oracle制造模块怎么跑mrp,MRP:运行MRP过程中相关程式功能与原理

来自oracle metalink,通过它来了解MRP运行一些功能与原理,尤其MRP执行Snapshots前后过程。

有时系统在运行此处被卡住。[@more@]

ECHNICAL OVERVIEW

I. Standard Planning Engine.

1. Exploder.

- Explodes through BOMs and ECOs.

- Determines low level codes.

- Identifies end assemblies.

2. Snapshot Workers.

- Snapshots item and transaction system data.

- Deletes snapshot data from previous planning run.

3. Planner, Planner Workers.

- Performs gross-to-net explosion.

- Deletes planning data from previous planning run.

- All processes are in the critical path.

- Each process "deletes, generates and loads".

II. Memory-Based Planning Engine.

1. The Exploder process is eliminated. These tasks are handled within the

snapshot.

2. Other than item information, the DELETE and LOAD tasks are removed

from the critical path.

3. DELETE and LOAD tasks are handled by separate processes.

4. MBP Concurrent processes.

- MRCNSP: Memory-Based Snapshot

- MRCMON: Snapshot Monitor

- MRCNSW: Memory Based Snapshot Worker

- MRCSDW: Snapshot Delete Worker

- MRCSLD: Loader Worker

- MRCNEW: Memory Based Planner

- MRCPDW: Planner Delete Worker

a. Memory-Based Snapshot.

- Writes a file containing the list of planned items. No

other snapshot task can begin until this file is written.

- Performs pre-processing tasks.

- Launches the Snapshot Monitor.

- Performs MB Snapshot Worker tasks after writing the file

containing the list of items.

b. Snapshot Monitor.

- Launches and controls remaining snapshot processes.

- Maintains list of snapshot and delete tasks.

- Uses inter-process communication to communicate with spawned

worker processes (database pipe).

c. Memory Based Snapshot Worker.

- Spawned by the Snapshot Monitor.

- Writes snapshot data to flat files.

- Snapshot Monitor assigns snapshot tasks to workers until all

tasks are complete.

- MRP:Snapshot Workers:

- If zero: none of these workers is spawned.

- If one or greater: corresponding number of workers is spawned.

d. Snapshot Delete Worker.

- Spawned by the Snapshot Monitor.

- Deletes snapshot and planning data from previous planning run.

- Number of workers spawned is determined by the site-level profile

MRP:Snapshot Workers. If this profile is set to zero, this

process runs single-threaded.

e. Loader Worker.

- Spawned by the Snapshot Monitor and MBP.

- Invokes SQL*Loader to load data written to flat files into

the database.

- One loader worker spawned per database table to be loaded.

- MRP:Use Direct Load Option:

- Yes: uses SQL*Loader direct path load.

- No : uses SQL*Loader conventional path load.

- Direct path load is faster, but requires exclusive locks

on the table (and its indexes) being loaded.

f. Memory Based Planner.

- Spawned by the Snapshot Monitor.

- Loads flat files written by snapshot workers into memory.

- Performs gross-to-net explosion.

- Is a single-threaded process.

- May launch Loader Workers if number of records to load exceeds

10,000.

g. Planner Delete Worker.

- Spawned by the MB Planner only when run without the snapshot.

- Planner Delete Worker is a single-threaded process.

- Deletes planning data from the previous planning run.

  • 0
    点赞
  • 2
    收藏
    觉得还不错? 一键收藏
  • 0
    评论
以下是 Oracle ADG (Active Data Guard) 的运行图: ``` +----------------------+ +-----------------------+ | Primary Database | | Standby Database (ADG) | |----------------------| |-----------------------| | LGWR ARCH | | LGWR ARCH | | | | | | | | | | v v | | v v | | SYNC ASYNC | | ASYNC SYNC | | | | | | | | | | v v | | v v | | MRP LNS |<------------| RFS MRP | +----------------------+ +-----------------------+ ``` 在 Oracle ADG ,主数据库 (Primary Database) 和备库 (Standby Database) 之间的数据同步可以通过 SYNC 或 ASYNC 两种方式进行。SYNC 方式下,主库的 LGWR 进程将 redo 日志传输到备库的 LNS 进程,备库的 LNS 进程将 redo 日志应用到备库的数据文件。ASYNC 方式下,主库的 LGWR 进程将 redo 日志传输到备库的 RFS 进程,备库的 RFS 进程将 redo 日志传输到备库的 MRP 进程,MRP 进程将 redo 日志应用到备库的数据文件。ARCH 进程负责将 redo 日志归档到归档日志文件,以供备份和恢复使用。 以下是这些进程的作用: - LGWR (Log Writer) 进程:LGWR 进程负责将 redo 日志写入磁盘。在 ADG ,LGWR 进程还负责将 redo 日志传输到备库的 LNS 进程或 RFS 进程。 - ARCH (Archiver) 进程:ARCH 进程负责将 redo 日志归档到归档日志文件,以供备份和恢复使用。 - MRP (Managed Recovery Process) 进程:MRP 进程在备库负责将主库的 redo 日志应用到备库的数据文件,以保持备库与主库的一致性。 - RFS (Remote File Server) 进程:RFS 进程在备库负责与主库建立连接,接收来自主库的 redo 日志,并将其传输到备库的 MRP 进程进行应用。 - LNS (Log Network Services) 进程:LNS 进程在备库负责接收来自主库的 redo 日志,并将其应用到备库的数据文件,以保持备库与主库的一致性。在 SYNC 方式下,LNS 进程直接从主库的 LGWR 进程接收 redo 日志;在 ASYNC 方式下,LNS 进程从 RFS 进程接收 redo 日志。

“相关推荐”对你有帮助么?

  • 非常没帮助
  • 没帮助
  • 一般
  • 有帮助
  • 非常有帮助
提交
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值