Shulou(Shulou.com)06/02 Report--

Editor to share with you the method of apache to modify the maximum number of concurrent connections. I hope you will gain a lot after reading this article. Let's discuss it together.

Apache provides several different MPM modules for different operating systems, such as mpm_beos, mpm_event, mpm_netware, mpmt_os2, mpm_prefork, mpm_winnt, mpm_worker. If conditions permit, we can compile the specified MPM module into our own Apache according to the actual requirements (Apache's source code is open, allowing users to compile on their own). However, if we do not have a choice at compile time, Apache will choose the corresponding MPM module according to the following table according to different operating systems, which is also the MPM module recommended by Apache for different platforms.

Default MPM modules on different operating systems

Operating system MPM module description

There is no need to introduce Windowsmpm_winnt:)

There is no need to introduce Unix/Linuxmpm_prefork:)

The official version of BeOSmpm_beos, a multimedia operating system developed by Be, has been stopped updating.

Netwarempm_netware is a network operating system introduced by NOVELL.

OS/2mpmt_os2 an operating system originally developed jointly by Microsoft and IBM and now developed separately by IBM (Microsoft abandoned OS/2 to develop Windows)

The mpm_event module can be seen as a variant of the mpm_worker module, but it is experimental and generally not recommended.

Of course, Apache also provides a finished Apache that has been compiled for MPM modules according to different operating systems on its official website. You can click here to go to the official website of Apache to download.

In addition, if we want to know what kind of MPM module is used inside an Apache, we can enter the Apache installation directory\ bin as a command line, and then type the command httpd-l to see what kind of MPM module is currently used inside Apache.

Use the httpd-l command to view the compilation module

As BeOS, NetWare, OS/2 and other operating systems are not common in the usual development work, here we mainly focus on the MPM module on Windows and Unix/Linux operating systems. On Windows and Unix/Linux operating systems, there are three main MPM modules: mpm_winnt, mpm_prefork and mpm_worker.

Mpm_prefork module

The mpm_prefork module is mainly used in the Apache server of the Unix/Linux platform, and its main working mode is that when the Apache server is started, the mpm_prefork module will create multiple sub-processes in advance (the default is 5). After receiving the request from the client, the mpm_prefork module will transfer the request to the child process, and each child process can only be used to process a single request at the same time. If the current number of requests will exceed the number of pre-created child processes, the mpm_prefork module creates new child processes to handle additional requests. Apache always tries to keep some spare or idle subprocesses to meet upcoming requests. In this way, the client's request does not have to wait for the child process to be generated after it is received.

Because each request corresponds to a child process in the mpm_prefork module, it consumes more system resources than the other two modules. However, the advantage of the mpm_prefork module is that each of its child processes processes the corresponding individual request independently, so that if there is a problem with one of the requests, the other requests will not be affected. At the same time, the mpm_prefork module can be applied to third-party modules that do not have thread safety (such as non-thread safe versions of PHP), and it is easy to debug on platforms that do not support thread debugging. In addition, the mpm_prefork module has higher stability than the mpm_worker module.

Mpm_worker module

The mpm_worker module is also mainly used in the Apache server of the Unix/Linux platform, which can be regarded as an improved version of the mpm_prefork module. The mpm_worker module works similar to the mpm_prefork module. However, because the same request is processed, process-based (such as mpm_prefork) takes up more system resources than thread-based processing. Therefore, unlike the mpm_prefork module, the mpm_worker module has each child process create a fixed number of service threads and one listener thread, and each service thread handles the client's request, which is used to listen for access requests and pass them to the service thread for processing and reply. Apache always tries to maintain a pool of spare or idle service threads. In this way, the client can be processed without waiting for a new thread or process to be established.

Compared with mpm_prefork module, mpm_worker module can further reduce the overhead of system resources. In addition, it also uses multiple processes, and each process has multiple threads, so it adds a certain degree of stability compared with the thread-based processing.

Mpm_winnt module

Mpm_winnt module is a MPM module optimized for Windows operating system. It only creates a single child process and takes turns generating multiple threads in that child process to process the request.

Modify MPM module configuration

With some understanding of the MPM module of Apache, we can modify the maximum number of concurrent connections configuration of Apache for different MPM modules.

