Guest

Cisco UCS Director

Release Notes for Cisco UCS Director Express for Big Data, Release 1.1

About Cisco UCS Director Express for Big Data

About Cisco UCS Director Express for Big Data

Cisco UCS Director Express for Big Data provides a single touch solution that automates deployment of Big Data infrastructure and provides a single management pane across both physical infrastructure and Hadoop software. It supports key Hadoop distributions including Cloudera, MapR, and Hortonworks.

Cisco UCS Director Express for Big Data delivers end-to-end automation of Hadoop cluster deployment that allows you to spin up and expand clusters on-demand. Configuration of the physical infrastructure, which includes compute, internal storage, network, and installation of operating system, Java packages, Hadoop along with provisioning of Hadoop services are handled automatically with minimal user input. This is achieved through the innovative Cisco UCS service profiles wherein both the physical infrastructure and Hadoop configuration are incorporated into a Hadoop cluster deployment profile.

System Requirements

The system requirements for Cisco UCS Director Express for Big Data is listed in the table below:
Table 1 VMware vSphere Minimum System Requirements

Component

Requirement

VMware

ESX or ESXi 4.x or 5.x

vCPU

4

Memory

Minimum = 4GB

High Loads and Scalability = 8 GB

Hard disk

40 GB

Supported Browser Versions

Cisco UCS Director Express for Big Data supports the following browsers and all browsers require the Adobe Flash Player plug-in version 11 or higher for those browsers:

  • Internet Explorer 8 or higher

  • Google Chrome 4.1 or higher

  • Firefox 3.5 or higher

  • Safari 4.0 or higher (for Apple Mac or Microsoft Windows)

Prerequisites

  • Cisco UCS Director, Release 5.3.1.2 patch. (Cisco UCS Director, Release 5.2.0.2 patch is the prerequisite for the Cisco UCS Director, Release 5.3.1.2)

  • Cisco UCS Director Express for Big Data Baremetal Agent, Release 5.2.

Installation Notes

Cisco UCS Director Express for Big Data is a 64-bit appliance that uses a standard virtual machine that is delivered in OVF format for VMware, and can be hosted on VMware vSphere or vCenter.

For installation instructions, see the following guides:
  • Cisco UCS Director Express for Big Data Installation on VMware vSphere, Release 1.1
  • Cisco UCS Director Express for Big Data Baremetal Agent Installation and Configuration Guide, Release 1.1

The installation guides are available on Cisco.com here: http:/​/​www.cisco.com/​en/​US/​products/​ps13050/​prod_​installation_​guides_​list.html.

Patch Installation Notes for Cisco UCS Director Express for Big Data Releases 1.x

You can use the cucsde_patch_1_1_0_3.zip patch file to upgrade both Cisco UCS Director Express for Big Data, Releases 1.0 and 1.1 to the current patch level.

Procedure
    Step 1   Install Cisco UCS Director Express for Big Data, Release 1.0 on Cisco UCS Director, Release 5.0 platform with the support for personality.
    Step 2   Log in to the current version of Cisco UCS Director Express for Big Data VM console.
    Step 3   In the shelladmin, choose Stop Services to stop all services.

    To verify that all services are stopped, choose Display Services Status.

    Note   

    You can skip Steps 4, 5, and 6 if you are already running Cisco UCS Director Express for Big Data, Release 1.1 on Cisco UCS Director, Release 5.2.0.2.

    Step 4   Choose Apply Patch to apply the patch application on Cisco UCS Director Express for Big Data.
    Step 5   When prompted, enter the location of the Cisco UCS Director 5.2.0.2 patch (cucsd_patch_5_2_0_2.zip) . (For example, ftp://username:password@hostname|IP_address/software_location_and_name).
    Step 6   Wait for the download and installation to complete for Cisco UCS Director, Release 5.2.0.2 patch.
    Step 7   Choose Apply Patch to apply the patch application on Cisco UCS Director Express for Big Data.
    Step 8   When prompted, enter the location of the Cisco UCS Director 5.3.1.2 patch (cucsd_patch_5_3_1_2.zip) . (For example, ftp://username:password@hostname|IP_address/software_location_and_name).
    Step 9   Wait for the download and installation to complete for Cisco UCS Director, Release 5.3.1.2 patch.
    Step 10   Choose Apply Patch to apply the patch application on Cisco UCS Director Express for Big Data.
    Step 11   When prompted, enter the location of the Cisco UCS Director Express for Big Data, Release 1.1.0.3 patch (cucsde_patch_1_1_0_3.zip). (For example, ftp://username:password@hostname|IP_address/software_location_and_name).
    Step 12   Wait for the download and installation to complete for Cisco UCS Director Express for Big Data, Release 1.1.0.3 patch.
    Step 13   When prompted, choose Start Services to start services and complete the patch upgrade process.
    Note   

    The patch process is not complete or successful until the Cisco UCS Director Express for Big Data services have started, Cisco UCS Director Express for Big Data is available, the login screen is displayed, and the admin user can log in to Cisco UCS Director Express for Big Data.

