Home PC Games Linux Windows Database Network Programming Server Mobile  
  Home \ Database \ Oracle background processes daemons     - Analysis JavaBean (Programming)

- tar decompression problems gzip: stdin: not in gzip format (Linux)

- Two network security scanning tools under ubuntu (Linux)

- GlusterFS distributed storage deployment (Server)

- Boot-Repair Tool - repair of frequent start-up problems (Linux)

- CentOS7 install MySQL 5.5 (Database)

- RedHat Linux 6.4 install Oracle 10g error (Database)

- Java Foundation - Getting Start (Programming)

- Construction Spark source and application development environment (Server)

- Install the Solaris 10 operating system environment over the network to sparc (Linux)

- Linux System Getting Started Learning: The Linux anacron command (Linux)

- C # Future: Method Contract (Programming)

- Docker container plaintext password problem-solving way (Server)

- Java String type time compare the size (Programming)

- Summary of Docker mounted directory (Server)

- Install Kali Linux via PXE network (Linux)

- How to make a U disk to install Ubuntu (Linux)

- Oracle 12C truncate table cascade (Database)

- MongoDB fragmentation (Cluster) (Database)

- Use the top command (Linux)

  Oracle background processes daemons
  Add Date : 2018-11-21      
  Oracle Process Type:
For the process of the database, the database instance contains or interact.
In accordance with the use of an object into the client process (client) ORACLE process, ORACLE process has divided daemon (background), service process (server) and from the process (slave). client server process execution request process based.
For example: server process to resolve SQL queries, these queries stored in the shared pool, create and execute the query plan for each query. And reads data from the buffer cache or disk.
Process structure depends on the operating system and ORACLE database selectable modes.
Services (server) process can be divided into exclusive (dedicated) and shared (shared) services
Process Structure exclusive mode

Through V $ PROCESS view to view information about the process:
col addr for a16
COL SPID FORMAT a8 - system processes
COL STID FORMAT a8 - system threads
col usrname for a20
col program for a25
select addr, pid, spid, stid, usrname, program from v $ process order by stid;

Client (client) process:
ps -ef | grep -e sample -e sqlplus | grep -v grep

The results in the client and server with the following command to view a different client displays only sqlplus process. The server displays a non-local connections.
On session connection:
When the statement is executed using ORACLE statistical information, it will automatically create a scene autotrace collection
SQL> select sid, serial #, paddr from v $ session where username = user;
---------- ---------- ----------------
        96 3 00000001F051C858
SQL> set autotrace on statistics; --- tracking SQL statement execution plan to rebuild an additional scene
SQL> select sid, serial #, paddr from v $ session where username = user;
---------- ---------- ----------------
        96 3 00000001F051C858
        98 5 00000001F051C858
SQL> SELECT PROGRAM FROM V $ PROCESS WHERE ADDR = hextoraw ( '00000001F051C858');
-------------------------------------------------- -------------------
oracle @ 021Y-SH-BKAP (TNS V1-V3)
Background (background) process

Background process can be divided into compulsory (mandatory) process, optional (Optional) process and from (Slave) Process
Compulsory (mandatory) the process comprising:
PMON - monitor other background processes, or when the service scheduling process exception terminated. PMON is used to clean up and release the data buffer cache client processes and resources used. Such as: PMON reset the transaction status table, the process releases the lock and remove idle idle ID.
LREG Examples of dynamic registration listening processes. When an instance starts, LREG process determine the listener is running. The process was introduced in the 12c.

SMON system system monitoring process is responsible for more than one level of clean-up work in the capital include the following:
1, perform instance recovery.
2, recovery due to file read or tablespace offline errors terminated transaction.
3, clean up unused temporary segment.
4, through the merger dictionary management table space continuous free range.
SMON will periodically check whether you need to use, will be called when other processes need SMON.

DBW database writer process, modify the Buffer cache write the contents of the data file. Usually a DBW system write process to meet demand, but also can open more than one to improve performance. Ranging from DBW0-9, DBWa-z, DBW36-99
DBW the number of processes to modify the parameters of db_writer_processes, this parameter is dynamic, effective after reboot.
1 show parameter db_writer_processes

The following two conditions trigger DBW process writes:
1, when the service process scanning buffer pool, can not be found when the available buffers in the threshold, triggering the DBW dirty buffers are written to disk asynchronously.
2, DBW regularly write buffers and trigger the checkpoint, and the starting point of the checkpoint in the redo thread is the strength of recovery. Recorded checkpoint location depends on the first dirty buffer.
In most cases, DBW blocks are written to disk hash, this speed will be slower than sequential writes by LGWR, DBW execution
Multiple concurrent writing to improve efficiency. Number of blocks written by the operating system to decide.

