Hitachi Command Suite 8 User Guide

							Storage SystemResourceValueProcessor ID
(for MP)The numeric value after the last hyphen (-) in the MP
column. For example, If the value of the MP column is
MPU-L0-00 , specify 00.RAID Group
NumberCharacter string with asterisk (*) wildcard characters
before and after the value of the Parity Group column.
For example, if the value of the Parity Group column is
1-3, specify *1-3*. If you want to further narrow down
the displayed information, change the wildcard character
specification.Port NameCharacter string of the Port columnPool IDNumerical value of the Pool column without "DP" at the beginning. For example, if the value of the Pool column is
"DP 10", specify 10.Universal Storage
Platform V/VM
Hitachi USPLDEV NumberValue of the Volume columnCLPR NumberValue of the CLPR columnAdaptor IDCharacter string before the second hyphen in the value of
the Processor column. For example, if the value for the
Processor column is CHA-1FL-7, specify CHA-1FL.Processor IDCharacter string after the second hyphen in the value of
the Processor column. For example, if the value for the
Processor column is CHA-1FL-7, specify 7.RAID Group
NumberCharacter string before the second hyphen in the value of
the Parity Group column, and with asterisk (*) wildcard
characters before and after. For example, if the value of
the Parity Group column is 1-2-1, specify *1-2*. If you want to further narrow down the displayed information,
change the wildcard character specification.Port NameCharacter string of the Port column.Pool IDNumerical value of the Pool column without the DP at the
beginning. For example, if the Pool column value is DP 10, specify 10.HUS 100
Hitachi AMS 2000
Hitachi SMS
Hitachi AMS/WMSLDEV NumberValue of the Volume column. If the value of the Volume column starts with 0, such as 01, specify the numerical
value without the 0 at the beginning.CLPR NumberValue of the CLPR columnControllerController value displayed in the chart titleProcessor IDCharacter string of the Processor columnRAID Group
NumberValue of the Parity Group columnUnit NumberNumerical value before the hyphen in the value of the
Drive column. For example, if the value of the Drive
column is 0-7, specify 0.HDU NumberNumerical value after the hyphen in the value of the
Drive column. For example, if the value of the Drive
column is 0-7, specify 7.Port NameCharacter string without the second character, T, in the
Port column. For example, if the value of the Port column is CTL0-B, specify CL0-B.Monitoring managed resources and resolving alerts421Hitachi Command Suite User Guide 
						

							Related concepts
•
About analyzing storage system performance  on page 403
Related tasks
•
Identifying storage system performance problems  on page 405
•
Analyzing and load balancing MP Blades or MP Units  on page 407
•
Creating periodic health check reports for storage systems performance
analysis  on page 412
Related references
•
Metrics of volumes of Virtual Storage Platform G1000, Virtual Storage
Platform, and HUS VM (Identify Performance Problems wizard)  on page
409
•
Metrics of volumes of Universal Storage Platform V/VM and Hitachi USP
(Identify Performance Problems wizard)  on page 410
•
Metrics of volumes of midrange storage systems (Identify Performance
Problems wizard)  on page 411
•
Metrics of resources of Virtual Storage Platform G1000, Virtual Storage
Platform and HUS VM (Health Check Report)  on page 416
•
Metrics of resources of Universal Storage Platform V/VM and Hitachi USP
(Health Check Report)  on page 417
•
Metrics of resources of midrange storage systems (Health Check Report) on page 418
•
Performance Reporter and Analytics tab display differences  on page 422
Performance Reporter and Analytics tab display differences The displayed report information differs between the Analytics tab and
Performance Reporter.
The following table describes how the displayed information of the Analytics
tab corresponds to that of Performance Reporter.
Table 12-8  Analytics tab and Performance Reporter display differences
Functionali
tyResourceMetricRecord
Name(Record ID)Field NameIdentify
Performance
ProblemsVolume 1IOPSVirtual Storage
Platform G1000,
Virtual Storage
Platform, Unified
Storage VM,
Universal Storage
Platform V/VM,
Hitachi USP: Logical
Device Summary
(PI_LDS, PI_LDS1,
PI_LDS2, PI_LDS3)Read I/O /sec + Write I/O /sec422Monitoring managed resources and resolving alertsHitachi Command Suite User Guide 
						

