Shulou(Shulou.com)06/01 Report--

User process

Ps-elf | grep oracleorcl | grep-v grep

Database process

Ps-elf | grep oracleorcl

Daemon process

Ps-elf | grep oha

Large pool: in SGA, large pool is an optional buffer area, which can be configured by administrators as needed. Large pool is needed as cache space during large-scale input / output and backup, such as big data operation, database backup and recovery.

Java pool: the Java pool is an optional buffer, but the Java pool must be set when installing java or java programs, which is used to compile Java language writing instructions.

Five core processes:

DBWn process: used to write dirty data from buffer cache to a data file on your hard disk.

What is dirty data? for example, I execute the following statement

SQL > update emp setsal=sal+100

At this time, the modified data is stored in memory, and not saved to the data file on the hard disk, the data is cached in buffer cache, and the transaction has not been committed (commit), so we call it dirty data after modifying the data that has not been written into the hard disk data file. In fact, when you open a session at this time, use the SCOTT user connection to see that the data in the emp table is unmodified. So in oracle, this kind of operation is asynchronous write, multiple processes can write at the same time, multiple accounts will have multiple DBWn processes, in fact, the process is called DMW, followed by n is the resulting process ID

SQL > update scott.emp setsal=sal+100

14 rows have been updated.

SQL >

LGWR: there is only one log writing process

The lgwr process is triggered when the user commit (commit transaction)

The lgwr process is triggered when 1/3 of the red log buffer space is used up.

Before writing dirty data to disk, the LGWR process writes log information to the log file, performing such an action every 3 seconds.

There cannot be more than one LGWR process because the logs are written sequentially and cannot be written in parallel, otherwise data recovery cannot be done.

CKPT process: checking whether the headers of all data files and control files are consistent is a mechanism to ensure that all modified data blocks in the data buffer are written to the data file, that is, a mechanism to ensure data integrity. However, after the checkpoint is completed, the CKPT process is responsible for updating the header information of the data file and control file, and saving the checkpoint information to ensure the synchronization of the database log file and the data file. The principle is that when the database is restored, you only need to find the last checkpoint saved by CKPT, and then you can determine the starting location of the recovered data in the log file according to it, and then re-execute the log information.

SMON process: if the data file and the control file header are not consistent, the SMON process will be triggered. When the instance starts, the SMON process checks whether the synchronization information of all data files is consistent with the checkpoint information recorded in the control file, and starts the instance recovery operation if it is inconsistent. In fact, the CKPT process is for use by the SMON process.

PMON process: the process monitors the process, and the PMON process manages the user session, which takes the appropriate action if something goes wrong with a session. For example, if a user's personal computer restarts when logging in to the database, PMON can detect the problem and organize any operations that the user is doing.

If you have any questions about this article, please add Wechat below to discuss it.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.