Skip to end of metadata
Go to start of metadata

You are viewing an old version of this page. View the current version.

Compare with Current View Page History

Version 1 Next »

Group control is affected by group state. Following are the commands and attributes related to control a group.


Table of Contents

모든 리소스 모니터링 / 모든 리소스 모니터링 안함

그룹내의 모든 리소스들을 사용하거나(모든 리소스 모니터링) 혹은 사용하지 않는 기능(모든 리소스 모니터링 안함) 입니다.
기본적으로 그룹 생성 후 리소스를 생성하게되면 '모든 리소스 모니터링'으로 리소스를 사용합니다.
유지보수 작업 같은 사용자의 요구에 따라서 그룹내의 모든 리소스를 일시적으로 사용하지 않게 될 경우 
'모든 리소스 모니터링 안함' 기능으로 그룹내의 모든 리소스를 사용하지 않을 수 있습니다.
이 때 그룹내의 모든 리소스들이 사용할 수 없는 상태가 되므로 그룹 시작, 종료 등이 불가능합니다.
유지보수 작업이 끝나면 다시 그룹의 리소스를 시작 해야 하므로 "모든 리소스 모니터링" 기능으로 그룹내의 모든 리소스를 사용할 수 있도록 해야 합니다.
그룹내의 모든 리소스가 사용할 수 있게 되면 그룹의 시작, 종료를 할 수 있습니다.


Group Online/Offline

It initiates an online/offline command for all resources in the group.
This means that MCCS bring online all the resources by order of dependency among those in a group on a node. 
For more details about dependency, please refer to "Chapter 5. 3 Dependency Management" of this document.
If there is an onlined group on a node in a failover group, online command will not work. However this will be work if the group state is partial online.
Offline command enables a node which group state is online or partial online. It is used when take offline all the onlined resources or deletes some of the resources or safely power off/reboot a node.
From the group management view of MCCS console, select the group and right click as shown in the picture below, then select ‘Online’ or ‘Offline’, the nodes of the group will be enabled which can be applied. 
These commands can only be initiated to enabled nodes.


[Figure] Group Online


[Figure] Group Offline



Switch

This is manual switch over the group from a node to opposite node. It is useful when you need to keep online a service during unplanned downtime like O.S update, patch, job related server rebooting and etc.
Or when you completed cluster configuration, you can group failover test using this without node shutdown. 
This command can be changed by selecting the node on service. It can test if the failover works properly or testing if there is any problem after failover and so on.
This is called 'Switch over' and 'Switch' command. A group should be online or partial online to enable this and standby node should not be in a failure status.

From the group management view of MCCS console, select the group and right click as shown in the picture below, then select ‘Switch’, the node of the group will be enabled which can be applied.
This command can only be given to enabled nodes.

[Figure] Manual Group Failover View


자동 비활성화 해제

자동 비활성화 해제 기능은 정상적인 상황에서는 사용하지 않는 기능입니다.
MCCS 가 구성 된 두 대의 서버는 부팅되어 있고 운영체제의 기능은 정상적으로 동작하는 상태에서 다른 한 서버에서 어떤 이유에서든 MCCS가 구동될 수 없을 경우에 사용합니다.
이 때 구동중인 MCCS 는 다른 서버의 MCCS 와의 통신을 할 수 없으므로 리소스들의 상태 및 정보를 알 수 없기 때문에 리소스들은 비활성화 상태가 됩니다.
이러한 비활성화 된 리소스들은 관리자에 의해서 반드시 확인되어야 하고, 관리자에 의해서 확인된 상황에서 응급으로 구동중인 MCCS 에서 서비스를 시작하기 위해서 이 기능이 사용됩니다.
따라서 이 기능을 사용하는 것은 관리자가 비정상인 상태에서 응급조치로써 MCCS에서 확인되지 않은 리소스를 직접확인하고 구동중인 MCCS 서버에서 리소스를 온라인 할 수 있게 합니다.

Group Lock

Group lock is useful when you need to maintain like test, update and patch the resources. MCCS monitor the status of resources, but do not online, offline and failover for ‘lock’ state group.
For an example, when you patch your own developed ‘a.exe’ which is managed by MCCS, you may need ‘a.exe’ process down for maintenance. 
When you get planned downtime for this, MCCS will detect this as failure and is trying to restart that or failover the group to secondary node which you do not want to be.
When group lock is set, command for Online/Offline/Switch for group and Online/Offline/Offline Prop/Enable for resources are disabled. 
But command for Add/Delete/Probe/Clear Fault for resource are enabled.
There are ‘Temporary Lock’ and ‘Lock’ and it’s attribute can be checked on Group Attributes view. When MCCS service is stop, ‘Temporary Lock’ will be unlocked even after this is brought up again. 
But ‘Lock’ remains the same after do this. From the group management view of MCCS console, select the group and right click as shown in the picture below, then select ‘Lock’, You can choose Temporary Lock or Lock from this. 

[Figure] Lock Group View

Temporary Lock

Unlocks when both nodes of MCCS service is stop and restarted. 

Permanent Lock

Group is lock permanently even after MCCS service is stop and restarted. This will be unlock when you set ‘Unlock’ command from management console.



Clear Fault

When a failure is occurred on a resource and MCCS decide this is failed, this will be marked failed resource with   icon on its resource and group icon. 
This mark is needed to be clear after you fix the problem of the resource. It is not allowed re-failover or manual switch to original node without clear this.
From the group management view of MCCS console, select the group or the resource and right click as shown in the picture below, then select ‘Clear Fault’, the node of the group will be enabled which can be applied. 

This command can only be given to enabled nodes.

[Figure] Clear Fault



Flush

This is used when commands like Online/Offline/Switch for a group is pending for some reasons (e.g. hang up) and this is needed to be initialized. 
All the commands will be cancelled by using 'Flush'.
From the group management view of MCCS console, select the group and right click as shown in the picture below, then select ‘Flush’, the node of the group will be enabled which can be applied. 
This command can only be given to enabled nodes.

[Figure] Group Flush View


  • No labels