This section describes the new features and enhancements that are available in BMC Discovery version 11.0. You can upgrade to BMC Discovery version 11.0 from versions 10.1 and 10.2. The upgrade procedure is described in Upgrading BMC Discovery.
If you are new to BMC Discovery, BMC recommends that you refer to Getting started as an introduction to using the product.
The following topics are covered in this section:
What's new in BMC Discovery version 11.0?
This section describes the new features and enhancements that have been introduced in BMC Discovery version 11.0.
Start anywhere application modeling
BMC Discovery version 11.0 introduces Start anywhere application modeling, this is a new approach to application modeling which enables you to choose any entry point, or points into an application and begin modeling from there. An application may be large, and sensible entry points differ depending on the view of the user. For example, an application owner may choose where the data is stored as the best entry point, and a user of the application may choose the server to which they connect, to access the application. A start anywhere approach also avoids parts of applications being missed as they are not currently connected to an entry point, such as a URL which may lead to a load balanced service or web server. You may also choose multiple entry points to model the application.
Start anywhere application modeling is based around the software visualization. The software visualization enables you to explore the nodes connected to the entry point or points, and select those that are interesting in the context of the application. You then save the view as a model definition that you can access through the application modeling home page, for later use, or you can continue to model. When you are satisfied that the definition represents the application, you can publish the model, which creates the BAI.
Essentially, the modeling process is this:
- Search for something in your application.
- View it in the visualize - software window.
- Add or remove components as required (usually remove).
- Save the definition.
- Publish, and view the BAI.
For a more detailed procedure about how to create a model, see Creating an application model.
Collaborative application mapping is still supported, as is application mapping using hand written TPL. Start anywhere application modeling is an additional feature, not a replacement feature.
The improvements in cluster discovery and modeling come in the following streams:
- Clustering software—the software that provides clustering capabilities. At the release of BMC Discovery 11.0, the following clustering technologies are supported:
- Oracle Clusterware
- HP ServiceGuard Cluster Server
- Microsoft Cluster Server
- Symantec Veritas Cluster Server
Clustered software—the software that runs on a cluster. The TKU team is updating patterns which use the new capabilities.
"Self-clustering" software—software that provides its own clustering capability, for example, BMC Discovery clusters. Self-clustering software is now modeled using a new node kind, the Software Cluster node, which acts as a logical container for the SIs representing the software running on individual hosts.
Cluster model changes
BMC Discovery 11.0 changes how clusters are modeled. In previous versions, the software running on clusters was modeled as SIs running on each host in the cluster. In version 11, the software running on a cluster is modeled as a single SI running on the cluster itself. At the time of release not all TKU patterns have been updated to build this new model. As updated patterns are activated and used to discover existing clusters they will automatically change the inferred model. When you scan a host which is a member of a cluster that has software recognized by one of these updated patterns, the system:
- creates a new SI linked to the cluster (rather than the host),
- remembers the value of the
typeattribute of the created SI,
- searches all hosts in the cluster for SIs that have the same
typeattribute value and deletes them.
Notes on cluster discovery
- Discovering full details of Windows Clusters requires you to use a BMC Discovery 11.0 Windows Proxy. Previous Proxy versions are able to discover Windows clusters, but not all the details of the clustered services.
- If you are upgrading, remember to activate the new patterns supplied by the upgrade - otherwise you will not see these changes.
EMC VPLEX discovery
This release introduces the ability to discover EMC VPLEX storage devices. The EMC VPLEX storage discovery pattern requires an additional type of credential, EMC VPLEX credentials.
discovery.restfulPost– Performs a POST request on the target using the RESTful system and returns a node containing information on the discovered system.
The BMC Discovery UI has been updated to give a cleaner, more modern look and feel. This is the first step on moving the UI to a task-based information architecture. See Navigating the BMC Discovery UI for more information on the layout and menus. Use the mapping table to see how to find previous UI equivalents in the new UI.
Unified credentials UI
The credentials UI has been substantially revised and simplified, in line with the ongoing UI updates. All device credentials are now configured on a single page, as are database credentials. Windows proxies are also managed on a single page. At the end of the upgrade the credentials must be modified to fit in with the new credential UI. To complete the upgrade, you must open the credentials vault using the existing passphrase. You can change the passphrase when the upgrade is complete.
The credential vault has always been securely encrypted with a passphrase. New in this version is an option to save the passphrase.
