Introduction to underlying processes of Oracle RAC

1. GSD global services daemon

Description of Oracle official documentation

The global services daemon (GSD) runs on each node with one GSD process per node. the GSD coordinates with the Cluster Manager to receive requests from clients such as the dbca, em, And the srvctl utility to execute administrative job tasks such as instance startup or shutdown. the GSD is not an oracle instance background process and is therefore not started with the Oracle instance.

GSD interacts with the RAC management tool dbca srvctl OEM to complete management tasks such as instance startup and shutdown.
To ensure that these management tools run normally, you must start GSD on all nodes, and a GSD process supports
Multiple RAC. GSD processes of a node are located in the $ ORACLE_HOME/bin directory, and the log file is $ ORACLE_HOME.
/Srvm/log/gsdaemon. Log
For example:
Assume that an OEM tool is used to start an instance. The OEM delivers the task to the corresponding Intelligent engine, which generates
A script file containing the srvctl command. The GSD process reads the script file and runs the script.
The execution result is returned to the intelligent engine, and the Intelligent engine returns to the OEM.
For example, if you use the srvctl tool to close all instances, the GSD process accepts requests from the srvctl tool and
Execute the request on the local node and return the execution result to the srvctl session.

2. lmon: Global enqueue Service Monitor
Description of Oracle official documentation
Global enqueue Service Monitor (lmon)

The background lmon process monitors the entire cluster to manage global resources. lmon manages instance deaths and the associated recovery for any failed instance. in particle, lmon handles the part of recovery associated with global resources. lmon-provided services are also known as cluster group services.

Lmon mainly monitors global queues and global resources in the cluster, manages instances, handles exceptions, and recovers corresponding cluster queues.

3. LMD: Global enqueue service daemon
Description of Oracle official documentation
Global enqueue service daemon (LMD)

The Resource Agent process that manages requests for resources to control access to blocks. The LMD process also handles Deadlock Detection and remote resource requests. Remote resource requests are requests originating from another instance.

The LMD process mainly manages access to global queues and resources, updates the status of corresponding queues, and Processes requests from other instances.
. The current status of each global queue is stored in the shared memory of the corresponding instance.
You have the right to use the resource. A special queue exists in the shared memory of an instance (master ).
Record the resource requests from other remote instances. When the LMD process of the remote instance sends a resource request, this request
Point to the LMD of the master instance. When the lmd process of the master instance receives this request, a special queue in the shared memory
To check whether the resource is invalid. If the resource is valid, the LMD process updates the status of the resource pair column and notifies the requested resource
The LMD process can be used by this resource queue. If the resource queue is being used by another instance or is currently invalid
The LMD process notifies the LMD process of an instance in use to release the resource. When the resource release becomes effective,
The lmd process of the master instance updates the status of the resource queue and notifies the LMD process of the requested resource instance of this resource.
The queue can be used. In addition, the LMD process is also responsible for the deadlock issue in the queue...

4. lmsn: Global cache service processes
Description of Oracle official documentation
Global cache service processes (lmsn)

Processes that manage remote messages. Oracle RAC provides for up to 10 Global cache service processes. The number of lmsn varies depending on the amount of messaging traffic among nodes in the cluster.

Global cache service process (N 0 ~ 9) the LMS process is mainly used to manage access to data blocks in the cluster and transmit blocks in the buffer cache of different instances.
Image. The LMS Process manages database requests across clusters and ensures that one of the requests is stored in the buffer cache of all instances.
Data block images can only appear once. The LMS process coordinates the access of data blocks by passing messages in the instance. When
When an instance requests a data block, the instance's LMD process sends a data block resource request.
The lmd process of the master data block instance.
When the process releases the resource, the LMS process of the instance that owns the resource will create a consistent read of the data block image.
Then pass the data block to the buffer cache of the Instance requesting the resource.
The LMS process ensures that only one instance can be allowed to update a data block at a time, and maintains
Image Record (including the status flag of the updated data block), RAC provides 10 LMS processes, the number of processes
The amount of data transmitted between nodes increases.

5. LCK
Description of Oracle official documentation

Instance enqueue Process
Processes are mainly used to manage resource requests between instances and cross-instance call operations. Call operations include data dictionary and other objects.
And handle caceh fusion's Chace resource requests (for example, Dictionary cache)

6. diag:

The diagnosability daemondiag process is mainly used to capture the diagnostic information of failed processes in the instance and generate the corresponding trace file (the trace
The file is saved in the backupground_dump_dest/cdmp_timestamp directory ).
Configuration is not recommended. This process is automatically started and does not need to be adjusted. If the process fails, it is automatically started.
Restart.

7. ONS Oracle Notification Services
Description of Oracle official documentation
A publish and subscribe service for communicating information about all fan events.

Responsible for communication between different nodes.

Reference http://download.oracle.com/docs/cd/B19306_01/rac.102/b14197/rac_glossary.htm#sthref1791

Search for more related topics: rac lms lmon GSD ons