EMC Storage Analytics

Transcription

1 EMC Storage Analytics Version 2.1 Installation and User Gide

2 Copyright 2013 EMC Corporation. All rights reserved. Pblished in USA. Pblished December, 2013 EMC believes the information in this pblication is accrate as of its pblication date. The information is sbject to change withot notice. The information in this pblication is provided as is. EMC Corporation makes no representations or warranties of any kind with respect to the information in this pblication, and specifically disclaims implied warranties of merchantability or fitness for a particlar prpose. Use, copying, and distribtion of any EMC software described in this pblication reqires an applicable software license. EMC², EMC, and the EMC logo are registered trademarks or trademarks of EMC Corporation in the United States and other contries. All other trademarks sed herein are the property of their respective owners. For the most p-to-date reglatory docment for yor prodct line, go to EMC Online Spport ( For docmentation on EMC Data Domain prodcts, go to the EMC Data Domain Spport Portal ( EMC Corporation Hopkinton, Massachsetts In North America EMC Storage Analytics 2.1 Installation and User Gide

3 CONTENTS Tables 5 Figres 7 Chapter 1 Introdction 9 Overview References Terminology Chapter 2 Installation and licensing 13 Installation overview Installation reqirements Installing vcenter Operations Manager for EMC Storage Installing the EMC Adapter and dashboards Adapter instances Adding an EMC Adapter instance for vcenter Adding EMC Adapter instances for yor storage system...24 Editing EMC Adapter instances for yor storage system...27 Chapter 3 EMC Storage Analytics Dashboards 29 Topology mapping...30 VNX Block topology...31 VNX File topology VMAX topology...33 VMAX topology rles EMC dashboards Storage Topology dashboard...35 Storage Metrics dashboard VNX Overview dashboard VMAX Overview dashboard Topology dashboards...44 Metrics dashboards Top-N dashboards...49 Dashboard XChange...52 Chapter 4 Resorce Kinds and Metrics 53 Resorce Kinds VNX Block metrics VNX File metrics VMAX metrics...69 Chapter 5 Trobleshooting 73 Lanching Unisphere Installation logs EMC Storage Analytics 2.1 Installation and User Gide 3

4 CONTENTS Error handling and event logging Viewing error logs Creating and downloading a spport bndle Editing the Collection Interval for a resorce...78 Freqently Asked Qestions - General Freqently Asked Qestions for VNX...83 Freqently Asked Qestions for VMAX Index 87 4 EMC Storage Analytics 2.1 Installation and User Gide

5 TABLES Installation and Upgrade options...15 VNX Block metrics for Array...55 VNX Block metrics for Storage Processor...55 VNX Block metrics for FAST Cache...56 VNX Block metrics for Pool LUN...57 VNX Block metrics for RAID Grop LUN VNX Block metrics for RAID Grop VNX Block metrics for Storage Pool VNX Block metrics for Tier VNX Block metrics for Disk...62 VNX File metrics for Array VNX File metrics for Data Mover VNX File metrics for dvol...67 VNX File metrics for File Pool VNX File metrics for File System VMAX metrics for Storage Grop VMAX metrics for Device VMAX metrics for Thin Pool VMAX metrics for Front-End Director...70 VMAX metrics Front-End Port...70 VMAX metrics SRDF Director...70 VMAX metrics Remote Replica Grop EMC Storage Analytics 2.1 Installation and User Gide 5

6 TABLES 6 EMC Storage Analytics 2.1 Installation and User Gide

7 FIGURES EMC adapter architectre...10 VNX Block topology VNX File topology...32 VMAX topology Storage Topology dashboard Storage Metrics dashboard...38 VNX Overview dashboard...40 VMAX Overview...42 Example of VNX Topology dashboard...45 Example of VMAX Topology dashboard Example of VNX Metrics dashboard Example of VMAX Metrics dashboard...48 Example of VMAX Top-5 Devices Example of Top-5 VNX File Systems...50 Example of Top-5 VNX LUNs...51 Dashboard XChange Example of lanching to Unisphere...74 Log Content Stats shows: Resorce down Workload shows: No Data...81 VM Resorces shows:?...82 EMC Storage Analytics 2.1 Installation and User Gide 7

8 FIGURES 8 EMC Storage Analytics 2.1 Installation and User Gide

9 CHAPTER 1 Introdction This chapter contains the following topics. Overview References Terminology Introdction 9

10 Introdction Overview VMware vcenter Operations Manager is a software prodct that collects performance and capacity data from monitored software and hardware resorces. It provides sers with real time information abot potential problems in the enterprise. vcenter Operations Manager presents data and analysis in several ways: Throgh alerts that warn of potential or occrring problems. In configrable dashboards and predefined pages that show commonly needed information. In predefined reports. EMC Storage Analytics links vcenter Operations Manager with an EMC Adapter. The adapter is bndled with a connector that enables vcenter Operations Manager to collect performance metrics. The adapter is installed with the vcenter Operations Manager ser interface. The connector types are shown in Figre 1 on page 10. EMC Storage Analytics leverages the power of existing vcenter featres to aggregate data from mltiple sorces and process the data with proprietary analytic algorithms. Figre 1 EMC adapter architectre vc Ops vapp vc Ops UI VM vc Ops Analytics VM EMC Adapter Cstom Dashboards Merge and Proxy (MOXY) Adapter Framework Resorce Discovery Metric Collection Services Event Processing Connectors vcenter Array Registration Topology Mapping VNX Block VNX File VMAX 10 EMC Storage Analytics 2.1 Installation and User Gide

11 Introdction References Terminology This topic provides a list of vcenter Operations Manager docmentation. VMware vcenter Operations Manager docmentation incldes: vcenter Operations Manager Release Notes. Contains descriptions of known isses and workaronds. vcenter Operations Manager vapp Deployment and Configration Gide. Explains installation, deployment, and management of vcenter Operations Manager. vcenter Operations Manager Getting Started Gide. Explains basic featres and se of vcenter Operations Manager. vcenter Operations Manager Administration Gide. Describes how to configre and manage the vcenter Operations Manager cstom interface. vcenter Operations Manager Adapter Gide. Provides information on how vcenter Operations Manager ses adapters to consme data from the enterprise. VMware docmentation is available at This topic contains a list of commonly sed terms. adapter A vcenter Operations Manager component that collects performance metrics from an external sorce like a vcenter or storage system. Third-party adapters sch as the EMC Adapter are installed on the vcenter Operations Manager server to enable creation of adapter instances within vcenter Operations Manager. adapter instance A specific external sorce of performance metrics, sch as a specific storage system. An adapter instance resorce is an instance of an adapter that has a one-to-one relationship with an external sorce of data, sch as a VNX storage system. dashboard A tab on the home page of the vcenter Operations Manager GUI. vcenter Operations Manager ships with defalt dashboards. Dashboards are also flly cstomizable by the end ser. health rating An overview of the crrent state of any resorce, from an individal operation to an entire enterprise. vcenter Operations Manager checks internal metrics for the resorce and ses its proprietary analytics formlas to calclate an overall health score on a scale of 0 to 100. metric A category of data collected for a resorce. For example, the nmber of read operations per second is one of the metrics collected for each LUN resorce. resorce Any entity in yor environment for which vcenter Operations Manager can collect data. For example, LUN 27 is a resorce. resorce kind A general type of a resorce, sch as LUN or DISK. The resorce kind dictates the type of metrics collected. References 11

12 Introdction 12 EMC Storage Analytics 2.1 Installation and User Gide

13 CHAPTER 2 Installation and licensing This chapter contains the following topics. Installation overview Installation reqirements Installing vcenter Operations Manager for EMC Storage Installing the EMC Adapter and dashboards Adapter instances Installation and licensing 13

14 Installation and licensing Installation overview EMC Storage Analytics 2.1 consists of two installation packages. The two packages are: vcenter Operations Manager for EMC Storage EMC Adapter vcenter Operations Manager for EMC Storage is a storage only version of vcenter Operations Manager 5.8 developed specifically for EMC storage cstomers. Installation instrctions are in the vcenter Operations Manager vapp Deployment and Configration Gide at Licensing depends on which version of the vcenter Operations Manager yo have installed: VMware isses the license for vcenter Operations Manager Advanced and Enterprise versions. EMC isses the license for vcenter Operations Manager for EMC Storage. NOTICE If yo have vcenter Operations Manager Advanced or Enterprise versions with a VMwareissed license, do not install the EMC-issed license for vcenter Operations Manager; installing the EMC-issed license restricts vcenter Operations Manager to storage-only mode. A 90-day trial license is available with EMC Storage Analytics 2.1 for all spported prodcts. The 90-day trial license provides the same featres as a permanent license, bt after 90 days of sage, the adapter stops collecting data. Yo can add a permanent license at any time. The EMC Adapter collects metrics from resorces. It spports vcenter Operations Manager for EMC Storage and the VMware vcenter Operations Manager Advanced and Enterprise versions. EMC Adapter 2.1 spports vcenter Operations Manager versions and higher. The adapter installation incldes: Installing the adapter and dashboards. Configring the adapter instances with licensing. Installing the EMC Adapter and dashboards on page 20 provides installation instrctions. The adapter is licensed when Adding EMC Adapter instances for yor storage system on page 24 or Editing EMC Adapter instances for yor storage system on page 27. Table 1 on page 15 lists the installation and pgrade options for different vcenter Operations Manager and the EMC Adapter combinations. To perform an installation or pgrade, refer to the instrctions for the task that matches yor objective. 14 EMC Storage Analytics 2.1 Installation and User Gide

15 Installation and licensing Table 1 Installation and Upgrade options Task Instrctions - - Install VMware vcenter Operations Manager 5.8 with EMC Adapter 2.1. vcenter Operations Manager vapp Deployment and Configration Gide at pbs/vcops-pbs.html Installing the EMC Adapter and dashboards on page 20 Adding EMC Adapter instances for yor storage system on page 24 Upgrade EMC Adapter 2.0 to EMC Storage Analytics 2.1 on a system rnning VMware vcenter Operations Manager or later. Install EMC Adapter 2.1 on a system rnning VMware vcenter Operations Manager 5.8. Install vcenter Operations Manager for EMC Storage and EMC Adapter 2.1. Installing the EMC Adapter and dashboards on page 20 Editing EMC Adapter instances for yor storage system on page 27 Installing the EMC Adapter and dashboards on page 20 Adding EMC Adapter instances for yor storage system on page 24 Installing vcenter Operations Manager for EMC Storage on page 19 Installing the EMC Adapter and dashboards on page 20 Adding EMC Adapter instances for yor storage system on page 24 Installation overview 15

