Brocade Monitoring and Alerting Policy Suite Administrator Manuel d'utilisateur

53-1003147-0127 June 2014Monitoring and AlertingPolicy SuiteAdministrator's GuideSupporting Fabric OS v7.3.0

• Brocade Supplemental Support augments your existing OEM support contract, providing directaccess to Brocade expertise. For more information, contact

Default FCIP monitoring thresholds for Brocade 7840 devices (Continued)TABLE 24 Monitoring statistic UnitsMAPS thresholds and actions perpolicyActio

Default Port Health monitoring thresholds for D_Ports (Aggressive Policy)TABLE 25 Monitoring statistic Unit Threshold ActionsCRC Errors (defALL_D_PO

Default Port Health monitoring thresholds for D_Ports (Moderate Policy) (Continued)TABLE 26 Monitoring statistic Unit Threshold ActionsCRC Errors (d

Default Port Health monitoring thresholds for D_Ports (Conservative Policy) (Continued)TABLE 27 Monitoring statistic Unit Threshold ActionsSync Loss

Default Port Health monitoring thresholds for E_Ports (Continued)TABLE 28 Monitoring statisticMAPS E_Port high/low thresholds and actions per policy

Default Port Health monitoring thresholds for Host F_Ports (Continued)TABLE 29 Monitoring statistic MAPS Host F_Port high/low thresholds and actions

Default Port Health monitoring thresholds for Target F_Ports (Continued)TABLE 30 Monitoring statisticMAPS Target F_Port high/low thresholds and acti

Default Port Health monitoring thresholds for non-F_Ports (Continued)TABLE 31 Monitoring statisticMAPS non-F_Port high/low thresholds and actions pe

Default resource monitoring thresholdsTABLE 32 Monitoring statisticMAPS thresholds and actions per policyActionsAggressive Moderate ConservativeFlas

SFP monitoring thresholdsThese are the default SFP monitoring thresholds used by the Monitoring and Alerting Policy Suite(MAPS).All SFP monitoring thr

About This Document● Supported hardware and software... 11● What's

Default SFP monitoring thresholds for QSFPs and all other SFPs (Continued)TABLE 35 Monitoring statisticMAPS thresholds and actions (allpolicies)Acti

Default Switch Status Policy thresholds for the MAPS aggressive policyTABLE 37 MonitoringstatisticMAPS thresholds (Marginal/Critical) ActionsBad Pow

Default Switch Status Policy thresholds for the MAPS moderate policyTABLE 38 MonitoringstatisticMAPS thresholds (Marginal/Critical) ActionsBad Power

Default Switch Status Policy thresholds for the MAPS conservative policyTABLE 39 MonitoringstatisticMAPS thresholds (Marginal/Critical) ActionsBad P

Monitoring statistic Threshold ActionsAggressiveModerateConservativeReceive Bandwidth usage percentage (RX)60 75 90RASLOG, SNMP, EMAILTransmit Bandwid

Brocade DCX Backbone familyTABLE 2 Gen 4 platform (8-Gpbs) Gen 5 platform (16-Gbps)Brocade DCX Brocade DCX 8510-4Brocade DCX-4S Brocade DCX 8510-8Wh

Monitoring and Alerting Policy Suite Overview● MAPS overview ...

CAUTIONMAPS activation is a non-reversible process. Downgrading the switch firmware to an earlierversion of Fabric OS will enable Fabric Watch with it

Interactions between Fabric OS features and MAPSTABLE 3 Feature MAPS interactionVirtual Fabrics When using Virtual Fabrics, different logical switch

Firmware downgrade considerationsWhen downgrading from Fabric OS 7.3.0 to any previous version of the operating system, thefollowing MAPS-related beha

Differences between Fabric Watch and MAPS configurationsThe monitoring and alerting configurations available in the MAPS are not as complex as those a

Differences between Fabric Watch and MAPS configurations18 Monitoring and Alerting Policy Suite Administrator's Guide53-1003147-01

MAPS Setup and Operation● Initial MAPS setup...

© 2014, Brocade Communications Systems, Inc. All Rights Reserved.Brocade, the B-wing symbol, Brocade Assurance, ADX, AnyIO, DCX, Fabric OS, FastIron,

The following example enables MAPS, loads the policy“dflt_conservative_policy”, sets the actions to “none”, and then sets approvedactions.switch:admin

The following example enables MAPS, loads the policy “fw_aggressive_policy”,sets the actions to “none”, and then sets approved actions.switch:admin>

Both of the following cases could indicate potential issues in the fabric.Configuring rules to monitor these conditions allows you to correct issuesbe

The following example sets “dflt_moderate_policy” as the active MAPS policy.switch:admin> mapspolicy --enable -policy dflt_moderate_policyswitch:ad

Resuming MAPS monitoring24 Monitoring and Alerting Policy Suite Administrator's Guide53-1003147-01

MAPS Elements and Categories● MAPS structural elements...2

• Security Violations on page 28• Fabric State Changes on page 29• Switch Resource on page 30• Traffic Performance on page 31• FCIP Health on page 32•

