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

This article mainly introduces "introduction to the usage of topas command in AIX". In daily operation, I believe many people have doubts about the usage of topas command in AIX. The editor has consulted all kinds of materials and sorted out simple and easy-to-use operation methods. I hope it will be helpful to answer the doubts about "introduction to the usage of topas command in AIX". Next, please follow the editor to study!

The topas command is updated every 2 seconds by default.

1. The topas command shows the performance of various indicators of the system in the form of regions, as shown in the following figure:

1. CPU: reflects the CPU performance area. If there are multiple CPU, press c twice to display the CPU list. Pressing c only once closes this area

The ratio of time occupied by User%: user processes to CPU. This number reflects the sum of the CPU time ratio occupied by all kinds of software that the user is running.

The ratio of CPU time occupied by the Kern%: operating system kernel. As the basic software, the operating system not only supports and serves the application, but also needs some CPU and memory resources to run. However, in general, the operating system does not take up a large percentage of CPU time. If the value of User%+Kern% is greater than 70% for a long time, there may be insufficient system resources, which needs our attention.

The ratio of Wait%:CPU waiting state to CPU time. CPU usually waits for a response from IO. If this value is high for a long time, it means that IO does not have time to respond to many IO requests. At this time, we can only find a way to optimize it from the IO level.

Idle%:CPU idle time ratio.

When the hot CPU list is first displayed in this area, sort the list by the User% field. However, you can sort the list by other fields, just move the cursor to the top of the desired column.

2, 2, Network: reflects the network usage area, press the n key to close this area. Press the n key again and the report summary of all network interface activities will be displayed.

KBPS: the number of KB data transfers per second is the sum of received and sent data per second.

I-Pack: the number of packets entered per second.

O-Pack: packets output per second.

KB-In: the number of bytes entered per second.

KB-Out: the number of bytes output per second.

3. 3. Disk: the area that reflects the disk utilization. Press d to close this area. Press d again and the report summary of all physical disk activity is displayed.

Disk: name of the physical disk

Busy%: indicates the percentage of time that the physical disk is active; when the Busy% of the disk is continuously greater than 85%, the disk is considered to be quite busy, which is worth paying attention to.

KBPS: the number of bytes read and written per second during monitoring (in K)

TPS: the amount of data transferred from the physical disk per second. A transmission refers to an Icano request.

KB-Read: K bytes read from the physical disk per second

KB-Write: the number of K bytes written to the physical disk per second.

4. FileSystem: the area that reflects file system usage:

KBPS: the number of bytes read and written per second during monitoring (in K)

TPS: the amount of data transferred by the file system per second.

KB-Read: K bytes read from the file system per second

KB-Write: the number of K bytes written to the file system per second.

5. The situation area that reflects the process information:

Name: the name of the process, that is, the name of the binary that starts when the process is executed.

PID: the ID of the process, the ID of the process is unique in the system, and it is an important value for us to know and track the process information.

Tracking process CPU usage, disk IO read and write, process memory and pagingspace usage, and so on.

The ratio of CPU time occupied by the CPU%: process.

PgSp: the amount of pagingspace space consumed by the process. In MB.

Owner: the owner of the process, that is, which user started the process.

Note: by default, this area lists the information of the first few processes that take up the highest CPU.

6. EVENTS/QUEUES reflects the system event / queue information area:

Cswitch: the number of content transitions per second during monitoring

Syscall: total number of system calls run per second during monitoring

Reads: the number of read system calls run per second during monitoring

Writes: the number of write system calls per second run during monitoring

Forks: the number of derived system calls that were run per second during monitoring

Execs: the number of calls to execute the system per second run during monitoring

Runqueue: average number of threads waiting for the processor to be idle to run

Waitqueue: average number of threads waiting for paging to complete

Readch: the number of bytes per second read by the read system call during monitoring

Writech: the number of bytes per second written by writing system calls during monitoring

Rawin: the number of naked bytes read through TTYs per second during monitoring

Ttyout: number of bytes written to TTYs per second during monitoring

Igets: the number of calls to the node lookup routine per second during monitoring

Namei: number of calls per second for path lookup routines during monitoring

Dirblk: the number of directory blocks scanned per second through the directory search routine during monitoring

7. PAGING reflects the paging space information area.

Faults: total number of page errors per second during monitoring

Steals: 4k frames of physical memory stolen by the virtual memory manager (VMM) per second during monitoring

PgspIn: the number of 4k pages read from page space per second during monitoring

PgspOut: the number of 4k pages written to the page space per second during monitoring

PageIn: the number of 4k pages read per second during monitoring, including page activity related to read operations from the file system

Subtracting the value of PgspOut from this value is the number of page activities associated with writing to the file system per second.