16 Installation and licensing Installation reqirements Before installing the EMC Adapter, verify that hardware and software component compatibility reqirements are satisfied. EMC Adapter reqirements Spported vcenter Operations Manager version vcenter Operations Manager and higher vcenter Operations Manager for EMC Storage EMC Storage Analytics reqires vcenter Operations Manager Advanced or Enterprise editions. EMC Storage Analytics does not spport vcenter Operations Manager Fondation and Standard editions. Note vcenter Operations Manager and higher is spported with EMC Adapter 2.1, bt vcenter Operations Manager 5.8 is recommended. Deploy the vapp for vcenter Operations Manager before installing the EMC Adapter. Check the vcenter Operations Manager vapp Deployment and Configration Gide at for system reqirements pertaining to yor version of vcenter Operations Manager. The deployment instrctions for vcenter Operations Manager 5.8 apply to vcenter Operations Manager for EMC Storage. Spported VNX series (1st generation) VNX5100 VNX5300 VNX5500 VNX5700 VNX7500 VG2 VG8 Spported VNX series (2nd generation) VNX5200 VNX5400 VNX5600 VNX5800 VNX7600 VNX EMC Storage Analytics 2.1 Installation and User Gide

17 Installation and licensing Spported CX series CX4-120 CX4-240 CX4-480 CX4-960 Spported NS series NS120 NS480 NS960 Spported VMAX series VMAX (all models) VMAXe VMAX 10K VMAX 20K VMAX 40K Spported web browser Internet Explorer 8 or 9. Firefox 10.0 or later. Google Chrome version 19 or later. VNX Block systems The EMC Adapter ses naviseccli to collect metrics from VNX Block systems. It is bndled into the EMC Adapter install file and is atomatically installed along with the adapter. Storage processors reqire IP addresses that are reachable from the vcenter Operations Manager server. Bidirectional traffic for this connection flows throgh port 443 (HTTPS). Statistics logging mst be enabled on each storage processor (SP) for metric collection (System > System Properties > Statistics Logging in Unisphere). VNX File systems CLI commands issed on the Control Station direct the EMC Adapter to collect metrics from VNX File systems. The Control Station reqires an IP address that is reachable from the vcenter Operations Manager server. Bidirectional ethernet traffic flows throgh port 22 sing Secre Shell (SSH). If yo are sing the EMC VNX nas_stig script for secrity (/nas/tools/nas_stig), do not se root in password credentials. Setting nas_stig to ON limits direct access for root acconts, preventing the adapter instance from collecting metrics for VNX File. SMI-S Provider for VMAX reqirements The EMC SMI-S Provider for VMAX mst be on the network and accessible throgh port 5988 (non-secre port) or port 5989 (secre port) for commnication. If the SMI- S Provider is also sed for vsphere vstorage API for Storage Awareness (VASA), follow the recommendations in the SMI-S Provider Release Notes to increase the nmber of available connections. The ser configred in the EMC Adapter instance and connecting to the SMI-S instance mst have the role of monitor or administrator. Installation reqirements 17

18 Installation and licensing VMAX software reqirements VMAX software reqirements inclde Enginity 5876 or higher and Soltions Enabler (for SMI-S Provider) or higher. Minimm OE reqirements Platform Spported OE versions - - VNX 1st generation series Block: R and higher R and higher File: and higher and higher VNX 2nd generation series CX4 series NS series Block: R and higher File: and higher R and higher and higher User acconts To create an EMC Adapter instance for a storage array, yo mst have a ser accont that allows yo to connect to the storage array or SMI-S Provider. For example, to add an EMC Adapter for a VNX array, se a global accont with operator or administrator role (a local accont will not work). To create an EMC Adapter instance for vcenter (where Adapter Kind = EMC Adapter and Connection Type = VMware vsphere), yo mst have an accont that allows yo access to vcenter and the objects it monitors. In this case, vcenter enforces access credentials (not the EMC Adapter). To create an EMC Adapter instance for vcenter, se, at minimm, an accont assigned to the Read-Only role at the root of vcenter, and enable propagation of permissions to descendant objects. Depending on the size of the vcenter, wait approximately 30 seconds before testing the EMC Adapter. More information on ser acconts and access rights is available in the vsphere API/SDK Docmentation available at: index.jsp?topic=%2fcom.vmware.wssdk.pg.doc%2fpg_athenticate_athorize. 8.3.html. DNS configration To se the EMC Adapter, the vcenter Operations Manager vapp reqires network connectivity to the storage systems to be monitored. DNS mst be correctly configred on the vcenter Operations Manager server to enable hostname resoltion by the EMC Adapter. Time zone and synchronization settings Ensre time synchronization for all EMC Storage Analytics resorces by sing Network Time Protocol (NTP). Also, set correct time zones for EMC Storage Analytics resorces (inclding the SMI-S Provider if sing an adapter for VMAX) and related systems. Failre to observe these practices may affect the collection of performance metrics and topology pdates. 18 EMC Storage Analytics 2.1 Installation and User Gide

19 Installation and licensing Installing vcenter Operations Manager for EMC Storage If no vapp for vcenter Operations Manager is deployed, install the vcenter Operations Manager for EMC Storage. Before yo begin Pre-reqisites: Obtain the OVA installation package for vcenter Operations Manager 5.8 from VMware or vcenter Operations Manager for EMC Storage. Obtain a vcenter Operations Manager for EMC Storage license key. Obtain a copy of the vcenter Operations Manager vapp Deployment and Configration Gide at To deploy the vapp for vcenter Operations Manager, refer to the vcenter Operations Manager vapp Deployment and Configration Gide. Procedre 1. Review the System Reqirements. 2. Follow instrctions in Installing vcenter Operations Manager. When assigning the vcenter Operations Manager license, type the vcenter Operations Manager for EMC Storage license from the EMC Storage Analytics Software Key Card. 3. Conclde the installation with Verifying the vcenter Operations Manager Installation. Installing vcenter Operations Manager for EMC Storage 19

20 Installation and licensing Installing the EMC Adapter and dashboards Before yo begin Pre-reqisites: Obtain the PAK file for the EMC Adapter. Note If sing Internet Explorer, the installation file downloads as a ZIP file bt fnctions the same way as the PAK file. WARNING Upgrading EMC Storage Analytics overwrites the standard dashboards. To cstomize a standard dashboard, clone it, rename it, and then cstomize it. To install the adapter and dashboards: Procedre 1. Save the PAK file in a temporary folder. 2. Start the vcenter Operations Manager administrative ser interface in yor web browser and log in as administrator. For example, 3. On the Update tab, click Browse to locate the adapter PAK file. 4. Click Update. The stats of the pdate process appears in the bottom pane. Note Installation of the PAK file can take several mintes or longer depending on the amont of metric data that already exists in the vcenter Operations Manager database. The vcenter Operations Manager will restart services following installation of the PAK file. 20 EMC Storage Analytics 2.1 Installation and User Gide

21 Installation and licensing Adapter instances The vcenter Operations Manager reqires an adapter instance for each resorce to be monitored. The instance specifies the type of adapter to se and the information needed to identify and access the resorce. With EMC Storage Analytics, the vcenter Operations Manager ses EMC Adapter instances to identify and access the resorces, sch as: Adapter instance for vcenter Adapter instance for the VNX File Adapter instance for the VNX Block Adapter instance for VMAX The adapter instance for vcenter is a prereqisite for other adapter instances. If this adapter is not configred, other adapter instances will fnction normally bt will not display visible connections between the VMware objects and the array objects. Note After adapter instances are created, the vcenter Operations Manager Collector reqires several mintes to collect statistics, depending on the size of the storage array configration. Large configrations can take p to 45 mintes to collect metrics and resorces and pdate dashboards. This is a one-time event; ftre statistical collections rn qickly. Adding an EMC Adapter instance for vcenter For sers to traverse health trees from the virtal environment into the storage environment, EMC Storage Analytics reqires yo to install an EMC Adapter instance for vcenter. All storage system adapter instances reqire the EMC Adapter instance for vcenter, which yo mst add first. To add an EMC Adapter instance for vcenter: Procedre 1. Start the vcenter Operations Manager cstom ser interface and log in as admin. For example in a web browser, type: 2. Select Environment > Configration > Adapter Instances. 3. Click the Add New Adapter Instance icon. The Add Adapter Instance screen appears. 4. Set the following: Collector Adapter Kind Adapter Instance Name Management IP Array ID (optional) Connection Type vcenter Operations Standard Server EMC Adapter Any descriptive name. For example, My vcenter. IP address of the vcenter server. This field mst be blank for VMware vsphere connection type. VMware vsphere Adapter instances 21

22 Installation and licensing License (optional) Log Level Credential Leave blank. Not applicable for EMC Adapter instance for vcenter. Configre log levels for each adapter instance. The for levels for logging information are ERROR, WARN, INFO, and DEBUG. ERROR The ERROR level logs only error conditions. The ERROR level provides the smallest amont of logging information. WARN The WARN level logs information when an operation completes sccessflly bt there are isses with the operation. INFO The INFO level logs information abot workflow. It describes how an operation occrs. DEBUG The DEBUG level logs all details related to an operation. This level is the highest level of logging. If logging is set to DEBUG, all other levels of logging information are displayed in the log file. Select any previosly defined credentials for this storage system, otherwise click Add. The Add Credential screen appears with the Adapter Kind set to EMC Adapter. 5. Set the following: Credential kind Instance name Username Password Password Credentials Any descriptive name. For example, My VMware Credentials. Username that EMC Storage Analytics ses to connect to the VMware vcenter system. If a domain ser is sed, the format for the sername is DOMAIN\USERNAME. Password for the EMC Storage Analytics sername on the VMware vcenter system. 22 EMC Storage Analytics 2.1 Installation and User Gide

23 Installation and licensing 6. Click OK. The Add Adapter Instance screen reappears. 7. To test the adapter instance, click Test. If the connection is correctly configred, a confirmation box appears. 8. Click OK. The new adapter instance polls for data every 5 mintes by defalt. At every interval, the adapter instance will collect information abot the VMware vsphere datastore and virtal machines with Raw Device Mapping (RDM). Consmers of the registered VMware service can access the mapping information. To edit the polling interval, select Environment > Environment Overview > Resorce Kinds > EMC Adapter Instance > All Atrribtes. Select the EMC Adapter instance and click the Edit Resorce icon. Adding an EMC Adapter instance for vcenter 23

24 Installation and licensing Adding EMC Adapter instances for yor storage system Before yo begin Pre-reqisites: Install the EMC Adapter for vcenter Obtain the adapter license key for yor storage system Each storage system reqires an adapter instance. All storage system adapter instances reqire the EMC Adapter instance for vcenter. Add the EMC Adapter instance for vcenter first. Then add the adapter instances for each storage system. Adapter instances are licensed per array. Note A VNX Unified array can se the same license for VNX File and VNX Block. Procedre 1. Start the vcenter Operations Manager cstom ser interface and log in as administrator. For example in a web browser, type: 2. Select Environment > Configration > Adapter Instances. The Manage Adapter Instances screen appears. 3. Click the Add New Adapter Instance icon. The Add Adapter Instance screen appears. 24 EMC Storage Analytics 2.1 Installation and User Gide