							Functionali
tyResourceMetricRecord
Name(Record ID)Field NameHUS 100, Hitachi
AMS 2000, Hitachi
SMS, Hitachi AMS/
WMS: Logical Device
Summary (PI_LDS)Pool 1Response
TimePool Summary
(PI_PLS)((Read-Response-Rate x Read I/O /
sec) + (Write-Response-Rate x Write
I/O /sec)) / (Read I/O /sec + Write
I/O /sec)IOPSPool Summary
(PI_PLS)Read I/O /sec + Write I/O /secPortBusy RatePool Summary
(PI_PTS)Avg Xfer/sec / actual-port-speed x 100CacheWrite
Pending
Rate + Side
File Usage
RateCLPR Summary
(PI_CLPS)Cache Write Pending Usage % +
Cache Side File Usage %Parity
GroupRead HitRAID Group
Summary (PI_RGS)Read Hit I/O Count / Read I/O Count
x 100Parity
GroupWrite HitRAID Group
Summary (PI_RGS)Write Hit I/O Count / Write I/O
Count x 100Health
CheckSystem
Total FE
IOPSIOPSLogical Device
Aggregation
(PI_LDA)Read I/O /sec + Write I/O /secSystem
Total FE
MB/sMB/sLogical Device
Aggregation
(PI_LDA)Read Xfer /sec + Write Xfer /secSystem
Total BE
IOPSIOPSLogical Device
Aggregation
(PI_LDA)Virtual Storage Platform G1000,
Virtual Storage Platform, HUS VM, Universal Storage Platform V/VM,Hitachi USP:
Read I/O /sec x (1 - Read Hit %) +
Write I/O /sec
HUS 100, Hitachi AMS 2000, Hitachi
SMS, Hitachi AMS/WMS:
Read I/O /sec x (1 - Read Hit %) +
Write I/O /sec x (1 - Write Hit %)Busy PG
by FE
IOPSIOPSRAID Group
Summary (PI_RGS)Read I/O /sec + Write I/O /secBusy PG
by FE
MB/sMB/sRAID Group
Summary (PI_RGS)Read Xfer /sec + Write Xfer /secBusy
Pools by
IOPSIOPSPool Summary
(PI_PLS)Read I/O /sec + Write I/O /secMonitoring managed resources and resolving alerts423Hitachi Command Suite User Guide 
						

							Functionali
tyResourceMetricRecord
Name(Record ID)Field NameBusy
Pools by
MB/sMB/sPool Summary
(PI_PLS)Read Xfer /sec + Write Xfer /secBusy PG
by BE
IOPSIOPSRAID Group
Summary (PI_RGS)Hitachi USP:
Read I/O /sec x (1 - Read Hit %) +
Write I/O /sec
Hitachi SMS, Hitachi AMS/WMS:
Read I/O /sec x (1 - Read Hit %) +
Write I/O /sec x (1 - Write Hit %)1  Response Time for volumes and pools are displayed in  microseconds in the Performance Reporter,
and is displayed in  milliseconds in the Analytics tab.
Related references
•
Configuring the Performance Reporter display  on page 420
Analyzing Universal Replicator performance This module describes procedures for analyzing Universal Replicatorperformance to identify potential problems.
About analyzing Universal Replicator performance The UR Performance Analysis window of the Replication tab provides
information for analyzing performance problems with data transfers between
the primary and secondary storage systems.
Universal Replicator (UR) asynchronously transfers data to the remote site.
Delay times occur and differ depending on the data transfer timing. This
delay is known as  C/T delta and is an indicator for Recovery Point Objective
(RPO).
You can use the UR Performance Analysis window to: • Identify instances where the C/T delta threshold was exceeded. View the top five copy groups and their C/T delta rates in a chart. You can
quickly identify copy groups that consistently exceed the maximum write time delay (C/T delta threshold).
• Analyze the C/T delta threshold against the performance of primary, secondary, and network resources. The analysis process supports two
modes:
○ Wizard mode:  a step-by-step guide that compares trends and helps
users identify the cause of the problem.
○ Advanced mode:  a selection of charts that lets advanced users
correlate multiple trends and identify the problematic resource.
424Monitoring managed resources and resolving alertsHitachi Command Suite User Guide 
						

