- Preface
- Cisco UCS Central Overview
- License Management
- Managing Administrative Settings
- Domain Management
- Remote Management
- Firmware Management
- Monitoring Inventory
- Managing Backup and Restore
- Working with Policies
- Service Profiles and Templates
- Server Configuration
- Network Configuration
- Storage Configuration
- Statistics Management
- System Management
- Monitoring Logs
- User Management
System Management
This chapter includes the following sections:
- Managing DNS Policies
- Managing Power Policies
- Managing Time Zones
- SNMP Policies
- About High Availability in Cisco UCS Central
Managing DNS Policies
Cisco UCS Central supports global DNS policies defining the DNS server and domain name. Registered Cisco UCS domains choosing to define DNS management globally within that domain's policy resolution control will defer DNS management to its registration with Cisco UCS Central.
- Configuring a DNS Policy
- Deleting a DNS Policy
- Configuring a DNS Server for a DNS Policy
- Deleting a DNS Server from a DNS Policy
Configuring a DNS Policy
Before configuring a DNS policy in a domain group under the Domain Group root, this policy must first be created. Policies under the Domain Groups root were already created by the system and ready to configure.
Deleting a DNS Policy
Deleting a DNS policy will remove all DNS server settings within that policy.
Configuring a DNS Server for a DNS Policy
Configure a DNS policy.
Deleting a DNS Server from a DNS Policy
Managing Power Policies
Cisco UCS Central supports global equipment policies defining the global power allocation policy (based on policy driven chassis group cap or manual blade level cap methods), power policy (based on grid, n+1 or non-redundant methods). Registered Cisco UCS domains choosing to define power management and power supply units globally within that client's policy resolution control will defer power management and power supply units to its registration with Cisco UCS Central.
- Configuring a Global Power Allocation Equipment Policy
- Deleting a Global Power Allocation Equipment Policy
- Configuring a Power Equipment Policy
- Deleting a Power Equipment Policy
Configuring a Global Power Allocation Equipment Policy
Before configuring a global power allocation equipment policy under a domain group, this policy must first be created. Policies under the Domain Groups root were already created by the system and ready to configure.
Deleting a Global Power Allocation Equipment Policy
Configuring a Power Equipment Policy
Before configuring a power equipment policy under a domain group, this policy must first be created. Policies under the Domain Groups root were already created by the system and ready to configure.
Deleting a Power Equipment Policy
Managing Time Zones
Cisco UCS Central supports global date and time policies based on international time zones and defined NTP server. Registered Cisco UCS Manager clients choosing to define date and time globally within that client's policy resolution control will defer the configuration for date and time to its registration with Cisco UCS Central.
- Configuring a Date and Time Policy
- Deleting a Date and Time Policy
- Configuring an NTP Server for a Date and Time Policy
- Configuring Properties for an NTP Server
- Deleting an NTP Server from a Date and Time Policy
Configuring a Date and Time Policy
Before configuring a date and time policy under a domain group, this policy must first be created. Policies under the Domain Groups root were already created by the system and ready to configure.
Deleting a Date and Time Policy
A date and time policy is deleted from a domain group under the domain group root. Date and time policies under the domain groups root cannot be deleted.
Deleting a date and time policy will remove all NTP server settings within that policy.
Configuring an NTP Server for a Date and Time Policy
To configure an NTP server for a domain group under the domain group root, a date and time policy must first have been created.
Configuring Properties for an NTP Server
An existing NTP server's properties may be updated before saving an NTP server instance. To change the name of an NTP server that is saved, it must be deleted and recreated.
Deleting an NTP Server from a Date and Time Policy
SNMP Policies
Cisco UCS Central supports global SNMP policies enabling or disabling, defining SNMP traps and SNMP users (with regular and privacy passwords, authentication types of md5 or sha, and encryption types DES and AES-128). Registered Cisco UCS domains choosing to define SNMP policies globally within that client's policy resolution control will defer all SNMP policies to its registration with Cisco UCS Central.
The SNMP Agent functionality provides the ability to remotely monitor the Cisco UCS Central. You can also change the Cisco UCS Central host IP, and then restart the SNMP agent on the new IP. SNMP is run on both the active and standby Cisco UCS Central servers and the configuration is persisted on both. Cisco UCS Central offers read-only access to only the operating system managed information base (MIB).Through the Cisco UCS Central CLI you can configure the community strings for SNMP v1, v2c, and create and delete the SNMPv3 users.