25 Installation and licensing 4. Set the following: Field Vale to enter - - Collector vcenter Operations Standard Server Adapter Kind Adapter Instance Name License (optional) Log Level Credential EMC Adapter Any descriptive name. For example, My Storage System or the array ID. License key reqired for the array. Verify that the license is for the array yo want to monitor. Configre log levels for each adapter instance. The for levels for logging information are ERROR, WARN, INFO, and DEBUG. ERROR The ERROR level logs only error conditions. The ERROR level provides the smallest amont of logging information. WARN The WARN level logs information when an operation completes sccessflly bt there are isses with the operation. INFO The INFO level logs information abot workflow. It describes how an operation occrs. DEBUG The DEBUG level logs all details related to an operation. This level is the highest level of logging. If logging is set to DEBUG, all other levels of logging information are displayed in the log file. Select any previosly defined credentials for this storage system, otherwise, click Add. Field VNX Block VNX File VMAX Management IP Use the IP address of one Storage Processor (SP) in a single array. Do not add an adapter instance for each SP. Use the IP address of the primary Control Station. Use the IP address of the configred SMI-S Provider. Array ID (optional) Not applicable Not applicable Reqired with prefixed zeros. Example: Connection Type VNX Block (se for VNX Block and CX4 arrays) VNX File (se for VNX File/Unified models, VG gateways, and NS models) VMAX (se for VMAX models) The Add Credential screen appears with the Adapter Kind set to EMC Adapter. 5. Set the following: Adding EMC Adapter instances for yor storage system 25

26 Installation and licensing Field Vale to enter - - Adapter kind EMC Adapter. Retain the setting. Credential kind Instance name Username Password Credentials. Any descriptive name. For example, Defalt Credentials. Username that EMC Storage Analytics ses to connect to the storage system. For VNX File, se the credentials of the Control Station. For VNX Block, se the credentials of the Storage Processors. For VMAX, se the credentials of an ECOM ser with monitor or administrator privileges. The defalt ser/password combination is admin/#1password. Password Password for the EMC Storage Analytics sername on the storage system. 6. Click OK. The Add Adapter Instance screen reappears. 7. Click Test to validate the vales yo entered. If the adapter instance is correctly configred, a confirmation box appears. NOTICE Testing an adapter instance validates the vales yo entered. Failre to do this step cases the adapter instance to change to the red (warning) state if yo enter invalid vales and do not validate them. 8. To finish adding the adapter instance, click OK. 26 EMC Storage Analytics 2.1 Installation and User Gide

27 Installation and licensing Editing EMC Adapter instances for yor storage system Before yo begin Pre-reqisites: Install the EMC Adapter Configre the EMC Adapter instance for yor storage system Obtain an adapter license key for yor storage system The EMC Adapter instances for storage systems reqire licenses. Adapter instances are licensed per storage array. A VNX Unified array can se the same license for VNX File and VNX Block. Procedre 1. Start the vcenter Operations Manager cstom ser interface and log in as administrator. For example in a web browser, type: 2. Select Environment > Configration > Adapter Instances. The Manage Adapter Instances screen appears. 3. Select the row of an EMC Adapter instance (Adapter Kind is EMC Adapter), and click the Edit Selected Adapter Instance icon. The Edit Adapter Instance screen appears. Editing EMC Adapter instances for yor storage system 27

28 Installation and licensing 4. Add the license from the EMC Storage Analytics Software Key Card. The EMC Adapter instance license key is issed for the storage system. Verify that the Management IP is for the same array model specified on the key card. 5. To test the configration, click Test. If the adapter instance is correctly configred, a confirmation box appears. 6. To finish editing the adapter instance, click OK. 28 EMC Storage Analytics 2.1 Installation and User Gide

29 CHAPTER 3 EMC Storage Analytics Dashboards This chapter contains the following topics. Topology mapping...30 EMC dashboards EMC Storage Analytics Dashboards 29

30 EMC Storage Analytics Dashboards Topology mapping Topology mapping is viewed and traversed graphically sing vcenter Operations Manager health trees. The dashboards developed for EMC Storage Analytics tilize topology mapping to display resorces and metrics. EMC Storage Analytics establishes mappings between: Storage system components Storage system objects and vcenter objects Topology mapping enables health scores and alerts from storage system components, sch as storage processors and disks, to appear on affected vcenter objects, sch as LUNs, datastores, and VMs. Topology mapping between storage system objects and vcenter objects ses a vcenter adapter instance. 30 EMC Storage Analytics 2.1 Installation and User Gide

31 EMC Storage Analytics Dashboards VNX Block topology EMC Storage Analytics implements the following topology for VNX Block. Figre 2 VNX Block topology VMware Host Array Instance Datastore Datastore Datastore LUN LUN LUN LUN SP A SP B Pool RAID Grop Fast Cache Tier Tier Tier Disk Disk Disk Disk Disk Disk Disk Disk Disk Disk VNX Block topology 31

32 EMC Storage Analytics Dashboards VNX File topology EMC Storage Analytics implements the following topology for VNX File. Figre 3 VNX File topology VMware Host Physical Host Datastore Datastore Array Inatance NFS Export NFS Export NFS Export File system File system File system Virtal Data Mover Data Mover (standby) Data Mover Pool for File Pool for File Data Mover dvol dvol dvol LUN LUN LUN 32 EMC Storage Analytics 2.1 Installation and User Gide

33 EMC Storage Analytics Dashboards VMAX topology EMC Storage Analytics implements the following topology for VMAX. Figre 4 VMAX topology VMware Datastore VNX dvol VMAX Array Storage Grop Storage Grop Remote Replica Grop Remote Replica Grop Device Device Device Device Device Device SRDF Director Device Device SRDF Director Front-End Director Thin Pool Front-End Director Thin Pool Front-End Port Front-End Port Front-End Port Front-End Port VMAX topology 33

34 EMC Storage Analytics Dashboards VMAX topology rles The following rles govern how objects are displayed in the VMAX topology dashboard and which metrics are collected for them. vcenter Operations Manager does not display devices that are nmapped and nbond. vcenter Operations Manager does not display devices that are mapped and bond bt nsed by VMware or VNX. Performance metrics for these devices are aggregated into the parent Storage Grop performance metrics. If the corresponding EMC vsphere adapter instance is rnning on the same vcenter Operations Manager appliance, then the vcenter Operations Manager displays devices that are mapped, bond, and sed by VMware datastores or RDMs. If the corresponding EMC VNX File adapter instance is rnning on the same vcenter Operations Manager appliance, then the vcenter Operations Manager displays devices that are mapped, bond, and sed by VNX File Disk Volmes. vcenter Operations Manager does not display Storage Grops with nmapped and nbond devices. vcenter Operations Manager displays Storage Grops that contain mapped and bond devices, and their metrics are aggregates of the member devices. 34 EMC Storage Analytics 2.1 Installation and User Gide

35 EMC Storage Analytics Dashboards EMC dashboards Use dashboards to view metrics. The standard dashboards are delivered as templates - if a dashboard is accidentally deleted or changed, yo can generate a new one. EMC Storage Analytics incldes these EMC dashboards: Storage Topology Storage Metrics VNX Overview Storage Topology dashboard VMAX Overview Topology dashboards Metrics dashboards Top-N dashboards Yo can se the standard vcenter Operations Manager dashboard cstomization featres to create additional dashboards that are based on yor site reqirements. The Storage Topology dashboard provides an entry point for viewing resorces and relationships between storage and virtal infrastrctre objects. Click the Storage Topology tab. Details for every object in every widget are available by selecting the object and clicking the Resorce Detail icon at the top of each widget circled in red in Figre 5 on page 36. EMC dashboards 35

36 EMC Storage Analytics Dashboards Figre 5 Storage Topology dashboard 36 EMC Storage Analytics 2.1 Installation and User Gide

37 EMC Storage Analytics Dashboards The Storage Topology dashboard contains the following widgets: Storage System Selector This Resorce widget filters the EMC Adapter instances that are fond in each storage system. To poplate the Storage Topology and Health widget, select an instance name. Storage Topology and Health This Health Tree widget provides a navigable visalization of resorces and virtal infrastrctre resorces. Single-click to select resorces, or doble-click to change the navigation focs. To poplate the Parent Resorces and Child Resorces widgets, select a resorce in this widget. Parent resorces This widget lists the parent resorces of the resorce selected in the Storage Topology and Health widget. Child resorces This widget lists the child resorces of the resorce selected in the Storage Topology and Health widget. Storage Topology dashboard 37

38 EMC Storage Analytics Dashboards Storage Metrics dashboard The Storage Metrics dashboard displays resorce and metrics for storage systems and allows the ser to view graphs of resorce metrics. Figre 6 Storage Metrics dashboard The Storage Metrics dashboard contains the following widgets: Storage System Selector This Resorce widget lists all configred EMC Adapter instances. To poplate the Resorce Selector widget, select an instance name. Resorce Selector This Health Tree widget lists each resorce associated with the adapter instance selected in the Storage System Selector. To poplate the Storage System Selector widget, select a resorce. Metric Selector This widget lists all the metrics that are collected for the resorce selected in the Resorce Selector widget. The search featre of this widget allows sers to locate specific objects. Doble-click a metric to create a graph of the metric in the Metric Graph widget. 38 EMC Storage Analytics 2.1 Installation and User Gide

39 EMC Storage Analytics Dashboards Metric Graph This widget graphs the metrics selected in the Metric Selector widget. Display mltiple metrics simltaneosly in a single graph or in mltiple graphs. Storage Metrics dashboard 39

40 EMC Storage Analytics Dashboards VNX Overview dashboard This collection of heat maps provides a single view of the performance and capacity for all VNX resorces with configred adapter instances. Heat maps on this dashboard grop the contents by adapter instance. Figre 7 VNX Overview dashboard The VNX Overview dashboard displays the following heat maps: CPU performance This displays the CPU Utilization, sch as % bsy, of each Storage Processor and Data Mover on each configred adapter instance. The color of the heat map entries shows % bsy: Green indicates 0% bsy. Red indicates 100% bsy. FAST cache performance This has two modes: Read Hit Ratio and Write Hit Ratio. To select the mode, se the Configration men. The Read/Write Cache Hit Ratio (%) is the nmber of FAST Cache read or write hits divided by the total nmber of read or write I/Os across all RG LUNs and Pools configred to se FAST Cache. The color of the heat map entries shows hit ratios: Green indicates a high FAST Cache hit ratio. Red indicates a low FAST Cache hit ratio. A low vale on an idle array is acceptable. Grey indicates that there is no FAST Cache present on the VNX systems identified by the adapter instances and a Heat Map not configred message appears with the heat map. 40 EMC Storage Analytics 2.1 Installation and User Gide

