Thursday Aug 29, 2013

Oracle Solaris 11 Zone support

I got a question about Oracle Solaris 11 Zones:

"I manually created an Oracle Solaris 11 Zone. I can discover it and bring it into Ops Center, but the Migrate Zone action is greyed out, whereas other zones created through Ops Center have all of their actions enabled. What's the difference? I thought that all zones were supposed to be equal in 12c."

All zones are equal in 12c, but sometimes it seems like some zones are more equal than others.

What's going on here is that, with a manually created zone, the zone metadata is stored locally and the zone storage isn't fully managed. To enable the Migrate Zone action, you need to use the Move Storage action to move the zone metadata and storage. Once you've done that, the storage will be managed and, as long as another global zone is available, the Migrate Zone action should be enabled.

Thursday Aug 22, 2013

Ops Center CLI functions

I see questions from time to time about the Ops Center CLI and what you can do with it, so I thought I'd mention a couple of resources.

The Ops Center CLI lets you connect to an Enterprise Controller and perform many tasks from the command line. These tables from the CLI chapter show the features that can be accessed through the CLI; many but not all of the product's features are CLI-accessible.

That same chapter also explains how to log in, how to use the different CLI modes, and how to run scripted jobs through the CLI.

Tuesday Aug 13, 2013

Database locations

I got a question about customer-managed database locations:

"We are preparing to install Ops Center with ECHA. The install docs say that ECHA uses a customer-managed database on a separate system. What if I partitioned a system, and had the EC node and DB node on separate partitions? Could that work?"

Theoretically yes, although I don't think we've tested that kind of setup.

The reason we say that the database should be on a separate system is because an Enterprise Controller failover doesn't include failover for the database. We assume that, with a customer-managed database, you'll have your own methods for protecting the DB from system failures. ECHA, then, just focuses on the Enterprise Controller, and assumes that the database will still be accessible when the new EC node comes up.

Thursday Aug 08, 2013

Upgrade question

I got a question about upgrading to 12.1.4:

"I'm trying to do a manual upgrade from 12.1.3 to 12.1.4. I've upgraded the Enterprise Controller, but now I can't find the Proxy Controller upgrade bundle. Where is that?"

This depends on how you're downloading the upgrade bundles.

If you're getting them through the UI, they're put in the /var/opt/sun/xvm/images/productUpdate directory. You can copy them from there to the Proxy Controller systems and do a manual upgrade (or upgrade through the UI if you're so inclined.)

You can also get the Proxy Controller bundles from either the OTN or e-Delivery, and then copy them to the Proxy Controller system.

However you do the download, we recommend that once you've started upgrading your environment, you upgrade the whole thing as soon as you can.

Thursday Aug 01, 2013

LDoms and Maintenance Mode

 I got a few questions about how maintenance mode works with LDoms.

"I have a Control Domain that I need to do maintenance on. What does being put in maintenance mode actually do for a Control Domain?"

Maintenance mode is what you use when you're going to be shutting a system down, or otherwise tinkering with it, and you don't want Ops Center to generate incidents and notification of incidents. Maintenance mode stops new incidents from being generated, but it doesn't stop polling, or monitoring, the system and it doesn't prevent alerts.

"What does maintenance mode do with the guests on a Control Domain?"

If you have auto recovery set and the Control Domain is a member of a server pool of eligible systems, putting the Control Domain in maintenance mode automatically migrates guests to an available Control Domain.  When a Control Domain is in maintenance mode, it is not eligible to receive guests and the placement policies for guest creation and for automatic recovery won't select this server as a possible destination. If there isn't a server pool or there aren't any eligible systems in the pool, the guests are shut down.

You can select a logical domain from the Assets section to view the Dashboard for the virtual machine and the Automatic Recovery status, either Enabled or Disabled.

To change the status, click the action in the Actions pane.

"If I have to do maintenance on a system and I do not want to initiate auto-recovery, what do I have to do so that I can manually bring down the Control Domain (and all its Guest domains)?"

Use the Disable Automatic Recovery action.

"If I put a Control Domain into maintenance mode, does that also put the OS into maintenance mode?"

No, just the Control Domain server. You have to put the OS into maintenance mode separately.

"Also, is there an easy way to see what assets are in maintenance mode? Can we put assets into, or take them out of, maintenance mode on some sort of group level?"

You can create a user-defined group that will automatically include assets in maintenance mode. The docs here explain how to set up these groups. You'll use a group rule that looks like this:


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


« August 2013 »