							Related tasks
•
Monitoring Universal Replicator performance  on page 425
•
Configuring network bandwidth for analysis  on page 427
•
Analyzing Universal Replicator performance in wizard mode  on page 431
•
Analyzing Universal Replicator performance in advanced mode  on page
432
Related references
•
Prerequisites for analyzing Universal Replicator performance  on page 425
•
Metrics for Universal Replicator performance analysis  on page 429
Prerequisites for analyzing Universal Replicator performance Before attempting analysis, verify the site configuration, necessary software,
and user access controls.
You must satisfy the following requirements before using the UR Performance
Analysis window:
• Ensure that your site configuration meets all the configuration requirements for Universal Replicator performance analysis described in
the  Hitachi Command Suite Administrator Guide .
• Ensure that Hitachi Tuning Manager and Replication Manager are installed and linked to HCS.
• Ensure that users have the necessary permissions (roles and resource access) for the UR Performance Analysis window of the Replication tab
described in 
Required roles and resource groups by function on page 108 .
Note:  The performance analysis function supports copy groups consisting of
open volumes. Copy groups that are defined by device group, use multiple
journal groups in a Universal Replicator MxN configuration, or use virtual
command devices, are not supported.
Related concepts
•
About analyzing Universal Replicator performance  on page 424
Related tasks
•
Monitoring Universal Replicator performance  on page 425
Related references
•
Required roles and resource groups by function  on page 108
Monitoring Universal Replicator performance
Prerequisites
You must satisfy the following requirements before using the UR Performance Analysis window:
Monitoring managed resources and resolving alerts425Hitachi Command Suite User Guide 
						

							• Ensure that your site configuration meets all the configurationrequirements for Universal Replicator performance analysis described inthe  Hitachi Command Suite Administrator Guide .
• Ensure that Hitachi Tuning Manager and Replication Manager are installed and linked to HCS.
• Ensure that users have the necessary permissions (roles and resource access) for the UR Performance Analysis window of the Replication tab.Note:  The performance analysis function supports copy groups consisting of
open volumes. Copy groups that are defined by device group, use multiple
journal groups in a Universal Replicator MxN configuration, or use virtual
command devices, are not supported.
You can monitor Universal Replicator performance (C/T delta) from the UR Performance Analysis window. If you see problems, you can then perform an
analysis to discover the root cause.
Tip:  By default, performance data is collected every four hours, and
configuration information (including storage systems, copy pairs, and pair management servers) is refreshed every 24 hours. The Refresh Performance
Data button allows you to do a manual refresh of the performance data, and
displays an option to refresh the configuration as well (Also refresh copy
group information).
Procedure
1. Open the  Replication  tab.
2. Examine the C/T delta trend in the UR Performance Analysis window:
• The five copy groups with the largest percentage of C/T delta threshold are displayed.
• The text in the bar graph displays the actual C/T delta and threshold values.
• The default Time Period is one day. Use the range buttons to adjust the view.
• The  Copy Groups  tab includes a table with copy group information
and key usage statistics.
• For reference, the  Alerts tab includes a summary of alerts that are
created by Replication Manager.
Note:  You cannot change alerts from the  Alerts tab. You can modify
them only from Replication Manager.
3. If the C/T delta threshold for a copy group is exceeded, you can perform
an analysis:
• In  advanced mode , you can perform the analysis yourself by viewing a
series of graphs to determine the root cause of the problem.
426Monitoring managed resources and resolving alertsHitachi Command Suite User Guide 
						

							• In wizard mode , you are guided through a selection of key metrics and
presented with a summary of possible causes and solutions.Caution:  The following operations will cause previous C/T delta values
and JNLG usage rates not to be displayed:
• Deleting copy pair configuration definition files • Making changes to the pair management server that manages a target copy group, including changing the WWNNote:  You cannot access previous C/T delta values during a takeover
operation (disaster recovery) because the relationship between the primary and secondary volumes is temporarily reversed. When the copygroup relationship is restored (takeback), the data can be accessed
again. For more information about disaster recovery operations, see the
Hitachi Command Suite Replication Manager User Guide .
Related concepts
•
About analyzing Universal Replicator performance  on page 424
Related tasks
•
Analyzing Universal Replicator performance in advanced mode  on page
432
•
Analyzing Universal Replicator performance in wizard mode  on page 431
•
Configuring network bandwidth for analysis  on page 427
Related references
•
Prerequisites for analyzing Universal Replicator performance  on page 425
•
Metrics for Universal Replicator performance analysis  on page 429
Configuring network bandwidth for analysis The UR Performance Analysis function gives you the option of supplying
values for the effective network bandwidth.
The  effective  network bandwidth is the actual speed at which data can be
transmitted on a remote path based on your replication environment. Check
your network and supply a proper bandwidth value for each path group.
Tip:
• In addition to the main  UR Performance Analysis  screen, this option is
also available during analysis (wizard and advanced mode).
• Setting the bandwidth is not required. If values are not set, a warning is displayed that can be ignored.Monitoring managed resources and resolving alerts427Hitachi Command Suite User Guide 
						