41 EMC Storage Analytics Dashboards Pool capacity This has for modes: RAID Grop Available Capacity, Storage Pool Capacity Utilization, Storage Pool Available Capacity, and File Pool Available Capacity. In Capacity Utilization mode, the color of the heat map entries shows the vale of the % fll metric for all non-raid Grop storage pools: Green indicates 0% fll. Red indicates 100% fll. In Available Capacity mode, the color of the heap map entries shows the vale of the Available Capacity (GB) metric: Green indicates the largest available capacity on any storage pool for any of the configred adapter instances. Red indicates 0 GB available. LUN and file system performance This has several modes. In LUN Utilization mode, the color of the heat map entries show the % bsy metric for all LUNs groped by adapter instance: Green indicates 0% bsy. Red indicates 100% bsy. In RAID Grop LUN Latency mode, the color of the heat map entries shows the vale of the Latency (ms) metric: Green indicates 0 ms latency. Red indicates 20 ms or greater latency and is configrable. Latency vales appear for RAID Grop LUNs. Pool LUNS appear in white with no latency vales reported. In LUN Read IOPs mode, the color of the heat map entries shows the relative nmber of read IO operations per second serviced by the LUN. The color ranges from light green to dark green. Dark green indicates the highest nmber of read IO operations per second serviced by any LUN listed in the heat map. In LUN Write IOPS mode, the color of the heat map entries shows the relative nmber of write IO operations per second serviced by the LUN. The color ranges from light green to dark green. Dark green indicates the highest nmber of write IO operations per second serviced by any LUN listed in the heat map. In File System Read IOPs mode, the color of the heat map shows the relative nmber of read IO operations per second serviced by the file system. The color ranges from light green to dark green. Dark green indicates the highest nmber of read IO operations per second serviced by any file system listed in the heat map. In File System Write IOPS mode, the color of the heat map entries shows the relative nmber of write IO operations per second serviced by the file system. The color ranges from light green to dark green. Dark green indicates the highest nmber of write IO operations per second serviced by any file system listed in the heat map. VNX Overview dashboard 41

42 EMC Storage Analytics Dashboards VMAX Overview dashboard This collection of heat maps provides a single view of the performance and capacity for all VMAX resorces with configred adapter instances. Heat maps on this dashboard grop the contents by adapter instance. Figre 8 VMAX Overview 42 EMC Storage Analytics 2.1 Installation and User Gide The VMAX Overview dashboard displays the following heat maps. For each heat map and selected metric, the configred VMAX Adapter is shown: Thin Pool Usage This heat map displays the Percent Allocated metric. Percent Allocated displays the allocated capacity in each thin pool. The color of the heat map entries ranges from green to red and corresponds to the percent allocated as follows: Green indicates that 0% of the thin pool is allocated. Yellow indicates that 50% of the thin pool is allocated. Red indicates that 100% of the thin pool is allocated. Storage Grop This has for modes: Total Reads (IO/s), Total Writes (IO/s), Read Latency (ms), and Write Latency (ms). Total Reads and Writes represent the aggregate reads or writes for all LUNs in the storage grop. Write and Read Latency is the average write or read latency of all LUNs in the storage grop. For Total Reads and Writes, the color of the heat map entries shows the relative nmber of total reads or writes across all the storage grops. The color ranges from light ble to dark ble. Dark ble indicates the storage grop(s) with the highest nmber of total reads or writes while light ble indicates the lowest. Becase the range of vales for total reads or writes has no lower or pper limits, the nmerical difference between light and dark ble may be very small. For Write and Read Latency, the color of the heat map entries is based on a scale of latency from 0 to 40ms. This scale is a based on average cstomer reqirements and may not represent a cstomer's particlar reqirements for latency. In sch cases,

43 EMC Storage Analytics Dashboards EMC recommends adjsting the scale appropriately. The color of the heat map entries ranges from green to red as follows: Green indicates a latency of 0ms. Yellow indicates a latency of 20ms. Red indicates a latency of 40ms. LUN Performance This has for modes: Reads (IO/s), Writes (IO/s), Read Latency (ms), and Write Latency (ms). Reads and Writes represent the total reads or writes for a particlar LUN. Read and Write Latency is the average read or write latency of all LUNs in the storage grop. For Reads and Writes the color of the heat map entries shows the relative nmber of reads or writes across all the LUNs. The color ranges from light ble to dark ble. Dark ble indicates the LUN(s) with the highest nmber of reads or writes while light ble indicates the lowest. Becase the range of vales for reads or writes has no lower or pper limits, the nmerical difference between light and dark ble may be very small. For Read and Write Latency the color of the heat map entries ranges from green to red and is based on a scale of latency from 0 to 40ms. This scale is a based on average cstomer reqirements and may not represent a cstomer's particlar reqirements for latency. In sch cases EMC recommends adjsting the scale appropriately. The color of the heat map entries range from green to red as follows: Green indicates a latency of 0ms. Yellow indicates a latency of 20ms. Red indicates a latency of 40ms. Front End Director This has two modes: Total Bandwidth (MB/s) and Total Operations (IO/s). Total Bandwidth is the cmlative amont of data transferred over all ports of the front-end director. Total Operations is the total nmber of operations taking place over all ports of a front-end director. The color of the heat map entries is the same for both metrics. It shows the relative total bandwidth or relative total nmber of operations, depending on the selected metric. The color ranges from light ble to dark ble. Dark ble indicates the frontend director(s) with the highest nmber of total operations or the greatest total bandwidth, depending on the selected metric. Light ble indicates the lowest nmber of operations or the least total bandwidth. Becase the range of vales for operations or bandwidth has no lower or pper limits, the nmerical difference between light and dark ble may be very small. SRDF Director This has two modes: Total Bandwidth (MB/s) and Total Writes (IO/s). Total Bandwidth is the cmlative amont of data transferred over an SRDF director. Total Writes is the total nmber of writes over an SRDF director. The color of the heat map entries is the same for both metrics. It shows the relative total bandwidth or relative total nmber of writes, depending on the selected metric. The color ranges from light ble to dark ble. Dark ble indicates the SRDF director(s) with the highest nmber of total writes or the greatest total bandwidth, depending on the selected metric. Light ble indicates the lowest nmber of writes or the least total bandwidth. Becase the range of vales for bandwidth or writes VMAX Overview dashboard 43

44 EMC Storage Analytics Dashboards has no lower or pper limits, the nmerical difference between light and dark ble may be very small. SRDF Grops This has for modes: Devices in Session (cont), Average Cycle Time (ms), Writes (IO/s), and Writes (MB/s). Devices in Session represents the nmber of devices in an SRDF session in the SRDF grop. The Average Cycle Time is an SRDF/A metric that provides the average elapsed time between data transfer cycles. Writes (IO/s) represents the nmber of writes per second on the devices in the SRDF grop. Writes (MB/s) represents the nmber of megabytes per second sent from the SRDF grop. The color of the heat map entries is the same for all metrics. It shows the relative devices in session, average cycle time, total bandwidth, or the relative nmber of writes, depending on the selected metric. The color ranges from light ble to dark ble. Dark ble indicates the SRDF grop(s) with the highest nmber of one these metrics and light ble indicates the lowest nmber of one of these metrics. Becase the range of vales has no lower or pper limits, the nmerical difference between light and dark ble may be very small. Topology dashboards The topology dashboards provide an entry point for viewing resorces and relationships between storage and virtal infrastrctre objects for VNX and VMAX adapter instances. Click the VNX Topology or VMAX Topology tab, depending on the EMC adapter instance yo installed. Details for every object in every widget are available by selecting the object and clicking the Resorce Detail icon at the top of each widget as shown. 44 EMC Storage Analytics 2.1 Installation and User Gide

45 EMC Storage Analytics Dashboards Figre 9 Example of VNX Topology dashboard Topology dashboards 45

46 EMC Storage Analytics Dashboards Figre 10 Example of VMAX Topology dashboard 46 EMC Storage Analytics 2.1 Installation and User Gide

47 EMC Storage Analytics Dashboards The topology dashboards contain the following widgets: Resorce Tree This widget shows the end-to-end topology and health of resorces across vsphere and storage domains. Yo can configre the hierarchy that is shown by changing the widget settings; changing these settings does not alter the nderlying object relationships in the database. Select any resorce in this widget to view related resorces in the stack. Health Tree The Health Tree widget provides a navigable visalization of resorces that have parent or child relationships to the resorce yo select in the Resorce Tree widget. Single-click to select resorces, or doble-click to change the navigation focs. Metric Sparklines This widget shows sparklines for the metrics of the resorce yo select in the Resorce Tree widget. Metrics dashboards The metrics dashboards display resorces and metrics for VNX or VMAX storage systems and allows the ser to view graphs of resorce metrics. Figre 11 Example of VNX Metrics dashboard Metrics dashboards 47

48 EMC Storage Analytics Dashboards Figre 12 Example of VMAX Metrics dashboard The metrics dashboards contain the following widgets: Resorce Tree This widget shows the end-to-end topology and health of resorces across vsphere and storage domains. Yo can configre the hierarchy that is shown by changing the widget settings; changing these settings does not alter the nderlying object relationships in the database. Select any resorce in this widget to view related resorces in the stack. Metric Selector This widget lists all the metrics that are collected for the resorce yo select in the Resorce Tree widget. Doble-click a metric to create a graph of the metric in the Metric Graph widget. Metric Graph This widget graphs the metrics yo select in the Metric Selector widget. Display mltiple metrics simltaneosly in a single graph or in mltiple graphs. 48 EMC Storage Analytics 2.1 Installation and User Gide

49 EMC Storage Analytics Dashboards Top-N dashboards Top-N dashboards enable yo to view yor top performers at a glance. The nmber of objects in yor top performer list is configrable. Top-N dashboards are available for VNX and VMAX systems as shown in these examples. Figre 13 Example of VMAX Top-5 Devices Top-N dashboards 49

50 EMC Storage Analytics Dashboards Figre 14 Example of Top-5 VNX File Systems 50 EMC Storage Analytics 2.1 Installation and User Gide