- SNMP Functional Overview
- SNMP Support in Cisco UCS Central
- SNMP Notifications
- SNMP Security Features
- Configuring an SNMP Policy
SNMP Functional Overview
The SNMP framework consists of three parts:
- An SNMP manager—The system used to control and monitor the activities of network devices using SNMP.
- An SNMP agent—The software component within Cisco UCS Central, the managed device, that maintains the data for Cisco UCS Central and reports the data, as needed, to the SNMP manager. Cisco UCS Central includes the agent and a collection of MIBs. To enable the SNMP agent and create the relationship between the manager and agent, enable and configure SNMP in Cisco UCS Central.
- A managed information base (MIB)—The collection of managed objects on the SNMP agent. Cisco UCS Central supports only the OS MIBs.
- RFC 3410 (http://tools.ietf.org/html/rfc3410)
- RFC 3411 (http://tools.ietf.org/html/rfc3411)
- RFC 3412 (http://tools.ietf.org/html/rfc3412)
- RFC 3413 (http://tools.ietf.org/html/rfc3413)
- RFC 3414 (http://tools.ietf.org/html/rfc3414)
- RFC 3415 (http://tools.ietf.org/html/rfc3415)
- RFC 3416 (http://tools.ietf.org/html/rfc3416)
- RFC 3417 (http://tools.ietf.org/html/rfc3417)
- RFC 3418 (http://tools.ietf.org/html/rfc3418)
- RFC 3584 (http://tools.ietf.org/html/rfc3584)
SNMP Support in Cisco UCS Central
Support for MIBs
Cisco UCS Central supports read-only access to OS MIBs. No set operations are available for the MIBs. The following MIBs are supported by Cisco UCS Central:
- SNMP MIB-2 System
-
HOST-RESOURCES-MIB -
UCD-SNMP-MIB -
SNMP MIB-2 Interfaces - IP-MIB
-
SNMP-FRAMEWORK-MIB - IF-MIB
- DISMAN-EVENT-MIB
- SNMP MIB-2 snmp
![]() Note | Cisco UCS Central does not provide support for IPV6 andCisco UCS Central MIBs. |
Authentication Protocols for SNMPv3 Users
Cisco UCS Central supports the following authentication protocols for SNMPv3 users:
AES Privacy Protocol for SNMPv3 Users
Cisco UCS Central uses Advanced Encryption Standard (AES) as one of the privacy protocols for SNMPv3 message encryption and conforms with RFC 3826. If AES is disabled but privacy password is set, then DES is used for encryption.
If you enable AES-128 configuration and include a privacy password for an SNMPv3 user, Cisco UCS Central uses the privacy password to generate a 128-bit AES key. The AES privacy password can have a minimum of eight characters. If the passphrases are specified in clear text, you can specify a maximum of 64 characters.
SNMP Notifications
A key feature of SNMP is the ability to generate notifications from an SNMP agent. These notifications do not require that requests be sent from the SNMP manager. Notifications can indicate improper user authentication, restarts, the closing of a connection, loss of connection to a neighbor router, or other significant events.
Cisco UCS Central generates SNMP notifications as traps. Traps are less reliable because the SNMP manager does not send any acknowledgment when it receives a trap, and Cisco UCS Central cannot determine if the trap was received.SNMP Security Features
SNMPv3 provides secure access to devices by a combination of authenticating and encrypting frames over the network. SNMPv3 authorizes management operations only by configured users and encrypts SNMP messages. The SNMPv3 User-Based Security Model (USM) refers to SNMP message-level security and offers the following services:
- Message integrity—Ensures that messages have not been altered or destroyed in an unauthorized manner and that data sequences have not been altered to an extent greater than can occur non-maliciously.
- Message origin authentication—Ensures that the claimed identity of the user on whose behalf received data was originated is confirmed.
- Message confidentiality and encryption—Ensures that information is not made available or disclosed to unauthorized individuals, entities, or processes.
SNMP Security Levels and Privileges
SNMPv1, SNMPv2c, and SNMPv3 each represent a different security model. The security model combines with the selected security level to determine the security mechanism applied when the SNMP message is processed.
The security level determines the privileges required to view the message associated with an SNMP trap. The privilege level determines whether the message needs to be protected from disclosure or authenticated. The supported security level depends upon which security model is implemented. SNMP security levels support one or more of the following privileges:
- noAuthNoPriv—No authentication or encryption
- authNoPriv—Authentication but no encryption
- authPriv—Authentication and encryption
SNMPv3 provides for both security models and security levels. A security model is an authentication strategy that is set up for a user and the role in which the user resides. A security level is the permitted level of security within a security model. A combination of a security model and a security level determines which security mechanism is employed when handling an SNMP packet.