							Procedure1. On the  Replication  tab, in General Tasks , select Configure
Thresholds for Network Analysis .
2. Supply one value for all remote path groups or select each individually.
Related concepts
•
About analyzing Universal Replicator performance  on page 424
Related tasks
•
Monitoring Universal Replicator performance  on page 425
Related references
•
Prerequisites for analyzing Universal Replicator performance  on page 425
•
Metrics for Universal Replicator performance analysis  on page 429
Configuring metric thresholds for M-JNL analysis The UR Performance Analysis window includes an option to set thresholds for
the following metrics:
• Host Write Transfer Rate to M-JNL • M-JNL Async Transfer Rate to RCU• Host Write IOPS to M-JNL
The threshold values are used to plot a horizontal lines in graphs indicating where the limit has been exceeded. Although defaults are defined, the valuesshould be based on your replication environment.
Tip:  In addition to appearing under General Tasks in the main screen, this
option is also available during analysis (wizard and advanced mode).
Procedure
1. On the  Replication  tab, in General Tasks , select Configure
Thresholds for M-JNL Analysis .
2. Select a metric using the M-JNL Threshold buttons.
3. Choose  All copy groups or  Select. You can select from the Available
Copy Groups and use the  Add/Remove buttons to modify the list of
Selected Copy Groups.
4. Enter a new value (or revert to the default) and click  Apply.
5. Repeat process for each of the desired metrics. Click  Close when you are
finished.
428Monitoring managed resources and resolving alertsHitachi Command Suite User Guide 
						

							Metrics for Universal Replicator performance analysisThe following table lists the metrics used to analyze Universal Replicatorperformance. See the figure for an illustration of how each metric relates to the data flow.
Table 12-9  Universal Replicator performance metrics
MetricDescriptionThreshold Value / NotesCache Write Pending
RateWrite pending rate for the
cache used by the MPB.Caution: > 30%Processor Busy RateOperating rate of the MPB.Caution: > 40%Journal (JNL) Usage RateUsage rate for the journal.This metric does not directly lead to C/T
delta degradation. The system status can
be determined by checking JNLG usage.
Secondary journal usage available in
advanced mode only.Array Port Transfer RateAmount of data output from
the array port.Caution: > network bandwidth value
See 
Configuring network bandwidth for
analysis on page 427 .
Host Write IOPS to M-
JNLAverage amount of data
input from the host to the
primary JNL per copy group
(MBs per second)Threshold value for graphs can be set as
described in 
Configuring metric thresholds
for analysis on page 428 . Default: 12,000.
Host Write Transfer Rate
to M-JNLFrequency of write
operations for the primary
JNL per unit of timeThreshold value for graphs can be set as
described in 
Configuring metric thresholds
for analysis on page 428 . Default: 250.
P-VOL Write Transfer
RateAmount of data transmitted
from the host or host group
per P-VOL (MBs per
second).Available in advanced mode only.
See 
About the P-VOL Write Transfer Rate
metric on page 435  for additional
information.
M-JNL Async Transfer
Rate to RCUAverage rate of transmission
from M-JNL to R-JNL (MBs
per second).Threshold value for graphs can be set as
described in 
Configuring metric thresholds
for analysis on page 428 . Default: 250.
Read M-JNL Process
TimeInterval between when the
MCU receives the RDJNL
command and the MCU
sends a data transmission
completion command to the
RCU (msec).Available in advanced mode only.End-to-End Journal Copy
TimeInterval between when the
RCU sends the RDJNL
command to the MCU and
the RCU receives a data
transmission completion
command (msec).Available in advanced mode only.R-JNL Transfer and
Process TimeThe time lag in network
transmission between M-JNL
and R-JNL.This value represents the difference
between the End-to-End Journal Copy
Time and Read M-JNL Process Time. (If
one of these values is not available, the R-Monitoring managed resources and resolving alerts429Hitachi Command Suite User Guide 
						

							MetricDescriptionThreshold Value / NotesJNL Transfer and Process Time value is
undefined.)
The following figure depicts points in the data flow associated with key
metrics. This is not intended as a master reference for all metrics, but serves
as an aid for performing an analysis in advanced mode. For example, the
expanded portion shows the relationship between the Read M-JNL Process
Time, R-JNL Transfer and Process Time, and the End-to-End Journal Copy
Time.
Related concepts
•
About analyzing Universal Replicator performance  on page 424
430Monitoring managed resources and resolving alertsHitachi Command Suite User Guide