oracle clusterware 10g/11g 进程详解
The Oracle 10g Clusterware processes on UNIX-based systems are:
Oracle Clusterware
Oracle-provided clusterware that manages clusterdatabase processing including node membership, group services, global resourcemanagement, and high availability functions.
CRSD- OracleClusterware Daemon
The primary Oracle Clusterware process
crsd:Performs highavailability recovery and management operations such as maintaining the OCR andmanaging application resources. This process runs as the root user, or by auser in the admin group on Mac OSX-based systems. This process restarts automatically upon failure.
The CRSD process that creates the backupsalso creates and retains an OCR backup for each full
day and at the end of each week .
crsd :执行高可用性的恢复和管理操作:维护ocr，管理应用资源，该进程以root用户身份运行。该进程在由于故障意外断后会自动重启。Crsd会在每天或者每周结束时对ocr进行备份。
The CRSD process maintains an internal counter to track how often a resourcehas been restarted.
Crsd进程维护内部的一个计数器来跟踪特定资源的重启次数：
- Engine for HAoperation
高可用性操作的引擎
- Manages ‘application resources‘
管理应用资源
- Starts, stops, and fails ‘application resources‘ over to other nodes
启动，停止，切换应用资源到其他节点
- Spawns separate ‘actions‘ to start/stop/check application resources
产生独立的操作来启动、停止、检查应用资源
- Maintains configuration profiles in the OCR (Oracle Configuration Repository)
维护ocr中的配置文件
- Stores current known state in the OCR.
把当前的状态信息存储在ocr中
- Runs as root
以root身份运行
- Is restarted automatically on failure
故障发生后能够自动重启
evmd: Event manager daemon. This process also starts the racgevt processto manage FAN server callouts.
Evmd：事件管理守护进程。这个进程也会启动racgevt进程来管理
Fan进程的调用
- Generates events when things happen
当意外发生时产生事件
- Spawns a permanent child evmlogger
产生一个永久子进程--evmlogger
- Evmlogger, on demand, spawns children
Evmlogger会根据实际情况自动产生子进程
- Scans callout directory and invokes callouts.
- Runs as Oracle.
以oracle用户身份运行
- Restarted automatically on failure
异常后能够自动重启
ocssd:Manages cluster node membership and runs as the oracle user; failureof this process results in cluster restart.
Ocssd：管理集群节点成员资格，以oracle用户身份运行；该进程在异常时会自动重启
- OCSSD isrequired with RAC databases and with non-RAC database when using ASM
使用 rac数据库，或者单实例使用asm的数据库都需要ocssd
- Provides access to node membership
提供对节点成员的访问
- Provides group services
提供组服务
- Provides basic cluster locking
提供基本的集群锁
- Integrates with existing vendor clusteware, when present
当有第3方集群软件时能和他自动整合
- Can also runs without integration to vendor clustware
即使没有第3方集群软件，也能够运行
- Runs as Oracle.
oracle用户运行
- Failure exit causes machine reboot(RAC cluster only).
该进程异常会导致机器重启
--- This is a feature to prevent data corruption in event of a split brain.
这个特性用来在发生脑裂时保护数据免收破坏。
oprocd : Process monitor for the cluster. Note that this process onlyappears on platforms that do not use vendor clusterware with OracleClusterware.
Oprocd：为集群提供进程监控。这个进程只有当没有使用第3方集群时才会出现
Oracle 11g clusterwareprocesses on UNIX-based systems are:
The Oracle Clusterware Stack
Oracle Clusterwareconsists of two separate stacks: an upper stack anchored by the ClusterReady Services (CRS) daemon (crsd) and a lower stack anchored by the OracleHigh Availability Services daemon (ohasd). These two stacks have severalprocesses that facilitate cluster operations.
Oracle集群件由两个独立的堆栈组成：上层堆栈由crs固定，下层堆栈由oracle高可用服务固定ohasd。
. The following sections describe thesestacks in more detail:
■ The Cluster Ready Services Stack
■ The Oracle High Availability Services Stack
The Cluster Ready Services Stack
The list in this section describes theprocesses that comprise CRS. The list includes components that are processes onLinux and UNIX operating systems, or services on Windows.
■ Cluster Ready Services (CRS): The primary program for managing highavailability operations in a cluster . The CRS daemon (crsd) manages clusterresources based on the configuration information that is stored in OCR for eachresource. This includes start, stop, monitor, and failover operations. The crsdprocess generates events when the status of a resource changes. When you haveOracle RAC installed, the crsd process monitors the Oracle database instance,listener, and so on, and automatically restarts these components when a failureoccurs.
