Thursday Apr 24, 2014

Proxy Controller High Availability

Proxy Controllers manage the bulk of your asset data. In the past, losing a Proxy Controller to a hardware failure meant losing all of that asset data, and the only way to resume management of the assets was to rediscover them using another Proxy Controller.

However, one of the enhancements in Ops Center 12.2 is full support for Proxy Controller High Availability. You can enable auto failover, and if a Proxy Controller goes down, its managed assets will automatically be migrated to another available Proxy Controller. If you're able to get the old Proxy Controller running, you can migrate the assets back, or migrate them to a new Proxy Controller once it's installed. You can migrate an asset to any Proxy Controller that's associated with the asset's network.

Take a look at the Proxy Controller High Availability section of the Administration Guide for more information.

Thursday Apr 17, 2014

Using the OCDoctor through the UI

The OCDoctor utility is a tool that's bundled along with Ops Center, and which you can also download from java.net. It has a wide variety of functions - it can check a system to see if it meets Ops Center's installation prerequisites, troubleshoot and fix some common issues, check a system's connectivity, and update itself.

You can download the OCDoctor and run it on a system from the command line, but it's also possible to use the OCDoctor's troubleshooting functions on managed systems through the UI. To do this, you select the asset and click the Self Diagnosis action in the Actions pane:


This option launches a Self Diagnosis job, which uses the OCDoctor to check the system for known issues. Once the job is complete, you can click on Self Diagnosis again to see the results and, if issues are found, attempt to fix them using the OCDoctor.


Take a look at the OCDoctor chapter and the Self Diagnosis procedure for information about the OCDoctor's other features.

Friday Mar 07, 2014

12.2 Release Documentation

A new version of Ops Center, version 12.2, has just been released. With it, we've made some improvements to the documentation (you can see all of the new documentation at Ops Center 12.2 documentation library). Here's a rundown of the big changes.

There's a section on the main docs page with bulletins about known issues and things you should be aware of before you install or upgrade. It's highlighted in the image below. You should read it.

Here are some of the big changes and improvements:
  • The What's New In This Release doc tells you about all of the new features, including improved OS Provisioning, LDOM management enhancements, and VLan Tagging.
  • The Known Issues section of the Release Notes tells you about the issues you should be aware of when you're upgrading or installing.
  • The Certified Systems Matrix lists the supported systems, some of which are new in 12.2.
  • There's a new Upgrade Guide, which includes end-to-end upgrade procedures for a variety of environments and preferences: HA or single Enterprise Controller, large or small, command line or UI.
  • The Ports and Protocols Guide lists all of the ports and protocols that Ops Center uses, which is good information if you're preparing to install.
  • There are quite a few new how-tos in the Deploy and Operate libraries, including how-tos for provisioning Oracle Solaris 10 and Oracle Solaris 11, and for provisioning root domains and I/O domains.
  • The Feature Reference Guide explains the product's main features, new and old, and the Feature Reference Appendix has some information that adds to that such as API information.
  • There's a new book that goes into detail on the use of the Ops Center command line interface.


Friday Feb 14, 2014

Adding an asset on a remote network

I saw an interesting question about discovering assets:

"I'm trying to discover a new chassis. I have three Proxy Controllers and all of them can ping the chassis, but only one has the firewall access to communicate with it. None of the Proxy Controllers has a direct connection to the network with the chassis, so that network isn't in the network list. How do I get the discovery job to route correctly?"

There is a solution in the discovery profile here. The discovery profile has a field for the target host or IP address, and a dropdown for the network. When you put the profile together, you supply the asset's IP address, and then select a network that has a route to the network with the chassis on it. Run that and you'll be able to discover the asset.

Thursday Feb 06, 2014

Assigning a User Roles for Specific Asset Groups

Ops Center's roles give you, as an administrator, a lot of fine-grained control over what each user can do. I heard a question recently that demonstrates this pretty well:

"Is it possible to set up a user defined Group and have a User Role restricted to only have specific roles for this group only?"

Yep. To do this, you select the user from the list in the Admin section and click Manage User Roles to bring up the wizard.


To assign a role to a specific group, you deselect the "Use the default Role associations" option. Then, when you go on to the next step, it lets you select which groups the role applies to:


In this case, the user will have the Asset Admin role only for the Example group. They'll be able to take all of the Asset Admin actions for the assets in that group.

Let me know if you have any questions.

Thursday Jan 23, 2014

How-To: Migrating Logical Domains

Using Ops Center, you can create Server Pools to allow for failover and migration of virtual systems such as Logical Domains. In some cases, you'll want to migrate systems manually, either to balance the load or to perform hardware maintenance.

There's a how-to in our library that walks you through the whole procedure, and explains what prerequisites and roles are required. It also gives you a pointer to some additional resources - the Feature Reference Guide has chapters about Oracle VM Servers for SPARC and Server Pools.

Thursday Jan 16, 2014

Database Failover Questions

In the post about installing Ops Center, one of the steps is setting up the database that Ops Center will use. You can use a co-located database on the EC system, or you can set up your own database and configure Ops Center to use it.

On that second option, I saw a couple of questions:

Can Ops Center be installed using a Data Guard customer managed database?

Data Guard is an Oracle tool for providing high availability and disaster recovery for an Oracle database. It can be used with Ops Center. The only thing to keep in mind is that, if your database fails over to a system with different connection parameters, you need to direct Ops Center to use the new system using the Changing the Customer-Managed Database Location procedure.

Can the RemoteDBProps.txt contain entries for both the active and standby databases?

No. The RemoteDBProps.txt file is used for the initial configuration. If you have a standby database in a separate location, you have to switch using the procedure described above.