PageOut: the number of 4k pages written per second during monitoring, including page activity related to write operations to the file system

The value of PgspOut subtracted from this value is the number of paging activities associated with writing to the file system per second.

Sios: the number of requests obtained through the virtual memory manager during monitoring

Description: paging space, that is, the space on the disk, is used as memory space in the AIX operating system. It is just a temporary storage place for memory pages, which are those that have not been used for a long time. If a large number of paging appears, there will be trouble at this time, indicating that there is not enough memory! This area is mainly concerned with PageIn,PageOut if these two values are greater than three digits, and for a long time greater than this value, there may be memory bumps, that is, constantly changing memory pages to disk space, and then reading memory pages from disk space, the memory use efficiency of the system becomes very poor, and the response performance of the system becomes slow.

8. MEMORY reflects the information area of memory usage:

Real,MB: the total amount of memory actually owned by the operating system, in MB.

% Comp: computational memory footprint ratio.

% Noncomp: the ratio of non-computational memory usage.

% Client: also non-computing memory, Noncomp includes client memory, jfs file system uses noncomp: for distinction, jfs2 and nfs use Client.

Description: computational memory is the memory actually used by the process, including stack, variables in the process and other values. When a process terminates, its physical and page space is freed. In VMM, when there is less free physical memory (that is, when the sum of computational memory and non-computational memory is close to 100), recently unused programs can be moved from RAM to paging space to help free physical memory, thus getting more practical work done.

Non-computational memory is the use of non-computational memory when there are a large number of file operations, such as compression, database dump/load, and so on. According to the default configuration of the system, non-computing memory takes up up to 90% of the total memory (AIX5L is 80%). Because the non-computing memory footprint is not actively released, it is not released until the file is uninstalled, the page is replaced, or the link to the file is unlinked. This may result in a shortage of memory resources and high Paging Space utilization. At this time, we can reduce the share of non-computing memory by reducing the parameter values of minperm and maxperm.

9. PAGING SPACE: reflects the usage area of the page space

Size,MB: the size of all page space in the system, in M

% Used: currently used page space as a percentage of total space

% Free: the current free page space as a percentage of the total space.

Setting of page space size: today's minicomputers generally have a large memory. When the memory exceeds 4GB, it needs to be determined according to the actual situation. Generally, you can start 3GB, and then observe the usage of paging space. If the utilization rate exceeds 70%, you need to increase paging space.

2. Topas command parameters:

-d: specifies the number of disks to be monitored, and number_of_monitored_hot_disks specifies the number of disks to be monitored.

-h: displays help information.

-I: sets the interval of monitoring time (seconds). The default is 2 seconds.

-m: displays in black and white mode.

-n: sets the number of network interfaces to be monitored.

-p: specifies the number of processes to be monitored.

-w: indicates the value of the WLM class being monitored.

-c: indicates the number of CPU being monitored.

-f: displays file system information.

-@: Wpar partition name

-P: displays the entire full-screen process.

-U:username shows the user's progress on full screen, which is used with the-P parameter.

-D: display the disk condition in full screen.

-E: full-screen display of the network.

-C: show the intersecting LPAR in full screen

-V: full screen display disk volume group usage

-M: full-screen analysis of memory applications

-L: full screen display of logical partitions (LPAR).

-W: displays the full screen WLM (load Manager) class.

The maximum value of the number of the parameters is determined by the available space on the screen. If this value exceeds the number of items installed in the system, only the installed items are monitored, and only the installed items are monitored and realistic. If the number is omitted, only two corresponding projects (except the number of processes) are monitored. If the number is 0, the corresponding project information is not displayed. The above parameters have been tested and only work in the AIX5L version, not in AIX6.

3. Subcommands of topas:

A: displays all monitored (CPU,network,disk,WLM,process) conditions, and when this command is executed, the content displayed by the original topas command will be restored.

C: switch between the standard interface and the CPU subinterface.

D: switch between standard interface and disk interface.

N: switch between the standard interface and the network subinterface.

W: switch between switches of the WLM class.

F: switch between the standard interface and the file system subinterface.

P: switch between the standard interface and the process subinterface.

P: full-screen process display replaces the default display.

W: this key replaces the default display with a full-screen WLM class display

At this point, the study on "introduction to the usage of topas commands in AIX" is over. I hope to be able to solve your doubts. The collocation of theory and practice can better help you learn, go and try it! If you want to continue to learn more related knowledge, please continue to follow the website, the editor will continue to work hard to bring you more practical articles!

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.