1. Enable MPM module profile

There is a configuration file called httpd-mpm.conf in the Apace installation directory / conf/extra directory. This file is mainly used for the configuration of the MPM module. However, the MPM module profile for Apache is not enabled by default. Therefore, we need to enable the configuration file in the httpd.conf file, as follows:

# Server-pool management (MPM specific) Include conf/extra/httpd-mpm.conf (remove the comment symbol "#" before the line)

two。 Modify the relevant configuration in the MPM module configuration file

After starting the MPM module configuration file, we can open the configuration file using a text editor, and we can see that there are many configuration nodes in the configuration file, as shown in the following figure:

Only if Apache uses the corresponding MPM module, the corresponding configuration will take effect.

At this point, we need to modify the parameter configuration under the corresponding node according to the MPM module used by the current Apache server. First, let's take a look at the default configuration under the mpm_winnt module:

# because the mpm_winnt module only creates one child process, setting the parameters of a single child process here is equivalent to setting the parameters of the entire Apache. ThreadsPerChild 15 recommendation setting: small website = 1000 medium website = 10000000000 large website = 2000~3500MaxRequestsPerChild recommendation setting: small = 10000 medium or large = 200000000000

The corresponding configuration parameters are as follows:

ThreadsPerChild

The maximum number of concurrent threads per child process.

MaxRequestsPerChild

The total number of requests allowed to be processed per child process. If the cumulative number of requests processed exceeds this value, the child process ends (and then determines whether to create a new child process as needed), and a value of 0 means that there is no limit to the total number of requests (the child process never ends).

It is recommended that this parameter be set to a non-zero value, which can bring the following two benefits:

It can prevent memory leaks that may exist in the program from going on indefinitely, thus running out of memory.

Give processes a limited lifespan, which helps to reduce the number of active processes when the server load is reduced.

Note: in the above parameters related to counting the number of requests, for KeepAlive connections, only the first request will be counted.

Next, let's look at the default configuration under the mpm_perfork module and the mpm_worker module:

# mpm_perfork module StartServers recommended setting: small = default medium = 20 "50 large = 50~100MinSpareServers" recommended setting: consistent with StartServers MaxSpareServers 1 "recommended setting: small = 20 medium = 30" 80 large = 80 "120 MaxClients 15" recommended setting: small = 500 medium = 500 "1500 large = 1500~3000MaxRequestsPerChild" recommended setting: small = 10000 medium or large = 10000mm 500000 (in addition, an additional ServerLimit parameter is required, which is preferably consistent with the value of MaxClients.)

# StartServers: number of server processes start # MinSpareServers: minimum number of server processes, save spare # MaxSpareServers: maximum number of server processes, save spare # MaxRequestWorkers: maximum number of server processes allowed to start # MaxConnectionsPerChild: maximum number of connections to a server process service

After the prefork control process initially established the "StartServers" child process, in order to meet the needs of the MinSpareServers setting, create one process, wait one second, continue to create two, and then wait one second to continue to create four. This increases the number of processes created exponentially to a maximum of 32 per second until the value set by MinSpareServers is met. This mode eliminates the need to generate new processes when the request arrives, thus reducing system overhead to increase performance. MaxSpareServers sets the maximum number of idle processes, and if the number of idle processes is greater than this value, Apache will automatically kill some extra processes. Do not set this value too high, but if the value is smaller than MinSpareServers, Apache will automatically adjust it to MinSpareServers+1. If the site is heavily loaded, consider increasing MinSpareServers and MaxSpareServers at the same time.

MaxRequestsPerChild sets the number of requests that each child process can handle. Each child process will automatically destroy after processing a "MaxRequestsPerChild" request. 0 means infinity, that is, child processes are never destroyed. Although the default setting of 0 allows each child process to process more requests, setting it to a non-zero value also has two important benefits:

1. It can prevent accidental memory leakage. 2. When the server load drops, the number of child processes will be reduced automatically.

Therefore, this value can be adjusted according to the load of the server.

The MaxRequestWorkers instruction set also limits the number of service requests. Any connection attempt in MaxRequestWorkerslimit will usually be queued, up to a few based on the ListenBacklog instruction.

In previous versions of apache2.3.13, MaxRequestWorkers was called MaxClients.