LGWR The process used to manage the online redo log buffer. LGWR will buffer the part of the record to the redo log. Changes in the operation of the separation buffer. The dispersion of dirty buffer is written to disk, and in rapid succession will redo is written to disk.
Following cases, LGWR after the last write operation of all the changes over the write buffer
1, the user submits the transaction (commit command).
2, online redo log switch occurs.
3, from the last LGWR write after over three seconds.
4, has been in the redo log buffer cache data of 1/3 or 1M.
5, DBW buffer changes must be written out to disk.
Before DBW write dirty data, the database must redo records associated with changes in buffer is written out to disk (write-ahead protocol), if they do not write redo records, DBW will send a signal to the LGWR process, and wait for the LGWR process Upon completion of the written record, and then dirty data written to disk.
CKPT The process is primarily used to update the checkpoint information control file and data file header, and triggers DBW the write operation. checkpoint information contains checkpoint counting points, SCN recovery and positioning the starting point and the like.

MMON / MMNL and AWR (Automatic Workload Repository) related tasks. For example, when the calculation result exceeds the threshold MMON recording state and create a snapshot of the state, the most recent statistics crawl SQL object changes. MMNL will SGA buffer of ASH (Active Session History) state written to disk when the ASH buffer is full write.

In a distributed database, RECO (recoverer process) automatically distributed transaction processing errors, a node of a distributed transaction RECO automatically connect uncertain other libraries. When RECO reconnect, automatically resolve all doubt transactions, removing a library each pending transaction table with these transaction-related uncertainty row.
Optional (Optional) Process:
ARCn occurs when the redo log switch, the archive process copies the online redo log to the line storage. The process can also collect transaction redo data and transmitted to the standby database application. ARCn process and enable automatic archiving is turned on only when the database archiving mode.
CJQ0 and Jnnn
Run job queue processes, Oracle dynamic queue management process, thus allowing the work queue clients to use multiple job queue processes the event of a request. When these processes are idle, the database using the new process to release resources.
Dynamic job queue processes can run more than one job at the same time according to the given interval. The sequence of events is as follows:
1, CJQ0 Database scheduled task when necessary, the coordination process automatic startup and shutdown, the coordination process regularly choose to perform, select a new working time according to the sort of work from JOB $ table.
2, the coordination process dynamically generated from the work queue process (Jnnn) to perform the work.
3, the process executes a work queue is CJQ0 selected work, every job queue process executes only one job until the work is completed.
4, after the process is complete a single job, the next job will be, if not an executable program, the process is sleeping.
To specify the maximum work queue processes through parameter job_queue_processes parameters.
- HomeKit User Interface Guidelines (Linux)
- Security Configuration SQL Server 2000 database tutorial (Linux)
- MySQL 5.7 can not log in problem (Database)
- Linux network security strategy (Linux)
- Ubuntu How to install and upgrade Linux Kernel 3.15 (Linux)
- Spring classic face questions Share (Programming)
- Lua non-blocking write log (Programming)
- Help you enhance Python programming languages 27 (Programming)
- VMware Workstation virtual machine startup error: Could not open / dev / vmmon in CentOS 6 (Linux)
- OpenJDK7 source compiler installation on CentOS 6.5 (Linux)
- Linux system started to learn: the Linux syslog (Linux)
- OpenvSwitch 2.1.2 shell script to start and stop (Linux)
- MySQL main and backup replication structures (using mysqld_multi) (Database)
- JSON Introduction and Usage Summary (Programming)
- ASP.NET 5 is connected with the Redis server on the Linux platform (Server)
- Two alert log ORA Errors (Database)
- ORA-12545: Connection failed because the target host or object does not exist (Database)
- Linux awk text analysis tool (Linux)
- Linux productivity tools and tips (Linux)
- Upgrading KDE Plasma 5.3 in Ubuntu 15.04 (Linux)
  CopyRight 2002-2020 newfreesoft.com, All Rights Reserved.