In previous versions of BMC Discovery, information on credential usage was displayed in a color-coded format on the Credentials page. In practice, the information required to diagnose discovery problems is not on a credential basis; it is, "which discovery accesses are failing?", "which accesses are failing because of credentials failures?", "which device identities have changed?", and "which devices are not responding and aging out of the model". This information is now displayed in a dashboard display on the Discovery status page (Manage > Discovery). This change results in improved load times for the Credential page, particularly in larger systems.
In BMC Discovery version 11.0 we have improved discovery of VMware ESX and ESXi hosts using vCenter. In earlier releases you needed to specify the IP addresses of ESX/ESXi hosts that could be discovered by a vCenter server. From BMC Discovery version 11.0, when a vCenter server is discovered, it is queried to find the ESX/ESXi hosts that it manages, and scans of those ESX/ESXi hosts are started automatically. This is indirect access, in which the results of one scan can cause the scan of additional IP addresses. Discovery using vCenter is now the preferred method of discovering ESX/ESXi hosts. Where an ESX/ESXi host is discovered through vCenter, and discovery is subsequently attempted using a vSphere credential, the IP address is skipped.
The following new TPL functions have also been provided to enable you to write custom patterns to discover VMware ESX and ESXi hosts using the VMware vSphere API.
Their use is described in VMware ESX and ESXi discovery using discovery.vSphere functions.
In previous releases you could develop custom patterns to discover information about the VMware ESX and ESXi hosts from the VMware vSphere API. To do this required definition of blocks and used integration points, while this is still supported, and described here, you can now perform the same discovery more simply using the discovery.vSphere* functions.
Software and infrastructure visualizations
The software context view has been extended to provide software and infrastructure visualizations. The infrastructure visualizations replaces the existing dependency visualizations, providing a simple editable view of the infrastructure related to any node you use to start viewing it.
The Visualize - Software view focuses on software and communication between items of software and shows some details of the hosts and load balancers involved. The Visualize - Infrastructure view focuses on the connectivity to network and storage components. It shows the switches a host is connected to, underlying storage for host filesystems, VMs on VM servers, or if the given host is a VM, the server on which it is hosted.
Processor information discovery
Information on host and host container processors is now discovered using patterns rather than normal discovery commands. The processor information is stored in a Processor Info node and can be views on the Processor information page. Using this approach, the attributes that were the least reliably discovered can be determined using a table of known values for particular processors, taken from the manufacturer's data. The attributes that are taken from the manufacturer's data are Cores per socket and Threads per core. The patterns will be updated for new processors and packaged with a monthly TKU release.
The following miscellaneous features have been added:
Selective baselining from the command line
The tw_baseline utility has been extended to enable you to perform selective baselines from the command line. The following options have been added:
--filter– limits the checks that are run
--list-checks– lists available baseline checks
LDAP user image
You can now specify in LDAP configuration that the system looks up an Active Directory user image. The retrieved image is displayed next to the username of the logged in user at the top of the BMC Discovery UI.
CMDB synchronization improvements
Two relationships are no longer created in the CMDB:
- the direct relationship between host and application (BMC_ComputerSystem CI to BMC_Application CI)
- the direct relationship (
HOSTEDVIRTUALSYSTEM) between physical host and hosted virtual host (BMC_ComputerSystem CI to BMC_ComputerSystem CI). The direct route from ComputerSystem to ComputerSystem was a duplication of the path through the VirtualSystemEnabler.
For more information, see CDM Mapping for Host.
New TPL functions
The following new TPL functions have been added:
- json.encode – Converts a string, table, or a list (including nested structures) to a JSON encoded string.
- json.decode – Decodes a JSON encoded string and returns a structure containing a string, table or list including nested structures.
- model.kind – Returns the node kind of a given node.
- related.hostingNode – Returns a node that is associated with the given triggers and is hosting the software instance.
BMC Atrium Single Sign-On integration
Integration with BMC Atrium Single Sign-On is deprecated in the version 11 release of BMC Discovery. In a future release support will be added to support integration with Remedy Single Sign-On.
The CiscoWorks import tool is deprecated in the version 11 release of BMC Discovery. It will be removed in a future release.
What's new in BMC Discovery version 11.0 patch 1?
The following changes were made in BMC Discovery version 11.0 patch 1:
- Start Anywhere Application Mapping has been enhanced to improve the user experience.
- vCenter discovery performance has been improved.
- CMDB synchronization mapping for clustered software has been updated.
- A number of defects were resolved.
- TPL version updated to TPL 1.12. It is a minor update to TPL version 1.10 which consists of some internal fixes. From the April 2016 TKU release, monthly TKUs for BMC Discovery 11.0.x require that you upgrade the system to version 184.108.40.206 (Patch 1).