"Oracle" 10gRAC noun explanation _ notes

Oracle Cluster Ready Service (Oracle Cluster Readiness Services)

Amnesia: Amnesia, configuration information in multiple nodes is not uniform, with the old configuration file, resulting in configuration loss.

Use OCR to resolve in RAC.

Split Brain: Brain fissure, RAC uses voting algorithms and Quorun device to solve.

IO Fencing:io isolation, when a cluster is cracked, it is necessary to determine which node should gain control of the cluster and which is evicted from the cluster.

Also make sure that the evicted node cannot manipulate the shared data. This requires an IO isolation solution.

--clusterware after installation, query voting disk location

$crsctl Query CSS Votedisk

------Oracle Clusterware Component---------Process

Event Manager Service EVMD

Evmdlogger

Oracle Notification Service (ONS)ONS

Process Monitor DaemonOPROCD

RACG Racgmain

Racgimon

----

The OCSSD process is a clusterware critical process that causes the system to restart unexpectedly.

Css:cluster Synchronization SERVICE,CSS Service provides real-time monitoring of cluster health status through a variety of heartbeat mechanisms, providing basic cluster service functions such as brain fissure protection.

The CSS service has two heartbeat mechanisms:

1. Network heartbeat via private networks, Max delay: MC (miscount), per second

2. Voting disk heartbeat, Max latency: IOT (I/O Timeout), per second

--------

-All components requiring high availability are registered to OCR in the form of CRS resource when installing the configuration.

CRSD: The process depends on the content in OCR to monitor those processes, such as monitoring, and how to resolve problems when they occur.

The CRSD process is responsible for monitoring the operating state of the CRS resource and starting to stop monitoring failover these resources.

CRS Resource:

GSD (Global service daemon)

ONS (Oracle nofitication service daemon)

Vip

Database

Instance

Service

These resources fall into two categories:

Noteapps class: GSD ONS VIP listener. Understanding: Each node needs only one enough

dfatabase-related Resource class: database instance service. Understanding: Not restricted by nodes.

The GSD, ONS, and VIPs are created and enlisted in OCR when the VIPCA is executed when the Oracle Clusterware is installed.

Database, listener, instance, and service are automatically or manually registered to OCR during their respective configuration

DLM (Distributed lock Management): Distributed lock manager. (data block is granular)

--DLM function to coordinate the competition of resources between instances to deal with multi-node concurrency problems

In DLM, resources are divided into two categories based on the amount of resource activity-intensive:

1.cache Fusion resource: specifically data blocks such resources, including: ordinary data block, index data block, Segment header block (segment header), Undo data Block

2.no-cache Fusion Resource: Non-block resources are all categorized as this class, including: Data file, control file, data dictionary view, library cache, row cache, etc.

----------------

GRD (Global Resource Directory): Records the distribution of each chunk in the cluster, in each instance of the SGA, but each instance of the SGA is part of the GRD, all instances of the GRD summary together is a complete GRD

GCS (global cache service)

GES (Global enqueue Service)

GCs and Ges manage status records for each data file and block per cache by using the Global Resource Directory (GRD)

GCS for GES of data blocks for locks, queue control

GRD content is distributed across all active instances, which effectively improves the SGA size of an Oracle RAC instance.

----process

1.lmsn-global Cache Service Process Global Cache Services processes: Manages data access requests between nodes and clusters

is the main process of cache fusion, responsible for the transfer of data blocks between instances, the corresponding service is called GCS (Global cache service)

This process name originates from the lock Manager service (9 starts this service renamed to global Cache Service[gcs])

Number of processes by parameter: Gcs_server_processes control, Value: 0~9

Gcs=>lmsn

2.LMD

The Global Enqueue Service (GES) is responsible for coordinating the order of access to data blocks across multiple instances to ensure consistent access to the data.

Together with the GCS service of the LMSN process and GRD, it is the core of the RAC feature cache fusion.

Ges=>lmd

3.LCK

This process is responsible for synchronizing access to Non-cache Fusion resources, with each instance having a lck process.

4.LMON Global Queue Service Monitor

Global Enqueue Service Monitor (LOCK monitor Processes)

Primary monitoring of global queues and global resources within the cluster, managing instances and handling exceptions, and corresponding cluster queues for recovery operations.

The Lmon process for each instance periodically communicates to check the health status of each node in the cluster, and when a node fails,

Responsible for the cluster reconfiguration, GRD recovery and other operations, it provides services called: Cluster Group Services (CGS).

Cgs=>lmon

5.crs-Cluster resource Service (cluster ready services)--CRSD

Basic procedures for managing highly available operations within a cluster

Any transaction managed by CRS is called a resource

Database, instance, listener, virtual IP, application process, etc.

CRS is based on resource configuration information stored in OCR to manage these resources

When the state of a resource changes, the CRS process generates an event

6.css-Cluster Synchronization Service (Cluster Synchronization Service)--CSSD

Managing Membership for cluster nodes

Control which node is the member of the cluster, the node notifies the cluster member when joining or leaving the cluster to control the cluster configuration information.

This process has failed to cause cluster restart

7.EVMD Incident Management Service (event Management)--EVMD

Event Management Daemon

Publish a background process for CRS creation events

8.ons-Event Publishing and subscription services (Oracle Notification service)

Rapid application of Communication notification event publishing and subscription services

9. Process Monitoring Daemon (OPROCD)-This process is locked in memory to monitor the cluster and provide I/O fencing.

---------------------------------------------------------

Lmon-lock Monitor Lock monitoring device

Lck0-instance Enqueue Process Instance queue processes

Lmd0-global Enqueue Service Daemon Global Queue Services Daemon

Lms0-global Cache Service Process global cache services processes

Diag-diagnostic Process Diagnostics

On Linux, Oracle cluster is configured in/etc/inittab. It uses three daemons to implement:

Crs-cluster Ready Service (CRSD) cluster readiness Services

Css-cluster Synchronization Service (OCSSD) cluster Synchronization Services

Evm-event Manager (EVMD) event Manager

-------

Online information, big talk Oraclerac, etc.

This article from "Despite the wrong, let me wrong to die!" "Blog, be sure to keep this provenance http://hxw168.blog.51cto.com/8718136/1567667

"Oracle" 10gRAC noun explanation _ notes