Common AIX Commands and parameter adjustments

Adjust netstat in UNIX using common AIX Commands and parameters to determine the network cache zone's urgent use, for example, the overflow information about queue in netstat. This information indicates the status of TCP and UDP queues. AIX provides the no command to set kernel configurations for the network. Common commands and parameters on www.2cto... AIX can be used to adjust netstat in UNIX to determine the network cache for urgent use, such as overflow information about queue in netstat. This information indicates the status of TCP and UDP queues. AIX provides the no command to set kernel configurations for the network. When www.2cto.com starts a process, VMM needs to allocate virtual memory space for the process to store the data and structure of the process. At this time, VMM maintains a free List of unallocated virtual memory page frames, which is used to handle process page defects in the system. In AIX, the VMM parameter settings can be obtained through the vmo command # vmo-aVMM divides the memory segments into two categories: computing segments (work segments) and File segments (persistent segments ). File segments use file memory, while computing segments use computational memory. For databases, shared memory is applied for when the database is started. the memory structure is the computing memory. For file access, because the file is stored in the corresponding storage location on the disk, it will be stored in the file segment memory until the page conversion or process disappears. Even if a swap operation occurs, we still want to distribute the file segments to the disk, instead of the computing segment memory structure. The page frame linked list used by the big data cache will be very large. To use AIX to allocate data caches to the application system more effectively and complete the effective height, you can set a large memory page structure to improve performance. By default, AIX does not allocate any memory to physical memory on large pages, but can be manually specified. it can support 4 kB, 64 KB, 16 MB, and 16 GB. Vmo-r-o lgpg_size = 16777216-o lgpg_regions = 256 the page pool with 16 MB of large memory is allocated and 256 are specified. If you want to use a large page for the shared memory, you also need to set the v_pinshm parameter of AIX to 1 (the default value is 0) so that the VMM can fix pages in the memory pool to avoid page switching. # Vmo-r-o maxpin % = 50 # After vmo-r-o v_pinshm = 1 is set, the application system and database can use a large memory pool. In the Oracle system, you can set the LOCK_SGA parameter to lock the Oracle shared memory in the memory. at the same time, when allocating shared memory, the request is larger. Minperm and maxperm must be set. These two parameters are the thresholds in the VMM memory replacement policy. When the actual percentage of memory occupied by file pages is lower than minperm, the page replacement algorithm replaces both file and computing pages, ignoring the re-distribution calculation. When the actual memory usage of a file page is higher than that of maxperm, the page replacement algorithm only replaces the file page. Therefore, these two parameters can be used to determine whether the AIX system tends to use computational memory or file memory. Set the number of idle lists minfree and maxfree parameters in AIX. when the number of pages in the aix vmm idle list falls below the value specified by minfree, VMM starts to run the lru daemon, convert the pages that have not been used recently. After the idle table is increased to the data entry specified by maxfree, the page conversion operation is stopped. The default