51 EMC Storage Analytics Dashboards Figre 15 Example of Top-5 VNX LUNs By defalt, the dashboard shows the top 5 devices in the following categories across all of yor VNX or VMAX systems: Top-5 by Read (IO/s) Top-5 by Write (IO/s) Top-5 by Read (MB/s) Top-5 by Write (MB/s) Either: Top-5 by Read Latency (ms) for VMAX Top-5 Devices dashboard Top-5 by Consmed Capacity for Top-5 VNX LUNs or Top-5 VNX File Systems dashboards Top performers are selected based on the crrent vale of the associated metric that yo configred for each widget. Yo can change the time period. Yo can also configre each widget on the dashboard to show more than 5 top performers in yor environment. Top-N dashboards 51

52 EMC Storage Analytics Dashboards Dashboard XChange The Dashboard XChange is a ser commnity page for sers to exchange EMC Storage Analytics cstom dashboards. EMC Storage Analytics provides a set of defalt dashboards that provide yo with a variety of fnctional views into yor storage environment. EMC Storage Analytics also enables yo to create cstom dashboards that allow yo to visalize collected data according to yor own reqirements. The Dashboard XChange is an extension of that featre that enables yo to: Export cstom dashboards to the Dashboard XChange to benefit a wider EMC Storage Analytics commnity Import cstom dashboards from the Dashboard XChange to add vale to yor own environment The Dashboard XChange, hosted on the EMC Commnity Network, will also host dashboards designed by EMC to showcase widget fnctions that may satisfy a particlar se-case in yor environment. Yo can import these dashboards into yor existing environment to enhance the fnctionality offered by EMC Storage Analytics. Yo can also edit imported dashboards to meet the specific reqirements of yor own storage environment. The Dashboard XChange, shown in Figre 16 on page 52, provides the following resorces to assist yo in creating cstom dashboards: How-to video that shows how to create cstom dashboards Best practices gide that provides detailed gidelines for dashboard creation Slide show that demonstrates how to import dashboards from or export them to the Dashboard XChange Figre 16 Dashboard XChange The EMC Storage Analytics Dashboard XChange is available at commnity.emc.com/commnity/connect/esa. Note that there are XChange Zones for Unified (VNX) and VMAX. 52 EMC Storage Analytics 2.1 Installation and User Gide

53 CHAPTER 4 Resorce Kinds and Metrics This chapter contains the following topics. Resorce Kinds VNX Block metrics VNX File metrics VMAX metrics...69 Resorce Kinds and Metrics 53

54 Resorce Kinds and Metrics Resorce Kinds Metrics collected in the vcenter Operations Manager are groped by Resorce Kind. Resorce Kinds for VNX Block EMC Storage Analytics incldes the following Resorce Kinds for VNX Block: EMC Adapter Instance Storage Processor FAST Cache Pool LUN RAID Grop LUN RAID Grop Storage Pool Tier Disk Resorce Kinds for VNX File EMC Storage Analytics incldes the following Resorce Kinds for VNX File: EMC Adapter Instance Data Mover Virtal Data Mover dvol NFS Export File Pool File System Resorce Kinds for VMAX EMC Adapter Instance Storage Grop Device Thin Pool Front-End Director Front-End Port Remote Replica Grop SRDF Director 54 EMC Storage Analytics 2.1 Installation and User Gide

55 Resorce Kinds and Metrics VNX Block metrics EMC Storage Analytics incldes the following metrics for VNX Block with one table for each Resorce Kind. Table 2 VNX Block metrics for Array Metric Additional information - - Elapsed collect time (ms) Time elapsed dring the collection. New metrics in each collect call New resorces in each collect call Nmber of down resorces Nmber of metrics collected Nmber of resorces collected Nmber of new metrics per collection. Nmber of new resorces per collection. Nmber of down resorces for this adapter instance. Nmber of metrics collected by this adapter instance. Nmber of resorces collected by this adapter instance. Table 3 VNX Block metrics for Storage Processor Metric Additional information - - Writes (IO/s) The nmber of writes per second at the time when the SP is polled. that is passed throgh the SP per second. Smaller reqests sally reslt in a higher write throghpt than larger reqests. Writes (MB/s) Reads (IO/s) Reads (MB/s) Bsy (%) Write Cache Flshes (MB/s) Dirty Cache Pages (%) Dirty Cache Pages (MB) The average write reqest size in megabytes that passes throgh the SP per second. Larger reqests sally reslt in higher bandwidth than smaller reqests. The average nmber of host read reqests that is passed throgh the SP per second. Smaller reqests sally reslt in a higher read throghpt than larger reqests. The average amont of host read data in megabytes that is passed throgh the SP per second. Larger reqests sally reslt in a higher bandwidth than smaller reqests. The percentage of time dring which the SP is serving reqests. When the SP becomes the bottleneck, the tilization will be at or close to 100%. And increase in workload will have no frther impact on the SP throghpt, bt the I/O response time will start increasing more aggressively. The average amont of data in megabytes that was written from the write cache to the disks per second. The vale is a measre of back-end activity. The amont of dirty cache pages by percentage. This metric is for 1st generation VNX models (see Spported VNX series (1st generation) on page 16 for list of spported VNX models). The amont of dirty cache pages in megabytes. This metric is for 2nd generation VNX models (see Spported VNX series VNX Block metrics 55

56 Resorce Kinds and Metrics Table 3 VNX Block metrics for Storage Processor (contined) Metric Additional information - - (2nd generation) on page 16 for list of spported VNX models). Read Cache Size (MB) Read Cache State Write Cache Size (MB) Write Cache State Read Cache Hit Ratio (%) Write Cache Hit Ratio (%) The size of the read cache in megabytes. This metric is only for 1st generation VNX models (see Spported VNX series (1st generation) on page 16 for list of spported VNX models). The enabled or disabled state of the read cache. The size of the write cache in megabytes. This metric is only for 1st generation VNX models (see Spported VNX series (1st generation) on page 16 for list of spported VNX models). The enabled or disabled state of the write cache. The ratio of read reqests that the SP Cache satisfied withot reqiring any disk access verss the total nmber of read reqests. The ratio of write reqests that the SP Cache satisfied withot reqiring any disk access verss the total nmber of write reqests. Table 4 VNX Block metrics for FAST Cache Metric Additional information - - Crrent Operation Creating or Destroying. Crrent Operation Stats Mode RAID Type Size (GB) Dirty (%) Flshed (MB) Read Cache Hit Ratio (%) Read Cache Hits (Hits/s) If there is a crrent FAST Cache operation in progress sch as destroying or creating, this displays the percentage complete. Read/Write. The RAID Type of FAST Cache. The capacity of FAST Cache. Appears in the Performance > SPA and Performance > SPB metric grops. Percentage of write cache pages owned by the SP that contain data that has not yet been flshed ot to the FAST Cache. Appears in the Performance > SPA and Performance > SPB metric grops. The average amont of data in megabytes that was written from the write cache to the FAST Cache. The ratio of read reqests that the FAST Cache satisfied withot reqiring any disk access verss the total nmber of read reqests. The higher the ratio the better the read performance. Average nmber of read reqests per second that were satisfied by the FAST Cache withot reqiring any disk access. Read reqests that are not FAST Cache hits are read misses. 56 EMC Storage Analytics 2.1 Installation and User Gide

57 Resorce Kinds and Metrics Table 4 VNX Block metrics for FAST Cache (contined) Metric Additional information - - Read Cache Misses (Misses/s) Average nmber of read reqests per second that reqired one or mltiple disk accesses. Write Cache Hit Ratio (%) Write Cache Hits (Hits/s) Write Cache Misses (Misses/s) Crrent Operation Complete (%) The ratio of write reqests that the FAST Cache satisfied withot reqiring any disk access verss the total nmber of write reqests. The higher the ratio the better the write performance. Average nmber of write reqests per second that were satisfied by the FAST Cache withot reqiring any disk access. Write reqests that are not FAST Cache hits are write misses. Average nmber of write reqests per second that reqired one or mltiple disk accesses. If there is a crrent FAST Cache operation in progress sch as destroying or creating, this displays the percentage complete. Table 5 VNX Block metrics for Pool LUN Metric Additional information - - Performance Tier Distribtion The distribtion (%) of the Performance Tier. (%) Extreme Performance Tier Distribtion (%) Capacity Tier Distribtion (%) Tiering Policy Initial Tier Consmed Capacity (GB) User Capacity (GB) Bsy (%) Qee Depth Read Cache State Write Cache State Reads (IO/s) The distribtion (%) of the Extreme Performance Tier. The distribtion (%) of the Capacity Tier. The tiering policy of this Pool LUN. The initial tier that was sed for initial placement of the new LUN. The amont of space consmed in the pool by the LUN pls overhead. The amont of space consmed in the pool by the LUN. Appears in the Performance > SPA and Performance > SPB metric grops. The fraction of an observation period dring which a LUN has any otstanding reqests. When the LUN becomes the bottleneck, the tilization is at or near 100%. However, since the I/Os can be serviced by mltiple disks, an increase in workload may still reslt in a higher throghpt. The length of the LUN qee. The enabled or disabled state of the read cache. The enabled or disabled state of the write cache. The average nmber of host read reqests that is passed throgh the LUN per second. Smaller reqests sally reslt in a higher read throghpt than larger reqests. VNX Block metrics 57

58 Resorce Kinds and Metrics Table 5 VNX Block metrics for Pool LUN (contined) Metric Additional information - - Reads (MB/s) The average amont of host read data in Mbytes that is passed throgh the LUN per second. Larger reqests sally reslt in a higher bandwidth than smaller reqests. Writes (IO/s) Writes (MB/s) Explicit trespasses Implicit trespasses Total Latency (ms) The average nmber of host write reqests that is passed throgh the LUN per second. Smaller reqests sally reslt in a higher write throghpt than larger reqests. The average amont of host write data in Mbytes that is passed throgh the LUN per second. Larger reqests sally reslt in higher bandwidth than smaller reqests. Appears in the Performance > SPA and Performance > SPB metric grops. Nmber of trespasses since the last poll. Defalt polling cycle is 5 mintes. Occrs as a reslt of an external command from a ser or the failover software. When an SP receives this command, LUN ownership is transferred to that SP. Appears in the Performance > SPA and Performance > SPB metric grops. Nmber of trespasses since the last poll. Defalt polling cycle is 5 mintes. Occrs as a reslt of software controls within the storage system. An implicit trespass occrs when the amont of I/O transferred across the non-optimal path exceeds the optimal path I/O by a specified threshold. The average time, in milliseconds, that a reqest to this LUN is otstanding, inclding its waiting time. Table 6 VNX Block metrics for RAID Grop LUN Metric Additional information - - Total Latency (ms) The average time in milliseconds, that it takes for one reqest to pass throgh the LUN, inclding any waiting time. The higher the qee length for a LUN, the more reqests are waiting in its qee, ths increasing the average latency of a single reqest. For a given workload, qee length and response time are directly proportional. Read Cache State Write Cache State Reads (IO/s) Reads (MB/s) The enabled or disabled state of the read cache. The enabled or disabled state of the write cache. The average nmber of host read reqests that is passed throgh the LUN per second. Smaller reqests sally reslt in a higher read throghpt than larger reqests. The average amont of host read data in Mbytes that is passed throgh the LUN per second. Larger reqests sally reslt in a higher bandwidth than smaller reqests. 58 EMC Storage Analytics 2.1 Installation and User Gide

