Chapter 2 Use VCs

Chapter 2 Use VCs
Objectives:
Manage applications on the cluster Platform
VCs management tools
Service Group operations
Use VCs Simulators Manage applications on the cluster Platform
When an application is under VCs control, you must change your management behavior. 2. Basic management methods:
1. Use VCs to start and stop service groups and resources
2. VCs does not intervene directly when you are managing the VCs through the solidification Service Group. Tools for managing VCs
Four types:
1. A simulator creates, modifies, and tests configurations.
2. Java graphical user interfaces on UNIX and Windows systems
3. A web-based graphical excuse to Support Web browsers at the same time
4. Command Line Service Group operations
Display attributes and status
Bring the service group online
Offline Service Group
Exchange Service Group
Solidification Service Group
Bring up resources online
Offline Resources
Clear fault Display attributes and status
You can view the attributes of cluster; system; Service Group; resources types.
The status information is displayed
1. Determine the cluster status
2. Analyze the errors and correct them as needed
Command:
Hastatus-sum [Mary]
Snapshot)
Hastatus
Displays the target status of a continuously updated cluster. Show logs
Had-Log
At/var/vrtsvcs/log/engine_a.log
Track all cluster activities
Helps solve configuration problems
Command-Log
Use GUI and CLI to track each released command
Help to learn CLI
Batch files can be created
Can be printed, but cannot be stored as a file on a disk Bringing service groups online
When a service group is brought online, the resources are brought up, and the relationship tree is composed of the bottom-layer child resources to the top-level parent resources, from the bottom up.
To bring a Failover service group online, all non-permanent resources in the cluster must be offline first. If any non-permanent resource is online in another system, the Service Group cannot be brought online.
A service group is considered to be online. If not less than its Automatic startup and related resources are online:
1. Set the autostart attribute of an autostart resource to 1.
2. Set the critical attribute of a critical resource to 1.
A service group is considered partially online if one or more non-permanent resources are online and at least one resource is:
1. autostart-enable
2. Critical
3. offline
Command:
Hagrp-Online service_group-sys System
Hagrp-Online service_group-any
-Any option: Bring the Service Group automatically started in the system Taking service groups offline
When a service group is offline, the resources start from top down and the parent resources start. The resources of each branch are offline based on the relationship tree until the last child Resource
Permanent resources cannot be offline (for example, nic). Therefore, the Service Group determines that offline resources are all non-permanent resources offline.
Command:
Hagrp-Offline service_group-sys System
Hagrp-Offline service_group-any Switching service groups
To ensure that failover is successful, you can test the switching service group between systems in the cluster.
Command:
Hagrp-switch service_group-to System Freezing a Service Group
A Service Group is solidified to prevent it from going offline, going online, or failover.
When you solidify a service group, VCs continues to monitor this resource, but does not allow the Service Group (or its resources) to be backed offline or online. failover is also prohibited, even if the resource fails.
Note: When solidification occurs, VCs cannot take actions on the Service Group even if you bring the service online on another system, resulting in parallel disruption.
Command:
Hagrp-freeze service_group
Hagrp-Unfreeze service_group
To permanently solidify a service group, you must enable the configuration first.
Haconf-makerw
Hagrp-freeze service_group-persistent
Hagrp-Unfreeze service_group-persistent
To determine whether a service group is solidified and shows frozen (permanent) and tfrozen (zero-Time) Service Group attributes for a service group
Hagrp-display service_group-attribute frozen Bringing resources online
Command:
Hares-online resource-sys System Taking resources offline
Offline resources. This can easily lead to a partially online service group, and the availability of the application is also affected.
1. If a resource needs to be offline, for example, to maintain the underlying hardware, switch the service group to another system.
2. If multiple resources need to be manually offline, they must follow the link tree from top to bottom.
Command:
Hares-Offline resource-sys System Clearing resource faults Use VCs Simulators
Simulators can run on UNIX and Windows platforms.
You can perform all the general operations and management jobs in the JAVA-GUI or CLI
You can use the pre-configuration file main. cf or create a service group and resources.
You can check how VCs reacts with a simulated error -------------------------------------------------------------------------------: under which circumstances should a Service Group be solidified ( Freezing a Service Group)? From: http://hi.baidu.com/nitar/blog/item/1ddfb6a98d249cbdca130cac.html