Crs：集群中管理高可用性操作的主要程序。Crs守护进程根据ocr中存储的每个资源的配置信息来管理集群资源。相应的操作包括：start、stop、monitior、failover。当资源的状态变化时crsd进程就会测试相应的事件。当安装完RAC后， crsd 进程会监控oracle数据库实例，监听，以及当这些组件异常时就会自动重启他们。
■ Cluster Synchronization Services (CSS): Manages the clusterconfiguration by controlling which nodes are members of the cluster and by notifyingmembers when a node joins or leaves the cluster. If you are using certifiedthird-party clusterware, then CSS processes interface with your clusterware tomanage node membership information.Thecssdagent process monitors the cluster and provides I/O fencing. This serviceformerly was provided by Oracle Process Monitor Daemon (oprocd), also Note: TheSCAN must resolve to at least one address on the public network. For highavailability and scalability, Oracle recommends that you configure the SCAN toresolve to three addresses .See Also: Your platform-specific Oracle GridInfrastructure Installation
Guide installation documentation forinformation about system requirements and configuring network addresses Overviewof Oracle Clusterware Platform-Specific Software Components1-8 OracleClusterware Administration and Deployment Guide known as OraFenceService onWindows. A cssdagent failure may resultin Oracle Clusterware restarting the node.
Css：通过控制集群成员资格和在节点加入或离开集群时通知集群成员来管理集群配置。如果使用第3方认证集群软件，则css进程会和第3方集群件进行交互，来管理节点成员资格。
Cssdagent进程监控集群和提供io隔离。在oracle10g中cssdagen的工作由oprocd来完成。
Scan在公网必须能解析出最少一个ip地址，从高可用性和扩展性来考虑oracle建议配置一个可以解析出3个ip的scan。
在windows环境下对应的服务为：oraFenceService
如果cssdagent异常肯会导致oracle集群重启节点。
■ Oracle ASM: Provides disk management for Oracle Clusterware and OracleDatabase.
■ Cluster Time Synchronization Service (CTSS): Provides timemanagement in a cluster for Oracle Clusterware.
Ctss：为oracle集群提供时间管理
■ Event Management (EVM): A background process that publishes eventsthat Oracle Clusterware creates.
Evm：用来发布oracle集群创建的events的后台进程。
■ Oracle Notification Service (ONS): A publish and subscribe servicefor communicating Fast Application Notification (FAN) events.
Ons:为通知fan events提供发布和订阅的服务
■ Oracle Agent (oraagent): Extends clusterware to supportOracle-specific requirements and complex resources。 Thisprocess runs server callout scripts when FAN events occur. This process wasknown as RACG in Oracle Clusterware 11grelease 1 (11.1).
Oraagent：扩展集群以支持oracle特定的需求，和复杂资源。当FAN事件发生时该进程会运行服务器调用脚本。
■ Oracle Root Agent (orarootagent): A specialized oraagent processthat helps crsd manage resources owned by root, such as the network, and theGrid virtual IP address. The Cluster Synchronization Service (CSS), EventManagement (EVM), and Oracle Notification Services (ONS) components communicatewith other cluster component layers on other nodes in the same cluster databaseenvironment. These components are also the main communication links betweenOracle Database, applications, and the Oracle Clusterware high availabilitycomponents. In addition, these background processes monitor and manage databaseoperations.
Orarootagent：一个特别的oraagent，用来帮助crsd管理root用户的资源：网络，gridvip。
Css，evm，ons会和其他节点上的集群组件进行通信。这些组件也是oracle数据库，应用，oracle高可用性组件间的主要通信链路。除此之外，这些后台进程也监控，管理数据库的操作。
The Oracle High Availability Services Stack
This section describes the processes thatcomprise the Oracle High Availability Services stack. The list includescomponents that are processes on Linux and UNIX operating systems, or services onWindows.
■ Cluster Logger Service (ologgerd): Receives information from allthe nodes in the cluster and persists in a CHM repository-based database. Thisservice runs on only two nodes in a cluster.
Ologgerd：从集群的所有节点接受信息并将信息存放到CHM信息库中。该服务运行在集群中的所有节点。
■ System Monitor Service (osysmond): The monitoring and operatingsystem metric collection service that sends the data to the cluster loggerservice. This service runs on every node in a cluster.
Osysmond：监控和os度量收集，并将数据发送到集群日志服务。这个服务运行在集群中的每个节点。
■ Grid Plug and Play (GPNPD): Provides access to the Grid Plug andPlay profile, and coordinates updates to the profile among the nodes of thecluster to ensure that all of the nodes have the most recent profile.
Gpnpd：通过访问gpnp配置文件，以及协调更新集群中所有节点的配置文件，确保所有节点拥有最近的配置文件。
■ Grid Interprocess Communication (GIPC): A support daemon thatenables Redundant Interconnect Usage.
Gipc：确保内联冗余的使用的守护进程。
■ Multicast Domain Name Service (mDNS): Used by Grid Plug and Play tolocate profiles in the cluster, as well as by GNS to perform name resolution.The mDNS process is a background process on Linux and UNIX and on Windows.
Mdns：被gpnp用来定位集群中的配置文件；gns用他来作命名解析。
■ Oracle Grid Naming Service (GNS): Handles requests sent by externalDNS servers, performing name resolution for names defined by the cluster.
Gns：处理外部dns服务器的请求，为集群定义的名字进行命名解析
Table 1–1 lists the processes and servicesassociated with Oracle Clusterware components. In Table 1–1, if a UNIX or aLinux system process has an (r) beside it, then the process runs as the rootuser.