values are 960 and 1088 (4 KB page ). For a 64-bit CPU server, it can be set to 8000 (32), 8512 (23 MB ). # Vmo-p-o maxfree = 8512 # vmo-p-o minfree = 8000 # mount-o cio/db/etc/filesystemsoptions = cio, when rw accesses a bare-device database, AIX uses the asynchronous I/O method to access the bare-device by default. at this time, the AIO process is a core process, no CPU context switch is required. When Oracle accesses the data files in the file system, it needs to call the AIO Server process to switch the CPU up and down. There is still a slight difference in the use of CIOs or bare devices. Alter system set filesystemio_options = setall scope = spfile; [aix202: root:/] vmstat 2 30 System configuration: lcpu = 4 mem = 3648 MB kthr memory page faults cpu ----- ----------- ------------------------ ------------ ----------- r B avm fre re pi po sr cy in sy cs us sy id wa 2 0 274445 593091 0 0 0 0 0 0 0 7 403425 8126 16 49 0 0 274445 0 0 0 0 0 0 0 0 8 593090 404238 16 16 48 0 0 8182 274445 0 0 0 0 0 0 8 406472 8182 16 16 0 1 0 274445 0 0 0 0 0 0 0 0 8 593088 406185 16 16 0 1 0 0 8182 0 0 0 0 0 0 0 0 0 8 406494 8193 35 16 48 0 1 0 274576 0 0 0 0 0 0 0 46 592951 406253 35 17 47 1 2 0 8184 0 0 0 0 0 0 0 9 274576 592950 35 16 48 0 2 0 274575 592950 0 0 0 0 0 0 10 407277 8178 35 16 48 0 2 0 274575 0 0 0 0 0 9 592949 404805 35 16 49 0 2 0 8172 0 0 0 0 0 0 0 1 1 406614 8134 35 16 48 0 2 0 274575 0 0 0 0 0 0 0 0 5 592948 405990 35 16 49 0 0 8127 0 0 0 0 0 0 6 274575 35 16 48 0 2 0 274574 592949 0 0 0 0 0 0 6 408020 8165 35 16 49 0 1 0 274574 0 0 0 0 0 8 592949 406316 35 16 48 0 2 0 8142 0 0 0 0 0 0 0 0 5 406821 8127 16 49 0 1 0 274574 0 0 0 0 0 0 4 592949 404438 35 16 49 0 re column: page height program input/output list fr column: released page (replacement page) sr column: scanned by page replacement algorithm Page cy column: the clock cycle Faults used by the page replacement algorithm, used to calculate the average value of traps and interrupt rates per second. In column: Device interruption statistics. Sy column: System call statistics. Cs column: kernel thread context switch statistics. [Aix202: root:/] iostat 5 5 System configuration: lcpu = 4 drives = 16 paths = 1 vdisks = 0 tty: tin tout avg-cpu: % user % sys % idle % iowait 0.0 11.8 36.1 15.1 48.8 Disks: % tm_act Kbps tps Kb_read limit 0.0 0.0 0.0 0.0 0 0dac1 0.0 0.0 0.0 0 0dac1utm 0.0 0.0 0.0 0 0dac0 0.0 0.0 0.0 0 0dac0utm 0.0 0.0 0.0 0.0 0 0hdisk1 0.0 0.0 0.0 0 0hdisk2 0.0 0 0hdisk3 0.0 0.0 0.0 0 0hdisk4 0.0 0.0 0.0 0 0hd Isk5 0.0 0.0 0.0 0 0hdisk6 0.0 0.0 0.0 0 0hdisk7 0.0 0.0 0.0 0 0hdisk8 0.0 0.0 0.0 0 0hdisk9 0.0 0.0 0.0 0 0hdisk10 0.0 0.0 0 0loop0 0.0 0.0 0.0 0 0tty partial statistics per second total number of messages received from terminals on the system, and the total number of characters per second that are output to the system for all summer corn. Tin, the total number of characters from all tty tout, the total number of characters written to all tty CPU statistics the global average statistical value of all processors, will be regularly updated by the kernel tty: tin tout avg-cpu: % user % sys % idle % iowait 0.0 599.8 39.5 19.5 20.4 Disks: % tm_act Kbps tps Kb_read limit 0.0 0.0 0.0 0.0 0 0dac1 12451.7 3112.9 2812 0.0 22496dac1utm 0.0 0.0 0.0 0 0dac0 0.0 0.0 0.0 0 0dac0utm 0.0 0.0 0.0 0.0 0 0hdisk1 0.0 0.0 0 0hdisk2 0.0 0.0 0 0hdisk3 0.0 0.0 0.0 0 0hdisk4 0.0 0.0 0.0 0 0hd Isk5 0.0 0.0 0.0 0 0hdisk6 0.0 0.0 0.0 0 0hdisk7 0.0 0.0 0.0 0 0hdisk8 0.0 0.0 0.0 0 0hdisk9 97.5 12451.7 3112.9 2812 22496hdisk10 0.0 0.0 0 0loop0 0.0 0.0 0 0.0 iowait, percentage of active physical disks. Kbps: the amount of data read and write to the disk, expressed in Kb/s. Kb_read: the number of KB read.