59 Resorce Kinds and Metrics Table 6 VNX Block metrics for RAID Grop LUN (contined) Metric Additional information - - Writes (IO/s) The average nmber of host write reqests that is passed throgh the LUN per second. Writes (MB/s) Qee Depth User Capacity Bsy (%) The average amont of host write data in Mbytes that is passed throgh the LUN per second. The length of the LUN qee. The amont of space available in the RAID Grop LUN. Appears in the Performance > SPA and Performance > SPB metric grops. The fraction of an observation period dring which a LUN has any otstanding reqests. When the LUN becomes the bottleneck, the tilization is at or near 100%. However, since the I/Os can be serviced by mltiple disks, an increase in workload may still reslt in a higher throghpt. Table 7 VNX Block metrics for RAID Grop Metric Additional information - - Fll (%) Percentage of total capacity that is consmed. User Capacity (GB) Raw Capacity (GB) Available Capacity (GB) Free Continos Grop of Unbond Segments (GB) Disk Cont LUN Cont Max Disks Max LUNs Defragmented (%) The amont of space available in the RAID Grop. The total amont of space available in the RAID Grop prior to RAID protection. The remaining free capacity of this RAID Grop. Size in GB of the largest continos span of free space in the RAID Grop. LUNs mst fit into a contigos span of free space. Nmber of disks in this RAID Grop. Nmber of LUNs in this RAID Grop. Maximm nmber of disks allowed for this RAID Grop. Maximm nmber of LUNs allowed for this RAID Grop. When a defragment operation in progress, this displays the percentage complete. VNX Block metrics 59

60 Resorce Kinds and Metrics Table 8 VNX Block metrics for Storage Pool Metric Additional information - - Available Capacity (GB) Capacity available for se in this Storage Pool. Consmed Capacity (GB) Disk Type Crrent Operation Crrent Operation Complete (%) Crrent Operation State Crrent Operation Stats FAST Cache Fll (%) Sbscribed (%) Threshold (%) Oversbscribed (GB) Disk Cont LUN Cont Ato-Tiering Relocation Stats Relocation Type Relocation Rate Relocation Start Time Relocation Stop Time Schedle Dration Remaining Ato-Tiering State Data Movement Completed (GB) Data to Move Up (GB) Data to Move Down (GB) Data to Move Within (GB) Estimated Time to Complete Dedplication State Capacity sed in this Storage Pool. Type of disks in this Storage Pool. Displays the crrent operation in the pool. If there is a thin pool operation in progress sch as a rebalance, this displays the percentage complete. Displays the crrent operation state. Displays additional descriptive information for the crrent state of the thin pool. The enabled or disabled state of the FAST Cache for this Storage Pool. Percentage of total capacity that is consmed. Percentage of total capacity that is sbscribed. Threshold as percentage of total capacity. How mch the Storage Pool is oversbscribed. Nmber of disks consmed by this Storage Pool. Nmber of LUNs hosted by this Storage Pool. Shows if ato-tiering is schedled. Relocation is active or inactive. Schedled or manal relocation. Rate at which relocation occrs. Start time for the relocation. Stop time for the relocation. If sing schedled relocation, this displays the remaining time for the relocation. The enabled or disabled state of ato-tiering. Amont of data that has been moved p or down. Amont of data that is going to be moved p. Amont of data that is going to be moved down. Amont of data to move within tiers. Estimated time to complete the data relocation. The dedplication state can take any of these vales: Idle (No dedplicated LUNs) Idle (No dedplicated LUNS) - Falted Pending 60 EMC Storage Analytics 2.1 Installation and User Gide

61 Resorce Kinds and Metrics Table 8 VNX Block metrics for Storage Pool (contined) Metric Additional information - - Pending - Falted Rnning (% complete, GB remaining) Rnning - Falted (% complete, GB remaining) Pased Pased - Falted Tiering Policy With FAST VP enabled, tiering policy can take any of these vales: Start High then Ato-Tier (recommended) Ato Tier Highest Available Tier Lowest Available Tier No Data Movement With FAST VP disabled, tiering policy can be: Optimize for Pool Performance (defalt) Highest Available Tier Lowest Available Tier Initial Tier Dedplication Rate Dedplicated LUNs Shared Capacity (GBs) Dedplication and Snapshot Savings (GBs) Initial tier can be any of the vales available for Tiering Policy (above). Rate of dedplication. Shared capacity of dedplicated LUNs. Capacity savings throgh dedplication and Snapshots. Table 9 VNX Block metrics for Tier Metric Additional information - - RAID Type Type of RAID applied to the tier. Disk Cont Higher Tier (GB) Lower Tier (GB) Available Capacity (GB) Consmed Capacity (GB) User Capacity (GB) Sbscribed (%) Nmber of disks in the tier. Amont of data targeted for higher tiers. Amont of data targeted for lower tiers. Capacity still available for se. Used capacity. Free capacity for sers. Percentage of tier that is sbscribed. VNX Block metrics 61

62 Resorce Kinds and Metrics Table 10 VNX Block metrics for Disk Metric Additional information - - Capacity (GB) Total capacity of the disk. LUN Cont State Hard Read Errors Hard Write Errors Reads (IO/s) Reads (MB/s) Writes (IO/s) Writes (MB/s) Bsy (%) Total Latency (ms) Qee Depth Total nmber of LUNs that the disk is serving. Crrent state of the disk. Nmber of hard read errors. Nmber of hard write errors. The average nmber of read reqests from the disk per second. The average amont of data read from the disk in Mbytes per second. The average nmber of write reqests to the disk per second. The average amont of data written to the disk in Mbytes per second. The percentage of time dring which the disk is servicing any reqests. The average time, in milliseconds, that it takes for one reqest to pass throgh the disk, inclding any waiting time. The average nmber of reqests within a polling interval that are waiting to be serviced by the disk, inclding the one crrently in service. 62 EMC Storage Analytics 2.1 Installation and User Gide

63 Resorce Kinds and Metrics VNX File metrics EMC Storage Analytics incldes the following metrics for VNX File with one table for each Resorce Kind. Table 11 VNX File metrics for Array Metric Additional information - - Elapsed collect time (ms) Amont of elapsed time for the collection. New metrics in each collect call New resorces in each collect call Nmber of down resorces Nmber of metrics collected Nmber of resorces collected Nmber of new metrics per collection. Nmber of new resorces per collection. Nmber of down resorces for this adapter instance. Nmber of metrics collected by this adapter instance. Nmber of resorces collected by this adapter instance. Table 12 VNX File metrics for Data Mover Metric Description Additional information Type Data Mover type. The vale can be Active (for the primary Data Mover), Standby, or VDM. - Bsy (%) DNLC Hit Ratio (%) Open File Cache Hit Ratio (%) Bffer Cache Hit Ratio (%) Reads (MB/s) Writes (MB/s) Total Bandwidth (MB/s) CPU tilization percentage dring this interval. Directory Name Lookp Cache (DNLC) hit ratio percentage sed for pathname resoltion logic. Open File Cache Hit ratio percentage. Bffer Cache Hit ratio percentage. Storage in Mbytes received from all server-storage interfaces. Storage in Mbytes sent to all server-storage interfaces. The total bandwidth for the Data Mover NFS Reads (IO/s) NFS read operations per second. - NFS Writes (IO/s) NFS write operations per second. - NFS Reads (MB/s) NFS Writes (MB/s) NFS read data response in Mbytes per second. NFS write data response in Mbytes per second. - - VNX File metrics 63

64 Resorce Kinds and Metrics Table 12 VNX File metrics for Data Mover (contined) Metric Description Additional information NFS Average Read Size (Bytes) Average size of data read. - NFS Average Write Size (Bytes) Average size of data written. - Network In Bandwidth (MB/s) Network Ot Bandwidth (MB/s) Network in bandwidth. Megabytes received per second. Network ot bandwidth. Megabytes sent per second. - - Read Calls/s Read calls per second These metrics appear in the Network > Read Errors/s Read errors per second NFSv2, NFSv3, and NFSv4 metric grops. Read Response Time (ms) Total read response time Write Calls/s Write Errors/s Write Response Time (ms) GetAttr Calls/s GetAttr Errors/s GetAttr Response Time (ms) SetAttr Calls/s SetAttr Errors/s SetAttr Response Time (ms) Lookp Calls/s Lookp Errors/s Lookp Response Time (ms) Access Calls/s Access Errors/s Access Response Time (ms) Write calls per second Write errors per second Total write response time Get file attribtes (GetAttr) per second GetAttr errors per second Total response time for GetAttr Set file attribtes (SetAttr) per second SetAttr errors per second Total response time for SetAttr Lookp calls per second Lookp errors per second Total lookp response time Access calls per second Access errors per second Total access response time These metrics appear in the Network > NFSv3 metric grop. Open Calls/s Open calls per second These metrics appear in the Network > Open Errors/s Open errors per second NFSv4 metric grop. Open Response Time (ms) Total open response time 64 EMC Storage Analytics 2.1 Installation and User Gide

65 Resorce Kinds and Metrics Table 12 VNX File metrics for Data Mover (contined) Metric Description Additional information Close Calls/s Close calls per second Close Errors/s Close Response Time (ms) Compond Calls/s Compond Errors/s Compond Response Time (ms) Close errors per second Total close response time Compond calls per second Compond errors per second Total compond response time CIFS Reads (IO/s) IOs per second for CIFS reads These metrics appear in the Network CIFS Reads (MB/s) Megabytes per second for CIFS metric grop. reads CIFS Average Read Size (KB) CIFS Writes (IO/s) CIFS Writes (MB/s) CIFS Average Write Size (KB) Average CIFS read size IOs per second for CIFS writes Megabytes per second for CIFS writes Average CIFS write size ReadX Calls/s ReadX calls per second These metrics appear in the Network > ReadX Max Response ReadX maximm response time SMB1 metric grop. Time (ms) ReadX Average Response Time (ms) WriteX Calls/s WriteX Max Response Time (ms) WriteX Average Response Time (ms) NTCreateX Calls/s NTCreateX Max Response Time (ms) NTCreateX Average Response Time (ms) Close Calls/s Close Max Response Time (ms) Close Average Response Time (ms) Average response time for ReadX WriteX calls per second WriteX maximm response time WriteX average response time NTCreateX calls per second NTCreateX maximm response time NTCreateX average response time Close calls per second Close maximm response time Close average response time VNX File metrics 65