Port Health category parameters (Continued)TABLE 6 MonitoredparameterDescriptionClass 3 timeouts(C3TXTO)The number of Class 3 discard frames because

The following table below lists the monitored parameters in this category. Possible states for all FRUmeasures are faulty, inserted, on, off, ready, a

Security Violations category parameters (Continued)TABLE 8 Monitored parameter DescriptionTS out of sync (SEC_TS) Time Server (TS) violations, which

ContentsPreface...7D

Fabric State Changes category parameters (Continued)TABLE 9 Monitored parameter DescriptionPercentage of devices in aFCR-enabled backbone fabric(LSA

Traffic PerformanceThe Traffic Performance category groups areas that track the source and destination of traffic. You canuse traffic thresholds and a

FCIP HealthThe FCIP Health category enables you to define rules for FCIP health, including circuit state changes,circuit state utilization, and packet

enough. To achieve this, MAPS monitors ports for the following states: IO_PERF_IMPACT andIO_FRAME_LOSS.The following table lists the monitored paramet

Switch Policy Status category parameters (Continued)TABLE 15 Monitored parameter DescriptionWWN (WWN_DOWN) Faulty WWN card (applies to modular switc

MAPS Groups, Policies, Rules, and Actions● MAPS groups overview...

logicalGroup --show fpm1 for the active Flow Vision flow “fpm1” that has been imported into, andbeing monitored through, MAPS.switch:admin> logical

Predefined MAPS groups (Continued)TABLE 16 Predefined group name Object type DescriptionALL_TUNNELS Tunnel All FCIP tunnels in the switch. This is s

Predefined MAPS groups (Continued)TABLE 16 Predefined group name Object type DescriptionALL_TUNNEL_F_QOS QoSSWITCH Switch Default group used for def

add or remove a member from the group you would have to use the logicalGroup command andspecify what you want to do (add or remove a member).To create

Switch Policy Status...33MAPS Groups, Policies, Rules, and Actions...

As an example of a dynamic definition, you could specify a port name or an attached device nodeWWN and all ports which match the port name or device n

Modifying a dynamic user-defined groupMAPS allows you to change the definition pattern used to specify a dynamic user-defined group afteryou have crea

group and then deletes the group. If a logical group is present in user-defined rules, the -force optiondeletes all the rules that are configured with

The following example restores all the deleted members and removes the addedmembers of the GOBLIN_PORTS group. First it shows the detailed view of the

defALL_D_PORTSLOSS_SYNC_3 RASLOG,SNMP,EMAIL ALL_D_PORTS(LOSS_SYNC/MIN>3) defALL_D_PORTSCRC_H90 RASLOG,SNMP,EMAIL ALL_D_PORTS(C

Fabric Watch legacy policiesWhen you migrate from Fabric Watch to MAPS, the following three policies are automatically created ifyou have used mapsCon

The following example shows the result of using the --show -summary option.switch:admin> mapsPolicy --show -summary Policy Name

The following example shows an excerpted result of using the --show -all option.The entire listing is too long (over 930 lines) to include.switch:admi

• To create a rule, enter mapsRule --create rule_name -group group_name -monitormonitored_threshold -timebase timebase -op op_value -value value -acti

The following example adds a rule to the policy named daily_policy, displays thepolicy, and then re-enables the policy so the change can become active

Additional MAPS Features...85Fabric performance mo

this threshold must be exceeded during the 60-second time base. If the counter reaches 11 within that60 seconds, the rule would trigger.NOTEMAPS condi

be issued but the port would not be fenced. To enable global actions, use the mapsConfig --actionscommands. For more details, refer to Enabling or dis

To disable all actions, enter mapsConfig --actions none. The keyword nonecannot be combined with any other action.The following example shows that RAS

SNMP MIB supportMAPS requires SNMP management information base (MIB) support on the device for managementinformation collection. For additional inform

BNA can decommission F_Ports based on CRC, ITW, PE, LR, STATE_CHG, or C3TXTO criteria.MAPS notifications are integrated with BNA, which in turn must c

The following example enables port fencing and port decommissioning for aswitch and then displays the confirmation.switch246:FID128:admin> mapsconf

The following example makes port fencing part of a rule and then displays theconfirmation.switch:admin> mapsrule --create crc_fence_Eport -group AL

The following example shows all rules on the switch. Notice that the policies arenot shown in the output.switch:admin> mapsrule --show -all -------

Creating a rule for a flowTo accommodate creating a rule for a flow, mapsRule accepts a flow name asa value for the -group parameter. The following ex

Changing multiple parametersThe following example modifies the rule “check_crc2” to generate a RASLogmessage and an e-mail message if the CRC counter

6 Monitoring and Alerting Policy Suite Administrator's Guide53-1003147-01

Cloning a rule and changing its valuesWhen you clone a rule, you can also specify the parameters you want to bedifferent from the old rule in the new

The following example shows that the rule port_test_rule35 exists intest_policy_1, deletes the rule from that policy using the -force keyword, andthen

Specifying multiple e-mail addresses for alertsThe following example specifies multiple e-mail addresses for e-mail alerts onthe switch, and then disp

1. Connect to the switch and log in using an account with admin permissions.2. Enter relayConfig --config -rla_ip relay IP address -rla_dname “relay d

The following example deletes the configured relay host address and relaydomain name for the switch, and then shows that these items have beendeleted.