SNMP Security Models and Levels
The following table describes the combinations of SNMP security models and levels supported in Cisco UCS Central.
|
Model |
Level |
Authentication |
Encryption |
What Happens |
|---|---|---|---|---|
|
v1 |
noAuthNoPriv |
Community string |
No |
Uses a community string match for authentication. |
|
v2c |
noAuthNoPriv |
Community string |
No |
Uses a community string match for authentication. |
|
v3 |
noAuthNoPriv |
Username |
No |
Uses a username match for authentication. |
|
v3 |
authNoPriv |
HMAC-MD5 or HMAC-SHA |
No |
Provides authentication based on the Hash-Based Message Authentication Code (HMAC) Message Digest 5 (MD5) algorithm or the HMAC Secure Hash Algorithm (SHA). |
|
v3 |
authPriv |
HMAC-MD5 or HMAC-SHA |
DES |
Provides authentication based on the HMAC-MD5 or HMAC-SHA algorithms. Provides Data Encryption Standard (DES) 56-bit encryption in addition to authentication based on the Cipher Block Chaining (CBC) DES (DES-56) standard. |
Configuring an SNMP Policy
Before configuring a SNMP policy under a domain group, ensure that a SNMP policy is first created. Policies under the Domain Groups root which were already created by the system and are ready to configure.
What to Do Next
Create SNMP traps and SNMP users.
Creating an SNMP Trap
Creating an SNMP User
Deleting an SNMP Policy
A SNMP policy is deleted from a domain group under the domain group root. SNMP policies under the domain groups root cannot be deleted.
Deleting an SNMP policy will remove all SNMP trap and SNMP User settings within that policy.
Deleting an SNMP Trap
Deleting an SNMP User
About High Availability in Cisco UCS Central
Cisco UCS Central provides high availability in a cluster setup when you deploy Cisco UCS Central in two virtual nodes. High availability provides stability and redundancy directly to your Cisco UCS Central and indirectly to your Cisco UCS Domains management. The high availability in Cisco UCS Central provides you the following:
- Simplified large scale Cisco UCS deployments with an increased number of servers, chassis, fabric interconnects, and data centers.
- UCS Central VM redundancy in a Hypervisor independent environment.
- A shared storage device to house database and image repositories.
- Built-in failure detection (DME, VM, host, or network failures) and automatic failover to ensure continuous operation.
High Availability Architecture
You will deploy Cisco UCS Central in two VMs on separate hosts to enable high availability. High availability -
- Requires at least one Cisco UCS Manager be registered with Cisco UCS Central for a cluster to support high availability
- Uses the same subnet for individual VMs and VIP addresses
- Allows you to configure a mirrored, multi-path shared storage disk on each VM that is accessible from both hosts
- Uses UCS Manager to store quorum data and determine primary node.
- Exchanges information such as heartbeat and election protocols in the same way as Cisco UCS Manager. This results in a simpler design, more code reusability, and easy to define failover conditions
Cautions and Guidelines for Using High Availability
The following are the guidelines to setup Cisco UCS Central in high availability:
- Make sure both VMs in the cluster should never be on the same server. Otherwise, a single host failure would end up bringing down the cluster.
- Each node in the cluster must have the following:
-
Separate network path for
management and storage network: Make sure the management network used
communications between the two Cisco UCS Central nodes are not on the same
network as the network that the nodes use to access the shared disk array. The
primary heartbeat mechanism relies on exchanging datagrams across the
management network. The secondary heartbeat mechanism uses quorum data on Cisco
UCS Manager. When you use separate network paths for management and shared disk
access, that provides redundant paths between the two nodes making it easier to
distinguish node failures from link failures.

Note
High availability is supported only in IPv4 addressing without the DHCP. You must configure the node IPs and cluster VIPs statically during the installation. These IP addresses are allocated from the production network over which the UCS Central cluster communicates with UCSMs.
- Both VMs must be configured on IP addresses that belongs to the same subnet.
- Make sure the cluster node infrastructure does not have a single point of failure. You can connect the cluster nodes my multiple, distinct networks. You can also construct the network with redundant switches and routers or similar hardware that removes single points of failure.
- For high availability Cisco UCS Central supports the most commonly used bus types, such as SAS , Fiber Channel (FC), and iSCSI. SCSI compatibility with Persistent Reservations (PRs) is recommended. LUN masking or zoning should be used to isolate the storage volumes accessed by the cluster from other hosts on the network.
Feedback