66 Resorce Kinds and Metrics Table 12 VNX File metrics for Data Mover (contined) Metric Description Additional information Trans2Prim Calls/s Trans2Prim calls per second Trans2Prim Max Response Time (ms) Trans2Prim Average Response Time (ms) Trans2Prim maximm response time Trans2Prim average response time Create Calls/s Create calls per second These metrics appear in the Network > Create Max Response Create maximm response time SMB2 metric grop. Time (ms) Create Average Response Time (ms) Close Calls/s Close Max Response Time (ms) Close Average Response Time (ms) Flsh Calls/s Flsh Max Response Time (ms) Flsh Average Response Time (ms) Read Calls/s Read Max Response Time (ms) Read Average Response Time (ms) Write Calls/s Write Max Response Time (ms) Write Average Response Time (ms) Ioctl Calls/s Ioctl Max Response Time (ms) Ioctl Average Response Time (ms) Qeryinfo Calls/s Create average response time Close calls per second Close maximm response time Close average response time Flsh calls per second Flsh maximm response time Flsh average response time Read calls per second Read maximm response time Read average response time Write calls per second Write maximm response time Write average response time IO Control (IOCTL) calls per second IO Control maximm response time IO Control average response time Qery information calls per second 66 EMC Storage Analytics 2.1 Installation and User Gide

67 Resorce Kinds and Metrics Table 12 VNX File metrics for Data Mover (contined) Metric Description Additional information Qeryinfo Max Response Time (ms) Qery information maximm response time Qeryinfo Average Response Time (ms) Qery information average response time Table 13 VNX File metrics for dvol Metric Additional information - - Capacity (GB) Total capacity of the disk volme. Average Read Size (Bytes) Average Write Size (Bytes) Average Completion Time (µsec/call) Average Service Time (µsec/call) IO Retries (IO/s) Qee Depth Reads (IO/s) Reads (MB/s) Total Operations (IO/s) Total Bandwidth (MB/s) Utilization (%) Writes (IO/s) Writes (MB/s) Average size of data read. Average size of data written. Average time for completion of an I/O. Average service time for sccessfl completion of I/O withot retries and qeing delays. Nmber of I/O retries per second. Length of disk qee Nmber of read operations on the disk per second. Mbytes read from the disk per second. Nmber of I/O operations on the disk volme per second. Total bandwidth of the disk volme. Percentage time that disk has been tilized. Nmber of write operations on the disk per second. Mbytes written to the disk per second. Table 14 VNX File metrics for File Pool Metric Additional information - - Capacity (GB) Total capacity of the file pool. Consmed Capacity (GB) Available Capacity (GB) Consmed capacity of the file pool. Capacity still available for se. Table 15 VNX File metrics for File System Metric Additional information - - Capacity (GB) Total space available for storage of ser data (does not inclde metadata). Consmed Capacity (GB) Consmed capacity of the File System. VNX File metrics 67

68 Resorce Kinds and Metrics Table 15 VNX File metrics for File System (contined) Metric Additional information - - Available Capacity (GB) Capacity still available for se. Thin Provisioning Max Capacity (GB) Read Reqests (Reqests/s) Write Reqests (Reqests/s) Reads (MB/s) Writes (MB/s) Average Read Size (Bytes) Average Write Size (Bytes) Reads (IO/s) Writes (IO/s) Read IO Ratio (%) Write IO Ratio (%) Tre indicates that the file system is enabled for virtal provisioning, an option that can only be sed with atomatic file system extension. Combining atomatic file system extension with virtal provisioning allows growth of the file system gradally and as needed. When virtal provisioning is enabled, NFS and CIFS clients receive reports for either the virtal maximm file system size or real file system size, which ever is larger. If atomatic extension is enabled, the file system will atomatically extend to this maximm size when the high water mark is reached. The defalt vale for the high water mark is 90 percent. Read operations per second in the interval. Write operations per second in the interval. Read data response in Mbytes per second. Write data response in Mbytes per second. Average size of data read. Average size of data written. The average read operations per second. The average write operations per second. Percentage of total IOs that are read IOs. Percentage of total IOs that are write IOs. 68 EMC Storage Analytics 2.1 Installation and User Gide

69 Resorce Kinds and Metrics VMAX metrics EMC Storage Analytics incldes the following metrics for VMAX with one table for each Resorce Kind. Table 16 VMAX metrics for Storage Grop Metric Definition - - Reads (IO/s) The nmber of read operations performed each second by the storage grop. Writes (IO/s) Reads (MB/s) Writes (MB/s) Total Operations (IO/s) Read Latency (ms) Write Latency (ms) Total Latency (ms) Total Bandwidth (MB/s) The nmber of write operations performed each second by the storage grop. The cmlative nmber of MBs read per second by the storage grop. The cmlative nmber of MBs written per second by the storage grop. The total reads and writes performed each second by the storage grop. The average time it took the Symmetrix array to serve one Read IO for this storage grop. The average time it took the Symmetrix array to serve one Write IO for this storage grop. The average time it took the Symmetrix array to serve one IO for this storage grop. The total nmber of MBs sent and received per second by the storage grop. Table 17 VMAX metrics for Device Metric Definition - - Reads (IO/s) The nmber of Read operations performed each second on the device. Writes (IO/s) Reads (MB/s) Writes (MB/s) Total Bandwidth (MB/s) Total Operations (IO/s) Read Latency (ms) Write Latency (ms) The total nmber of Write IO operations performed each second by the Symmetrix volme (LUN). The cmlative nmber of MBs read per second from the device. The cmlative nmber of MBs written per second to the device. The total nmber of read and write MBs performed each second on the device. The total reads and writes performed each second on the device. The average time it took the Symmetrix array to serve one Read IO for this device. The average time it took the Symmetrix array to serve one Write IO for this device. VMAX metrics 69

70 Resorce Kinds and Metrics Table 18 VMAX metrics for Thin Pool Metric Definition - - Total Capacity (GB) The total thin pool capacity. Allocated Capacity (GB) Used Capacity (GB) Fll (%) The allocated thin pool capacity. The sed thin pool capacity. The percent that the thin pool has been allocated. Table 19 VMAX metrics for Front-End Director Metrics Definition - - Total Operations (IO/s) The total reads and writes the front-end director processes per second. Reads (IO/s) Writes (IO/s) Total Hits (IO/s) Total Bandwidth (MB/s) The total read operations the front-end director processes per second. The total write operations the front-end director processes per second. The total nmber of reqests that were immediately serviced from cache. The total nmber of MBs sent and received per second by the director. Table 20 VMAX metrics Front-End Port Metrics Definition - - Total Operations (IO/s) The total reads and writes the front-end port processes per second. Total Bandwidth (MB/s) The nmber of MBs transferred per second. Table 21 VMAX metrics SRDF Director Metrics Definition - - Total Operations (IO/s) The total reads and writes the SRDF director processes per second. Reads (IO/)s Writes (IO/s) Total Bandwidth (MB/s) The total read operations the SRDF director processes per second. The total write operations the SRDF director processes per second. The total nmber of MBs sent and received per second by the RDF director. Table 22 VMAX metrics Remote Replica Grop Metrics Definition - - Minimm Cycle Time (ms) The setting for the minimm nmber of seconds for a cycle. Average Cycle Time (ms) Delta Set Extension Threshold The average time it takes for each cycle to complete. The percent of write pendings before DSE activates. 70 EMC Storage Analytics 2.1 Installation and User Gide

71 Resorce Kinds and Metrics Table 22 VMAX metrics Remote Replica Grop (contined) Metrics Definition - - Devices in Session The nmber of devices in the grop. (cont) Writes (IO/s) Writes (MB/s) HA Repeat Writes (conts/s) The nmber of write IOs coming in for the volmes in this session. The total nmber of MBs sent per second by the grop. Writes to a slot already in the active cycle. Total host adapter repeat writes, measred in write commands to SRDF/A volmes only. This conter helps estimate the cache locality of reference, i.e., how mch cache is saved by the re-writes. This does not give any indication to the bandwidth locality of reference. VMAX metrics 71

72 Resorce Kinds and Metrics 72 EMC Storage Analytics 2.1 Installation and User Gide

73 CHAPTER 5 Trobleshooting This chapter contains the following topics. Lanching Unisphere Installation logs Error handling and event logging Editing the Collection Interval for a resorce...78 Freqently Asked Qestions - General Freqently Asked Qestions for VNX...83 Freqently Asked Qestions for VMAX Trobleshooting 73

74 Trobleshooting Lanching Unisphere EMC Storage Analytics provides metrics that enable yo to assess the health of monitored resorces. If the metrics for VNX Block or VNX File resorces indicate that yo need to trobleshoot those resorces, EMC Storage Analytics provides a way to lanch Unisphere on the VNX array in context. Select the resorce and click the lanch icon as shown in the example in Figre 17 on page 74. The lanch icon is available on most widgets. Note This featre reqires a fresh installation of the EMC Adapter (not an pgrade) and vcenter Operations Manager or later. Yo mst select the VNX Block or VNX File object to lanch Unisphere. There is no Unisphere lanch capability for VMAX objects. Figre 17 Example of lanching to Unisphere Installation logs This topic lists the log files to which errors in the EMC Storage Analytics installation are written. Errors in the EMC Storage Analytics installation are written to log files in the following directory on the UI VM of the vcenter Operations Manager: /var/log/emc Log files in this directory follow the naming convention: install :54:19.log. Use a text editor to view the installation log files. 74 EMC Storage Analytics 2.1 Installation and User Gide

75 Trobleshooting Error handling and event logging Errors in the EMC Storage Analytics operation are written to log files on the Analytics VM of the vcenter Operations Manager. Error logs are available in the /data/vcops/log directory. This directory contains the vcenter Operations Manager logs. Adapter logs (inclding adapters other than the EMC Adapter) are in /data/ vcops/log/adapters. Logs for each EMC Adapter instance are in folders nder /data/vcops/log/ adapters/emcadapter. For example, if yo have five EMC Adapter instances, a directory appears for each of them. Log files in this directory follow this naming convention: VNX_File-131.log.9. The log filename begins with the name of the EMC Adapter instance (for example, VNX_Block or EmcAdapter). Filenames beginning with EmcAdapter are common to all connectors. The nmber that follows the EMC Adapter instance name is the adapter instance ID, which corresponds to a VMware internal ID. The last nmber in the filename indicates the roll-over increment; the lowest nmber represents the most recent log. Each roll-over is 10 MB, and ten roll-overs are possible. View logs relating to EMC Storage Analytics operation in the vcenter Operations Manager GUI. Create and download a spport bndle sed for trobleshooting. Viewing error logs EMC Storage Analytics enables yo to view error log files for each adapter instance. Procedre 1. Start the vcenter Operations Manager cstom ser interface and log in as administrator. For example in a web browser, type: 2. Select Admin > Spport. Select the Logs tab. 3. Expand the vcenter Operations Collector. Select vcenter Operations Collector > adapters > EmcAdapter. Log files appear nder the EmcAdapter folder and are separated for each adapter instance. Doble-click a log entry in the log tree. Entries appear in the Log Content window. Error handling and event logging 75