Thursday Jan 09, 2014

Notifications

Notifications are one of the tools you can use to keep track of your systems. They are automatic messages that are sent to specific users when their conditions are met - such as when a critical incident occurs.

Each user has their own notification profile, which specifies what level of notifications are sent - whether it's everything that happens to an asset, all incidents, or just critical incidents. It also specifies where these notifications are sent - the user interface, a specified email address, or a specified pager. You can set different levels for each destination, so that critical incidents go to the pager but less crucial issues go to the email address. If you have the user admin role you can configure a notification profile using this procedure.

The Command Line Interface for Ops Center lets you view and delete notifications as well, although configuring a notification profile can only be done through the UI.

Thursday Jan 02, 2014

Installing and Configuring Ops Center

Our previous post was about planning an Ops Center installation. Once you've done the planning, you're ready to install and configure the software.

The initial installation is done from the command line. You put the correct bundle on the EC and PC systems, unpack it, and install it. Depending on your planning choices you'll use a few different options with the install commands.

Once the install tells you that it's done, your final step is to log in to the UI and configure the Enterprise Controller. During the config process you will choose a connection mode (disconnected mode for dark sites or connected mode if the EC has internet access), set up the libraries for Linux, Oracle Solaris 8-10, and Oracle Solaris 11 content, and set up DHCP for OS provisioning if you need it.

Installing Ops Center is a complex process, and this post is just an overview and not a quick-start. The Oracle Solaris Installation Guide and the Linux Installation Guide go into much greater detail. If you have specific questions about installing Ops Center, let me know.

Thursday Dec 12, 2013

New OCDoctor released

There's a new version of the OCDoctor available - 4.26. It makes the connectivity check option a bit clearer, and checks for several new issues in troubleshooting.

You can get the new version at https://updates.oracle.com/OCDoctor/OCDoctor-latest.zip. If you have an Ops Center installation in connected mode, an automated job will get the new version for you.

Thursday Dec 05, 2013

Using Connected Mode

When you get started using Ops Center, you choose a connection mode. You can use Disconnected Mode, which works for sites without internet access; Connected Mode, which uses an internet connection; or you can set up the connection mode later.

Using Ops Center in Connected Mode can make your life easier in many ways. It lets you download packages and updates, get automatic firmware updates, download the latest OCDoctor utility and Ops Center updates, and file service requests. However, if you're currently using Disconnected Mode or if you didn't pick one during config, there are a couple of things you need to do in order to start using Connected Mode.

First, you need to register your Enterprise Controller.

Next, supply your MOS credentials in the Authentications. These MOS credentials are used to access the Oracle Knowledge Base.

Once you've taken these steps, you can switch to Connected Mode.

Thursday Nov 21, 2013

Tracking the energy cost for your servers

Keeping energy costs down is an ongoing challenge in any datacenter. Knowing how much energy each of your systems consumes, and how much they each cost to run, is an important piece of data.

Well, Ops Center can provide this data - it can give you data about the energy consumption of monitored servers, and about how much that energy costs.

The first step in getting this working (apart from discovering the server assets) is providing the cost of energy per kilowatt-hour in your datacenter, which you do using the Edit Energy Cost wizard in the Administration section.

Once Ops Center has that bit of data, it can give you detailed information about how much each server costs to run. You can select an asset and click the Energy tab to view the system's power consumption, power policy, and total cost for one day, or click the Charts tab and see the system's power consumption over time.


Thursday Nov 14, 2013

Optimizing OS performance

With the right information, there's a lot that you can do to improve the performance of an operating system. However, it can be time-consuming to gather that information in a large environment with hundreds or thousands of operating systems, Oracle Solaris Zones, or Oracle VM Servers.

Ops Center's OS Analytics tools can help you gather the information that you need to tune your operating systems. For operating systems managed by Ops Center, you can view OS performance info like CPU and memory utilization, process details, and virtualization capacity. For Oracle Solaris systems, you can also view the status and details of SMF services.

We put together a how-to that walks you through the process of viewing all of this information.

Thursday Nov 07, 2013

Plugins in Ops Center and Cloud Control

Cloud Control just released an updated plugin for Oracle Virtual Networking, so I thought I'd mention a bit about how both Ops Center and Cloud Control use plugins.

On the Ops Center side, we have a plugin to connect Ops Center to Cloud Control, letting them share monitoring data. This guide explains how to install and use that plugin.

In Cloud Control, they have a more extensive collection of plugins, letting you link Cloud Control with a variety of other products. The Cloud Control library plug-in tab goes into more detail about how you can use these plugins in your environment.

Thursday Oct 31, 2013

Agent versus Agentless management

I got a couple of questions about Agentless asset management:

"What does agentless management do for an asset?"

Agentless management is one of the two ways that you can manage an operating system. Rather than installing an Agent Controller on the OS, agentless management uses SSH to regularly check the system and gather monitoring data. Many of the actions that would be available on an agent-managed system are available on an agentless system, but actions such as running reports or updating an Oracle Solaris 10 or Linux OS are not available. A table showing the capabilities of agentless management is here.

"What permissions does agentless management require?"

Agentless management still requires root credentials. If you can't log into the system as root, you can provide one set of credentials for the login, and then a set of root credentials to switch to.

About

This blog discusses issues encountered in Ops Center and highlights the ways in which the documentation can help you

Search

Archives
« April 2014
SunMonTueWedThuFriSat
  
1
2
3
5
6
7
8
9
11
12
13
14
15
16
18
19
20
21
22
23
24
25
26
27
28
29
30
   
       
Today