Discovery for Storage: Frequently Asked Questions (FAQ)

Version 12

    This document is a summary of frequently asked questions that relate to the BMC Discovery for Storage add-on. Feel free to ask for additions and I will be happy to update.

     

    Table of contents:

     

    Coverage

    What systems does BMC Discovery for Storage discover?

    The following systems are discoverable:

    • Dell Compellent
    • Dell EquaLogic
    • EMC Celerra
    • EMC Clariion CX
    • EMC Symmetrix, DMX, VNX, V-MAX
    • EMC DataDomain
    • EMC Isilon
    • EMC VPLEX
    • EMC XtremIO
    • Fujitsu Eternus
    • Hitachi AMS, HUS, USP-V, VSP and VSP-G
    • Hitachi HNAS
    • HP StorageWorks P2000 (DotHill)/P6000 (also known as EVA)/P9000
    • HP 3PAR
    • HP P2000, P6000, P9000, P9500
    • Huawei OceanStor
    • IBM XiV
    • IBM DS 6000, 8000 Series
    • IBM SVC, Storwize v7000/v3700
    • IBM FlashSystem
    • NetApp (including Cluster Mode)
    • Oracle ZFS NAS
    • Pure Storage FlashArray
    • Brocade SAN switches
    • Cisco SAN switches

     

    More will be added over time via TKU.

    What elements of the storage infrastructure are discovered (arrays, NAS, SAN switches, tape libraries etc)?

    BMC Discovery discovers the SAN and NAS devices, their hardware (disks, processors, ports etc) and logical partitioning (pools, volumes, file systems), as well as storage management software. It also discovers SAN switches.

    BMC Discovery does not currently discover tape libraries, nor Content-Addressable Storage.

    A complete taxonomy for storage is described in the documentation

    Architecture/implementation

    Does storage discovery need an agent?

    No, storage discovery is agent-less and can be triggered either upon discovery of storage management software or discovery of storage devices.

    What are the pre-requisites to discover storage systems?

    The pre-requisite to discover NetApp NAS devices is a valid read SNMP community string. This is also true for SAN switches.

    For any other device, the pre-requisite is credentials to access an SMI-S provider (via WBEM). The SMI-S provider can either be running on the SAN device or on a supported storage management software server. Detailed pre-requisites for each vendor/model are available in Configipedia.

    To discover and link storage consumers (hosts), credentials are required for accessing any host which is a consumer of the storage, and that the HBA ID is discovered.

    More details on pre-requisites can be found in the documentation.

    What permissions are needed to discover storage systems?

    Only read-only access via SNMP or WBEM is necessary.

    What protocols does Discovery leverage to discover storage?

    SNMP for NetApp and WBEM for SAN devices.

    How do storage elements map into CIs in the CMDB?

    The following table descibes how data is synchronized in CMDB:

    Node KindCMDB Class
    StorageDeviceBMC_Cluster (NetApp clusters only)
    StorageSystemBMC_ComputerSystem
    StoragePoolBMC_ResourcePool
    StorageProcessorBMC_HardwarePackage
    FibreChannelPortBMC_NetworkPort
    StorageVolumeNot mapped

    The reason why volumes are not mapped by default is that there can be an awful lot of them, and that they do not play a direct role in CMDB integrations with monitoring, automation etc. However, synchronization is customizable, so volumes can be added.

    How do I enable storage discovery?

    Discovery for Storage is a licensed add-on to BMC Discovery. It is provided as a TKU file that includes all patterns needed to discover the storage infrastructure.Users who are entitled to this SKU will be able to download the storage TKU from the EPD and apply it to their  appliances.

    What information do we discover if the storage TKU is not loaded?

    The taxonomy includes storage devices (i.e. storage array hardware that Discovery has talked to) and storage systems (i.e. storage array system that was either discovered by communicating to a storage device or to storage management software).
    Basic discovery can identify storage devices, but will not gather their configuration, nor relationships with hosts.

    Is it possible to extend storage patterns?

    Absolutely. It is possible to use SNMP or WBEM to extend what BMC provided patterns discover.

    Once the storage data is integrated into CMDB, how can it be used?

    The data provided enables storage components to be included in ITSM processes such as incident, change and configuration management and managed consistently across ITOM processes such as monitoring and automation.
    Storage discovery was developed in a consistent manner with storage modules for BMC Atrium Orchestrator, TrueSight Operations Management and Capacity Optimization. This guarantees that the systems in those tools can be matched with discovery data so that use cases like smartflows are seamlessly implemented and service-aware IT management is possible.

    Features

    Are there any out-of-the-box reports specific to storage?

    Storage related reports can be found in different places:

    • The Storage Overview dashboard offers out-of-the-box charts and access to storage-specific reports
    • The Infrastructure tab has a section for storage reports
    • From a Storage System or Storage Volume, a report can be accessed to display Consumer Hosts
    • From a host, two reports are available to show Volumes and Systems that are consumed by the host
    • From an application instance, two reports are available to show Volumes and Systems that are consumed by the application

    There are multiple ways to discover the relationship between a host and the storage it consumes:

    • The relationship with SAN is done by identifying the World Wide Port Name (WWPN) from a host's HBA card, and matching it with the WWPN of the corresponding volume in a storage array
    • Relationship with file systems uses techniques that differ based on the OS, leveraging multipath information, LUNs, NAA IDs etc. Currently Linux, Windows and VMware linkage is supported
    • For NAS, this is done by identifying servers that have mounted file systems as well as LUNs.

    Further TKU and releases will add other techniques.