Shulou(Shulou.com)06/01 Report--

Problem description:

In a RAC environment, you want to open archive logging, where you use the

SQL > alter system set cluster_database=false scope=spfile;-- > modify the cluster to a single instance

After the master node is modified, the cluster double nodes are restored.

SQL > alter system set cluster_database=true scope=spfile;-- > modify the cluster to double instances

The current database master node has open status, and the query status is open

The standby node is already in shutdown status. Launch the current node Times:

ORA-01102: cannot mount database in EXCLUSIVE mode

After consulting the information from the Internet, it may be as follows. I will copy the relevant solution ideas and you can refer to them.

Analyze the reasons:

1. In the HA system, other nodes have started instances to take up resources shared by both computers (such as bare devices on the disk array and clusters that are not turned on correctly).

2. It indicates that when oracle is abnormally closed, some resources are not released. Generally speaking, there are several possibilities

1. The shared memory segment or semaphore of Oracle is not released.

2. The background processes of Oracle (such as SMON, PMON, DBWn, etc.) are not closed.

3. The files lk and sgadef.dbf used to lock the memory have not been deleted.

The solution is as follows:

When a 1102 error occurs, you can check and troubleshoot according to the following process:

If it is a HA system, check whether other nodes have started an instance to check whether the Oracle process exists. If so, kill the process to check whether the semaphore exists. If so, clear the semaphore to check whether the shared memory segment exists. If so, clear the shared memory segment to check whether the locked memory files lk and sgadef.dbf exist. If so, delete them.

Specific methods:

First of all, although our system is a HA system, the instance of the standby node is always closed, which can be confirmed by checking the database status on the standby node.

Secondly, it is because the system has modified the cluster status and whether it did not take effect after the modification. At that time, it was to query the current cluster status.

SQL > show parameter clusterNAME TYPE VALUE-- cluster_database boolean fales

Cluster_database_instances integer 2cluster_interconnects string

Here, we may roughly judge the impact of the problem. It may be that the current modification of the parameters does not take effect, and the modification needs to take effect again. This is the first case.

Restart the mount status of the database

SQL > alter system set cluster_database=true scope=spfile

Modify the database to open state, and then try whether the backup node can start successfully, and it is found that the startup is successful.

After modifying the RAC environment parameters, it is best to reconfirm whether the modification is successful.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.