76 Trobleshooting Figre 18 Log Content Creating and downloading a spport bndle Procedre 1. On the Logs tab, for Spport bndle, click the Create Spport Bndle icon. 76 EMC Storage Analytics 2.1 Installation and User Gide

77 Trobleshooting The bndle encapslates all necessary logs. 2. Select the bndle name and click the Download Bndle icon. Creating and downloading a spport bndle 77

78 Trobleshooting Editing the Collection Interval for a resorce From the vcenter Operations Manager UI, yo can edit the Collection Interval for a resorce. The interval time is 5 mintes by defalt. Changing this time will affect the freqency of collection times for the VNX Block and VNX File metrics, bt the EMC Adapter will only recognize the change if the resorce is the EMC Adapter instance. This is normal vcenter Operations Manager behavior. Instrctions on configring Resorce Management settings are provided in the vcenter Operations Manager online help. 78 EMC Storage Analytics 2.1 Installation and User Gide

79 Trobleshooting Freqently Asked Qestions - General How does a trial license work? A 90-day trial license is provided for each platform that EMC Storage Analytics spports (for example, NS960). The 90-day trial license provides the same featres as a permanent license, bt after 90 days, the adapter stops collecting data. Yo can add a permanent license at any time dring or after the trial period. How do health scores work? Health scores measre how normal a resorce is acting and grades it on a scale of A health score of 100 indicates normal behavior while a lower health score indicates that the resorce is acting abnormally. The resorce may not be in an nhealthy state bt there is an abnormality. Health scores are calclated by a proprietary algorithm which accont for several factors inclding thresholds and historical statistics. vcenter Operations Manager may take p to 30 days to gather enogh information to determine what is considered normal in yor environment. Until then, yo may not see any changes in yor health scores. How many arrays can I add? It is difficlt to arrive the exact nmber of arrays yo can add becase the scale limits depend on the nmber of monitored objects and the metrics applied to them. When yo deploy vcenter Operations Manager 5.8 and EMC Storage Analytics 2.1 in fll profile mode, it can spport p to 15,000 objects and p to 5 million metrics. Examples of objects are File Systems, NFS Exports, LUNs, Storage Pools, and Disks. For more details, refer to vcenter Operations Manager 5.8 Release Notes at spport/vcops/doc/vcops-571-vapp-release-notes.html. I deleted a resorce. Why does it still appear in the vcenter Operations Manager? vcenter Operations Manager will not delete any resorces atomatically becase it retains historical statistics and topology information that may be important to the ser. To remove the resorce, delete it on the Environment Overview page. What does the ble qestion mark in the health score indicate? The ble qestion mark indicates that vcenter Operations Manager was nable to poll that resorce. It will retry dring the next polling interval. What does it mean when a resorce has a health score of 0? This indicates that the resorce is either down or not available. Why are my EMC Adapter instances marked down after pgrading to the latest version of the EMC Adapter? EMC Adapter instances reqire a license to operate. Edit yor EMC Adapter instances to add license keys obtained from EMC. Select Environment Overview > Configration > Adapter Instances. I have added license keys for EMC Adapter instances. Why are my resorces still marked down? License keys are specific to the model for which the license was prchased. Verify that yo are sing the correct license key for the adapter instance. Yo can also click the Test btton to test the configration after editing adapter instances. I have mltiple EMC Adapter instances for my storage systems, and I have added license keys for each of them. Why are they still marked down? After adding a license, click the Test btton to test the configration and validate the license key. If yo saved the configration withot performing a test and the license is invalid, the adapter instance will be marked down. To verify that a valid license exists, select Environment Overview. The list that appears shows the license stats. Freqently Asked Qestions - General 79

80 Trobleshooting Figre 19 Stats shows: Resorce down What is the difference between vcenter Operations Manager 5.8 from VMware and vcenter Operations Manager inclded with EMC Storage Analytics? vcenter Operations Manager inclded with EMC Storage Analytics is based on vcenter Operations Manager 5.8 (Advanced edition) from VMware. However, EMC Storage Analytics is restricted to storage view. For vcenter resorces sch as virtal machines, yo can only see storage or disk related metrics. How is the detailed view of vcenter resorces affected in EMC Storage Analytics? Any changes in the disk system affects the health of vcenter resorces sch as virtal machines, bt EMC Storage Analytics does not show changes in other sbsystems. Metrics for other sbsystems will either show No Data or?. 80 EMC Storage Analytics 2.1 Installation and User Gide

81 Trobleshooting Figre 20 Workload shows: No Data Freqently Asked Qestions - General 81

82 Trobleshooting Figre 21 VM Resorces shows:? Can I see relationships between my vcenter and EMC storage resorces? Yes. Relationships between resorces are not affected and yo can see a top to bottom view of the virtal and storage infrastrctres if the two are connected. What if I want to see metrics for other sbsystems instead of No Data or?? To view metrics for other sbsystems, pgrade from vcenter Operations Manager for EMC Storage to the vcenter Operations Manager 5.8 Advanced edition. No pgrade is reqired for the EMC Adapter. 82 EMC Storage Analytics 2.1 Installation and User Gide

83 Trobleshooting Freqently Asked Qestions for VNX Which VNX arrays does EMC Storage Analytics spport? It spports first- and second-generation VNX platforms. Spport for CX4 covers CX4-120, CX4-240, CX4-480, and CX4-960 platforms. Spported NS series incldes NS120, NS240, and NS960 platforms. A complete list of the spported models in each of these series is available in Installation reqirements on page 16. Will EMC Storage Analytics contine to collect VNX statistics if the primary SP or CS goes down? Storage Analytics will contine to collect statistics throgh the secondary Storage Processor if the primary Storage Processor goes down. EMC Storage Analytics will atomatically collect metrics from the secondary Control Station in the event of a Control Station failover. Note that the credentials on the secondary Control Station mst match the credentials on the primary Control Station Does the Unisphere Analyzer for VNX need to be rnning to collect metrics? No. VNX Block metrics are gathered throgh naviseccli commands and VNX File metrics are gathered throgh CLI commands. However, statistics logging mst be enabled on each storage processor (SP) on VNX Block, and statistics logging will have a performance impact on the array. No additional services are reqired for VNX File. How does the FAST Cache heatmap work? The FAST Cache heatmaps are based on the FAST Cache read and write hit ratios. This heatmap will trn red if these ratios are low becase that indicates that FAST Cache is not being tilized efficiently. These heatmaps will trn green when FAST Cache is servicing a high percentage of I/O. I prchased a license for the model of the VNX array that I plan to monitor. When I configre the adapter instance for VNX File, why does an "invalid license" error message appear? Control Station may not be reporting the correct model or the array. Log into Control Station and check the array model with the command: /nas/sbin/model. Verify that the array model retrned matches the model on the Right to Use certificate. After a Control Station failover, why is the VNX File adapter instance marked down and why does metric collection stop? The failover may have been sccessfl, bt the new Control Station may not be reporting the correct model of the array. This reslts in a failre to validate the license and all data collection stops. Log into Control Station and check the array model with the command: /nas/sbin/model. If the model retrned does not match the actal model of the array, Prims case emc in the EMC Knowledgebase provides possible soltions. The disk tilization metric is not visible for my VNX Block array. Why not? The disk tilization metric is not spported on VNX arrays rnning a VNX Block OE earlier than Release 32. Upgrade to VNX Block OE Release 32 or later to see this metric in vcenter Operations Manager. I am nable to sccessflly configre an EMC Adapter instance for VNX File when sing a ser with read-only privileges. Why does this happen? A ser with administrative privileges is reqired while configring an EMC Adapter instance for VNX File arrays rnning an OE earlier than Upgrade to VNX File OE or later to be able to configre an adapter instance sing a ser with read-only privileges. Freqently Asked Qestions for VNX 83

84 Trobleshooting The ser LUNs on my VNX Block valt drives are not reporting performance metrics. Why not? Performance metrics are not spported for ser LUNs on valt drives. Place ser LUNs on drives other than valt drives. 84 EMC Storage Analytics 2.1 Installation and User Gide

85 Trobleshooting Freqently Asked Qestions for VMAX I deleted a VMAX adapter, bt all the dependent objects still remain in vcenter Operations Manager. Why? Deleting a VMAX adapter instance does not remove its dependent objects. The objects enter an nknown state (ble) bt retain their historical metrics. If desired, delete the objects from the Environment Overview page. If I test a VMAX connection and it fails, how do I know which field is wrong? Unfortnately, the only field that prodces a niqe message when it is wrong is the license nmber field. If any other field is wrong, the ser is informed the connection was not sccessfl. To resolve the isse, verify all the other fields are correct. Remove any white spaces after the end of the vales. Why is my VMAX array showing as red rather than ble (nknown) jst after adding it? If the array is red, the SMI-S may be down or there may be an incorrect parameter in the adapter configration. Retrn to the adapter instance configration and test the configred instance. If it fails, verify all fields are correct. Determine if the SMI-S server is rnning and can recognize the VMAX array that yo are configring. All my configred VMAX adapter instances changed to red. Why? This sally happens when the SMI-S Provider is navailable or there is a clock synchronization problem. This might be network-related or application-related. The SMI-S Provider may have reached the connection limit if the network IP is available. Look for errors in the cimom.log file on the SMI-S box similar to: 26-Mar :47: E- WebServer: The webserver hits its connection limit, closing connection. Follow the steps in the SMI-S Provider Release Notes to increase the cont. A message next to the array will indicate a clock synchronization problem if a problem exists. Can I modify or delete the VMAX Overview (or another) dashboard? Yes, the environment can be cstomized to sit the needs of the ser. Rename the dashboard so that it is not overwritten dring an pgrade. I received the following error when I attempted to modify the VNX Overview dashboard althogh I have only VMAX arrays: Is this a problem? No, this is a generic error that VMware prodces when yo attempt to modify a component yo do not have. Why do some of the boxes appear white in the VMAX Overview dashboard? While the metrics are being gathered for an adapter instance, some of the heat maps in the dashboard may be white. This is normal. Another reason the boxes may appear white is that the adapter itself or an individal resorce (e.g. a VMAX device) has been deleted, bt the resorces remain ntil they are removed from the Environment Overview page. Freqently Asked Qestions for VMAX 85