    All Cisco UCS Director Express for Big Data services must be started before you attempt to perform other shelladmin procedures, such as apply additional patches, take a database backup, or restore a database from a backup.

    Step 14   When the patch application is complete, choose Show Version in shelladmin to verify the current version of Cisco UCS Director Express for Big Data.
    Step 15   Log in to the Cisco UCS Director Express for Big Data Baremetal Agent (BMA) server using the default credentials (root/pxeboot). Copy the cucsde_bma_patch_5_3_1_2.zip (BMA patch) file and extract the BMA patch file. Change directory to scripts/bin, add execute permissions to copyfiles.sh (chmod+x copyfiles.sh), and execute the copyfiles.sh (./copyfiles.sh).
    Step 16   Log in to the Cisco UCS Director Express for Big Data GUI and start the BMA services.

    Open and Resolved Bugs

    The open and resolved bugs for this release are accessible through the Cisco Bug Search Tool. This web-based tool provides you with access to the Cisco bug tracking system, which maintains information about bugs and vulnerabilities in this product and other Cisco hardware and software products.


    Note


    You must have a Cisco.com account to log in and access the Cisco Bug Search Tool. If you do not have one, you can register for an account.


    For more information about the Cisco Bug Search Tool, see the Bug Search Tool Help & FAQ.

    Open Bugs for This Release

    All open bugs for this release are available in the Cisco Bug Search Tool. The results of that search include workarounds for the open bugs, if any.

    Defect ID

    Symptom

    Workaround

    CSCup20580

    With no data, the Hosts summary report is continuously trying to update the Role-CPU Usage and it is not showing any graph for the Managed node.

    Not available.

    CSCup65770

    Duplicate/multiple Delete Node submissions for the same selected node in the same cluster is accepted.

    Not available.

    CSCur20682

    Re-submission of Multi UCSM/Single UCSM workflows should resubmit all its child workflows.

    You can re-submit child and parent workflows individually from the Service Requests page.

    CSCus23190

    Validation does not happen when the Ganglia policy for Cloudera/MapR distribution is selected.

    Ganglia policy will not be visible for Cloudera and MapR distribution. You can ignore it.

    CSCut07395

    The value is not getting populated for the field "Select from existing UUID" for storage pools during cluster creation.

    Select any other field in the page which triggers the validation and makes UUID inputs visible.

    CSCut07971

    Validation does not happen to add node using Add Managed Node and Add Live Node simultaneously.

    Not supported. Recommended to wait for ten minutes between performing actions on nodes.

    CSCut07984

    Validation is required on Boot Order Policy in the UCS SP Template for Big Data.

    You can avoid modifying the boot order in the Boot Order Policy as no validation is currently in place in the UCS SP Template for Big Data.

    CSCut09390

    Hadoop cluster is not created for RAID0 OS and RAID1 Data nodes scenario.

    You have to configure RAID0 on Data nodes instead of RAID1 in the case of disks that belong to different types.

    CSCut11804

    The Cloudera Manager user interface displays SentryServer warning.

    Kerberos setup is not supported.

    CSCut17332

    The MapR 4.0.x Hadoop cluster Yarn statistics is not getting populated.

    Not supported by MapR distribution.

    CSCut15961

    MapR cluster Delete Baremetal Nodes result in alarms.

    Ignore volumes specific alarms post node deletion.

    CSCut17193

    Cloudera cluster Add/Delete Baremetal Nodes results in HDFS health issue.

    The issue is inconsistent. Perform Stop All / Start All services.

    CSCut19303

    Hortonworks 2.2 cluster Add/Delete Baremetal Node results in DataNode stopped.

    The issue is inconsistent. Perform Stop Data Node / Start Data Node.

    CSCuu30029

    Configure limit.config settings in Red Hat for Hadoop clusters.

    Update manually.

    Resolved Bugs in This Release

    Resolved Bugs in Release 1.1

    All resolved bugs for this release are available in the Cisco Bug Search Tool.

