storage Devices
Brocade SAN Switches Cisco SAN Switches Dell Compellent Arrays Dell EqualLogic PS Series Array Groups EMC CLARiiON CX Series Arrays (CLI) EMC CLARiiON, VNX, or VMAX Storage Arrays (SMI-S) EMC Isilon EMC VPLEX Hitachi Data Systems AMS, USP, and VSP HP EVA Storage Arrays HP 3PAR Arrays NetApp Filers Configuration Procedures
Managing Agents Using Foglight for Storage Management Monitoring Storage Performance Investigating Storage Devices Investigating Storage Components Investigating VPLEX Storage Troubleshooting Storage Performance Managing Data Collection, Rules, and Alarms Understanding Metrics Appendix: Collection Target Support Matrix Online-Only Topics

Units of Measurement

Foglight for Storage Management uses the following units of measurement:

•
B/s — Number of bytes per second. Frequently converted to KB/s or MB/s.
•
ops/s — Number of operations per second.
•
ms — Milliseconds. Frequently converted to μs (microsecond).
•
m — milli — Thousandth.
•
ms/op — Milliseconds per operation.
•
MB — Capacity metrics are captured in MB (megabytes), but are frequently displayed in GB (gigabytes) or TB (terabytes).

Performance Metrics

In general, performance is gauged in terms of rates (such as data rates, ops to disk rates, frame rates), latency, and utilization. The selection of performance metrics displayed in any given dashboard depends on the selected device type and the metrics available from the device vendor.

The following tables define each metric and shows the topology name for the metric. You need the topology name when referencing the metric in user-defined rules and dashboards.

Performance metrics are organized into the following categories:

•
Fabrics and FC Switches — Performance Metrics
•
Storage Arrays and Filers — Disk I/O Performance Metrics
•
Clustered Storage Arrays — Network Performance Metrics

Fabrics and FC Switches — Performance Metrics

For fabrics and switches, performance is assessed in terms of send and receive values for the FC switch ports and is expressed as the average number of bytes per second or frames per second. Port errors are also an important indicator of port performance, because a higher number of errors correlates with lower send and receive values.

 
Table 21. Performance metrics
Display Name
Description
Topology Name
Baseline

A range of normal values, as determined by IntelliProfile by looking at historical data for the same time range. A baseline range displays after seven days of data collection.

n/a

Data Rate

Bytes per second sent and received though a port.

bytesTotal

Data Receive Rate

Average number of bytes per second received through a port. The data rate, compared against the Baseline of typical activity, indicates whether the current traffic is typical or out of the norm.

bytesRcvd

Data Send Rate

Also Data Xmit Rate

Average number of bytes per second transmitted through a port. The data rate, compared against the Baseline of typical activity, indicates whether the current traffic is typical or out of the norm.

bytesXmit

Frame Rate

Average number of frames per second received and transmitted through a port.

framesTotal

Frame Receive Rate

Average number of frames per second received through a port.

framesRcvd

Frame Send Rate

Also Frame Xmit Rate

Average number of frames per second transmitted through a port.

framesXmit

Link Error Rate

Average number of link errors per second on a port. Link errors are caused by or affect the link status of the connection. Link errors include:

•
Link resets
•
Loss of signal on the port
•
Loss of synchronization

If an error rate value is very small, it may be displayed as 0.0. If a fabric or switch has an actual error rate of 0, it is not displayed.

errorsLink

Link Speed

Current speed of a port in Mb/second.

currentSpeedMb

Non-Link Error Rate

Average number of non-link errors per frame on a port. Non-link errors include:

•
Resource constraint that causes frames to be retried
•
CRC errors
•
Frame length errors
•
Address errors

If an error rate value is very small, it may be displayed as 0.0. If a fabric or switch has an actual error rate of 0, it is not displayed.

nonLinkErrors

Rcvd Utilization

= Data Receive Rate / Link Speed, expressed as a percentage. Utilization values help you identify ports that may be overloaded.

bytesRcvdUtilization

Xmit Utilization

= Data Send Rate / Link Speed, expressed as a percentage. Utilization values help you identify ports that may be overloaded.

bytesXmitUtilization

Storage Arrays and Filers — Disk I/O Performance Metrics

Rate and latency metrics describe the performance of ports, controllers, pools, LUNs, disks, and NASVolumes. The selection of performance metrics displayed in any given dashboard depends on the selected device type and the metrics available from the device vendor.

 
* 
NOTE: For controllers, the Date Rate, Latency, and Ops Rate sets of metrics include Block and File versions as well to track block and file operations separately. For example, Data Rate has Data Rate (File) and Data Rate (Block). These sets of metrics are not included in this table.
 
Table 22. Disk I/O Performance Metrics
Display Name
Description
Topology Name
% Busy

Also displayed as Pct Busy or Busiest

Average percentage of time a component is busy doing I/O during the collection period.

busy

Average Queue Depth

Average number of outstanding I/O operations at the start of each new I/O request in a pool, LUN, or disk.

avgQueueDepth

Cache Hit Rate

Percentage of read operations that can be satisfied from the cache.

cacheHits

L1 Cache Hit Rate

Percentage of read operations that can be satisfied from this cache.

cacheHitRate_L1

L2 Cache Hit Rate

Percentage of read operations that can be satisfied from this cache.

cacheHitRate_L2

Data Rate

Bytes per second read and written.

bytesTotal

Data Read Rate

Bytes per second read.

bytesRead

Data Write Rate

Bytes per second written.

bytesWrite

Disk Latency

Average latency for read/write operations to the disks in a pool, calculated in milliseconds per operation. At the pool level, Avg Disk Latency reflects the average latency for I/O to the physical disks.

latencyTotalDisk

Latency

(R/W Latency)

Average latency for read/write operations, calculated in milliseconds per operation. At the pool level, Latency reflects I/O to the pool only. If the block of data to be read is already in a cache, then the latency for that operation is very low, because no disk access needs to be performed.

NOTE: Latency is not the sum of read latency and write latency. The number of read operations and write operations is different.

latencyTotal

Other Latency

Average latency for other operations

latencyOther

Read Latency

Average latency for read operations.

latencyRead

Write Latency

Average latency for write operations.

latencyWrite

Ops Rate

Average read and write operations per second.

opsTotal

Read Ops Rate

Read operations per second.

opsRead

Write Ops Rate

Write operations per second.

opsWrite

Other Ops Rate

Other operations per second.

opsOther

Related Documents

The document was helpful.

Select Rating

I easily found the information I needed.

Select Rating