MaxClients is the most important of these instructions, setting requests that Apache can process at the same time, and is the parameter that has the greatest impact on Apache performance. The default value of 150 is far from enough, and if the total number of requests has reached this value (which can be confirmed by ps-ef | grep http | wc-l), subsequent requests will be queued until a request has been processed. This is the main reason why there are a lot of system resources left and HTTP access is slow. Although in theory, the higher this value, the more requests can be processed, the default limit for Apache cannot be greater than 256. )

# mpm_worker module StartServers recommendation settings: small = default medium = 3 "5 large = 5~10MaxClients 15" recommended settings: small = 500 medium = 500 "1500 large = 1500~3000MinSpareThreads 2" recommended settings: small = default medium = 50 "100 large = 100~200MaxSpareThreads 7" recommended settings: small = default medium = 80 "160 large = 200" 400 ThreadsPerChild 2 "recommended settings: small = default medium = 50" 100 large = 100~200MaxRequestsPerChild "recommended settings: small = 10000 medium or large = 10000mm 50000 (in addition, if the MaxClients/ThreadsPerChild is greater than 16, additional ServerLimit parameters are required ServerLimit must be greater than or equal to the value of MaxClients/ThreadsPerChild.)

The corresponding configuration parameters are as follows:

StartServers

The number of child processes created when Apache was started.

MinSpareServers

The number of minimum processes that are idle.

The so-called idle child process refers to no child process that is processing the request. If the current number of idle child processes is less than MinSpareServers, then Apache will generate new child processes at a maximum rate of one per second. This parameter needs to be adjusted only on very busy machines. This value should not be too large.

MaxSpareServers

The maximum number of child processes that are idle.

This parameter needs to be adjusted only on very busy machines. This value should not be too large. If you set the value of the instruction to be less than MinSpareServers, Apache will automatically change it to MinSpareServers+1.

MaxClients

The maximum number of requests allowed to connect at the same time.

Any request that exceeds the MaxClients limit will be queued until the maximum limit of the ListenBacklog instruction is reached.

For non-threaded MPM (that is, mpm_prefork), MaxClients represents the maximum number of child processes that can be used to process client requests, with a default value of 256. To increase this value, you must also increase ServerLimit.

For threaded or hybrid MPM (that is, mpm_beos or mpm_worker), MaxClients represents the maximum number of threads that can be used to process client requests. The default value for threaded mpm_beos is 50. For mixed MPM, the default value is 16 (ServerLimit) multiplied by 25 (ThreadsPerChild). So to increase the MaxClients to more than 16 processes to provide, you must also increase the value of ServerLimit.

MinSpareThreads

The minimum number of threads that are idle.

Different MPM processes this instruction differently:

The default value for mpm_worker is 75. This MPM will monitor the number of idle threads based on the entire server. If the total number of idle threads in the server is too small, the child process will generate new idle threads. The default value for mpm_netware is 10. Since this MPM runs only a single child process, this MPM of course also monitors the number of idle threads based on the entire server. Mpm_beos and mpmt_os2 work in much the same way as mpm_netware. The default value for mpm_beos is 1, and the default value for mpmtactuos2 is 5.

MaxSpareThreads

The maximum number of threads that are idle.

Different MPM processes this instruction differently:

The default value for mpm_worker is 250. This MPM will monitor the number of idle threads based on the entire server. If there are too many total free threads in the server, the child process will kill the extra idle threads. The default value for mpm_netware is 100. Since this MPM runs only a single child process, this MPM of course also monitors the number of idle threads based on the entire server. Mpm_beos and mpmt_os2 work in much the same way as mpm_netware, with a default value of 50 for mpm_beos and 10 for mpmtactuos2.

Note: ServerLimit represents the maximum number of processes allowed to be created by Apache. It is worth noting that Apache has a hard limit of ServerLimit 20000 internally at compile time (ServerLimit 200000 for mpm_prefork modules). You can't go beyond that limit.

Be very careful when using this instruction. If ServerLimit is set to a much higher value than is actually needed, too much shared memory will be allocated. If ServerLimit and MaxClients are set to exceed the processing capacity of the system, Apache may not start, or the system may become unstable.

Note: when configuring relevant parameters, make sure that the server has sufficient hardware performance (for example, CPU, memory, etc.). If you find that the memory usage of the server increases with the increase of the running time of the server since startup, which may be due to a memory leak in the program, please adjust the value of the parameter MaxRequestsPerChild down to reduce the impact of the memory leak, and then find out the problem in the program as soon as possible.

After reading this article, I believe you have a certain understanding of apache's method of modifying the maximum number of concurrent connections. Want to know more about it. Welcome to follow the industry information channel. Thank you for your reading!

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.