    Defect ID

    Description

    CSCuq14714

    Inputs that are given before starting to create a cluster in the Multi UCSM Managed cluster UI page (in a specific order using Instant Hadoop Cluster Creation) is not editable, once submitted.

    CSCur51778

    Node IP address is not released to the Big Data IP pool after performing the Delete Node to Baremetal action on that node.

    CSCur51789

    Deleted node is getting populated in the Add Managed Node dialog after performing the Delete Node to Baremetal action.

    CSCur76365

    Server pools in an Instant /Customized Hadoop cluster creation dialog are filtered based per UCS Manager.

    CSCus85952

    Server with enclosure ID 252 is not configured for RAID during RHEL boot.

    CSCus87823

    NTP server synchronization for the Hadoop cluster.

    Resolved Bugs in Patch Release 1.1(1)

    All resolved bugs for this release are available in the Cisco Bug Search Tool.

    Defect ID

    Description

    CSCut12988

    The Big Data Account tab displays that an account is added successfully with an incorrect IP address.

    CSCut25341

    Operating system to map drives with the device names such as sda, sdb, sdc is based on the UUID of the drives.

    CSCut29209

    Inventory collected before Hadoop cluster creation is complete.

    CSCut47360

    Mount point is not created for the cleared old partitions/mounts for M4 Servers.

    CSCut76907

    Hortonworks disk usage is less when nodes in a cluster have different disk count.

    CSCut77235

    Triggering a cluster should warn users if there is already a template available in the system.

    CSCut95006

    Make all DataNode roles selected as default for Customized Hadoop Cluster Template.

    Resolved Bugs in Patch Release 1.1(2)

    All resolved bugs for this release are available in the Cisco Bug Search Tool

    Defect ID

    Description

    CSCut71232

    Remove version dependancy for MapR to support MapR 4.1.0.

    CSCuu03997

    Displays incorrect UCS Service Profile Template ID when viewing the details in the Hadoop Clusters tab.

    CSCuu04044

    If the values populated in the fields of the Add BareMetal Nodes dialog box are untouched, then Add BareMetal Node fails when executing the Create UCS Service Profile Template task with an error displaying that the UCS Service Profile Template already exists.

    CSCuu19592

    Changes made in rc.local are not getting reflected like vm.swappiness and redhat_transparent_hugepage.

    CSCuu19598

    SSH key exchange should be done also for user account 'mapr' in MapR cluster.

    CSCuu28850

    MAC addresses are not assigned to vNICs as per the user selection in the instant or customized Hadoop cluster creation dialog box.

    CSCuu32722

    An instant Hadoop cluster require RAID 0 configuration for all data drives for MapR.

    Resolved Bugs in Patch Release 1.1(3)

    All resolved bugs for this release are available in the Cisco Bug Search Tool

    Defect ID

    Description

    CSCut94963

    Enable to use all the disks in a node and configure HDFS and YARN directories for newly added nodes to assign configurations to the groups globally by grouping nodes that have the same number of disks.

    CSCuv66637

    Cloudera cluster creation failed with the UCS CPA Node Baremetal workflow at Power On UCS Server.

    CSCuv79094

    Displays the same First MAC Address in the vNIC Template table irrespective of the MAC Address Pool selected in the Instant Hadoop Cluster or Customized Hadoop Cluster creation dialog box.

    CSCuv92616

    When you decide not to use boot drives for M4 servers, then the "Use HDD drives on M4 Servers with insufficient standalone boot drives" option is disabled.

    CSCuv92656

    Allow the user to modify host name prefix while adding additional nodes to the Hadoop cluster.

    CSCuv95615

    MAC Address Pool is not set for Add BareMetal Nodes flow.

    CSCuw02799

    Unable to add baremetal nodes to the Hadoop cluster with a new UCSM Policy Name prefix other than the default.

    CSCuw08610

    Custom cluster creation pages auto-populate Hadoop distributions and Hadoop distribution versions.

    Related Documentation

    In addition to these release notes, you can find documentation for Cisco UCS Director, Release 5.2 in the following locations on Cisco.com:

    Obtaining Documentation and Submitting a Service Request

    For information on obtaining documentation, using the Cisco Bug Search Tool (BST), submitting a service request, and gathering additional information, see What's New in Cisco Product Documentation, at: http:/​/​www.cisco.com/​c/​en/​us/​td/​docs/​general/​whatsnew/​whatsnew.html.

    Subscribe to What's New in Cisco Product Documentation, which lists all new and revised Cisco technical documentation as an RSS feed and delivers content directly to your desktop using a reader application. The RSS feeds are a free service.