Port Monitoring Using MAPS● Port monitoring and pausing... 65●

Port monitoring using port namesFabric OS allows you to monitor ports based on their assigned names.Because the port name is an editable attribute of

1. Connect to the switch and log in using an account with admin permissions.2. Enter logicalGroup --addmember group_name -member member_listThe elemen

The following example walks through the steps above for the groupALL_HOST_PORTS, first showing that port 5 is not part of the group, thenadding it to

Rules based on the ALL_D_PORTS group are part of the default policies, and have error thresholdsspanning multiple time windows or bases. If any of the

Preface● Document conventions...7● Brocade resource

Using the mapsDb --show command shows any error or rule violation duringdiagnostics tests on a D_Port.switch:admin> mapsdb --show1 Dashboard Inform

Monitoring Flow Vision Flows with MAPS● Viewing Flow Vision Flow Monitor data with MAPS... 71● Ex

‐ Number of SCSI I/O bytes read as recorded for the flow.‐ Number of SCSI I/O bytes written as recorded for the flow.For more information on Flow Visi

RASLog message is generated. If you are certain that you want to import that flow and monitor it usingthe existing rules for that flow, you must use t

Monitoring frames for a specified set of criteriaThe following example watches for frames in a flow going through a port that contain SCSI ABORTsequen

MAPS Dashboard● MAPS dashboard overview...75● MAPS dashboard

MAPS dashboard sectionsThe MAPS dashboard output is divided into three main sections: high-level dashboard information,general switch health informati

• FCIP Health on page 32• Fabric Performance Impact on page 32The following output extract shows a sample Summary Report section.3.1 Summary Report:==

information recorded since the previous midnight. The historical data log stores the last seven days onwhich errors were recorded (not the last seven

1. Connect to the switch and log in using an account with admin permissions.2. Enter mapsDb --show followed by the scope parameter: all, history, or d

Convention Descriptionvalue In Fibre Channel products, a fixed value provided as input to a commandoption is printed in plain text, for example, --sho

1. Connect to the switch and log in using an account with admin permissions.2. Enter mapsDb --show with no other parameters to display the summary sta

Sub-flow rule violation summariesIn the MAPS dashboard you can view a summary of all sub-flows that have ruleviolations.When a rule is triggered, the

1. Connect to the switch and log in using an account with admin permissions.2. Enter mapsDb --show all to display the detailed status.The following ex

Viewing historical dataTo view what has happened on a switch since the previous midnight, enter mapsDb --show history toview a summarized status histo

The following example displays historical port performance data for January 9,2014 for a chassis-based platform. Because the health status of the curr

Additional MAPS Features● Fabric performance monitoring using MAPS... 85● Scalability l

NOTENo existing bottleneck daemon logic or behaviors have been removed from Fabric OS 7.3.0.Enabling MAPS Fabric Performance Impact monitoringNOTEIf y

In the following extract, the last three lines list bottlenecks, with the final bottleneck caused by a timeoutrather than a numeric value. Be aware th

MAPS Fabric Performance Impact monitoring and legacy bottleneckmonitoringThe following conditions apply to MAPS Fabric Performance Impact (FPI) monito

Fabric State Changes |Out of operating range |No Errors |Switch Resource |In operating range |Out of operating

Brocade resourcesVisit the Brocade website to locate related documentation for your product and additional Brocaderesources.You can download additiona

The following example shows a typical RASLog entry for exceeding thethreshold for the number of Fibre Channel routers in the Backbone fabric:2014/05/2

• The “LSAN-imported device” metric is only monitored in switches that are a part of a Backbonefabric.• Scalability limits that are determined interna

Rule for LSAN device countsIn the following example, when the total device count in all switches that are part of the metaSAN(edge plus Backbone) fabr

Using only “--show”In this form, the report lists the following information for each port:• Port Number• Port type‐ D (disable port)‐ DIA (D_Port)‐ DP

Using “--show memory”The following example shows the output for mapsSam --show memory.switch:admin> mapssam --show memoryShowing Memory Usage: Me

On triggering the rules, the corresponding RASLogs will appear under the summary section of thedashboard. In the following example, there is one RASLo

Brocade 7840 FCIP monitoring using MAPS96 Monitoring and Alerting Policy Suite Administrator's Guide53-1003147-01

MAPS Threshold Values● Viewing monitoring thresholds... 97● Fabri

The following example shows all the thresholds for the ALL_D_PORTS group inthe policy named “dflt_conservative_policy”.switch:admin> mapspolicy --s

FCIP monitoring thresholdsThe following tables list the default monitoring thresholds for Fiber Channel over IP (FCIP) criteria usedby MAPS. All actio