18.4. Managing Kernel Resources

Postgres Pro can sometimes exhaust various operating system resource limits, especially when multiple copies of the server are running on the same system, or in very large installations. This section explains the kernel resources used by Postgres Pro and the steps you can take to resolve problems related to kernel resource consumption.

18.4.1. Shared Memory and Semaphores

Postgres Pro requires the operating system to provide inter-process communication (IPC) features, specifically shared memory and semaphores. Unix-derived systems typically provide System V IPC, POSIX IPC, or both. Windows has its own implementation of these features and is not discussed here.

The complete lack of these facilities is usually manifested by an Illegal system call error upon server start. In that case there is no alternative but to reconfigure your kernel. Postgres Pro won't work without them. This situation is rare, however, among modern operating systems.

Upon starting the server, Postgres Pro normally allocates a very small amount of System V shared memory, as well as a much larger amount of POSIX (mmap) shared memory. In addition a significant number of semaphores, which can be either System V or POSIX style, are created at server startup. Currently, POSIX semaphores are used on Linux and FreeBSD systems while other platforms use System V semaphores.

Note

Prior to PostgreSQL 9.3, only System V shared memory was used, so the amount of System V shared memory required to start the server was much larger. If you are running an older version of the server, please consult the documentation for your server version.

System V IPC features are typically constrained by system-wide allocation limits. When Postgres Pro exceeds one of these limits, the server will refuse to start and should leave an instructive error message describing the problem and what to do about it. (See also Section 18.3.1.) The relevant kernel parameters are named consistently across different systems; Table 18.1 gives an overview. The methods to set them, however, vary. Suggestions for some platforms are given below.

Table 18.1. System V IPC Parameters

NameDescriptionValues needed to run one Postgres Pro instance
SHMMAXMaximum size of shared memory segment (bytes)at least 1kB, but the default is usually much higher
SHMMINMinimum size of shared memory segment (bytes)1
SHMALLTotal amount of shared memory available (bytes or pages)same as SHMMAX if bytes, or ceil(SHMMAX/PAGE_SIZE) if pages, plus room for other applications
SHMSEGMaximum number of shared memory segments per processonly 1 segment is needed, but the default is much higher
SHMMNIMaximum number of shared memory segments system-widelike SHMSEG plus room for other applications
SEMMNIMaximum number of semaphore identifiers (i.e., sets)at least ceil((max_connections + autovacuum_max_workers + max_worker_processes + 5) / 16) plus room for other applications
SEMMNSMaximum number of semaphores system-wideceil((max_connections + autovacuum_max_workers + max_worker_processes + 5) / 16) * 17 plus room for other applications
SEMMSLMaximum number of semaphores per setat least 17
SEMMAPNumber of entries in semaphore mapsee text
SEMVMXMaximum value of semaphoreat least 1000 (The default is often 32767; do not change unless necessary)

Postgres Pro requires a few bytes of System V shared memory (typically 48 bytes, on 64-bit platforms) for each copy of the server. On most modern operating systems, this amount can easily be allocated. However, if you are running many copies of the server, or if other applications are also using System V shared memory, it may be necessary to increase SHMALL, which is the total amount of System V shared memory system-wide. Note that SHMALL is measured in pages rather than bytes on many systems.

Less likely to cause problems is the minimum size for shared memory segments (SHMMIN), which should be at most approximately 32 bytes for Postgres Pro (it is usually just 1). The maximum number of segments system-wide (SHMMNI) or per-process (SHMSEG) are unlikely to cause a problem unless your system has them set to zero.

When using System V semaphores, Postgres Pro uses one semaphore per allowed connection (max_connections), allowed autovacuum worker process (autovacuum_max_workers) and allowed background process (max_worker_processes), in sets of 16. Each such set will also contain a 17th semaphore which contains a magic number, to detect collision with semaphore sets used by other applications. The maximum number of semaphores in the system is set by SEMMNS, which consequently must be at least as high as max_connections plus autovacuum_max_workers plus max_worker_processes, plus one extra for each 16 allowed connections plus workers (see the formula in Table 18.1). The parameter SEMMNI determines the limit on the number of semaphore sets that can exist on the system at one time. Hence this parameter must be at least ceil((max_connections + autovacuum_max_workers + max_worker_processes + 5) / 16). Lowering the number of allowed connections is a temporary workaround for failures, which are usually confusingly worded No space left on device, from the function semget.

In some cases it might also be necessary to increase SEMMAP to be at least on the order of SEMMNS. If the system has this parameter (many do not), it defines the size of the semaphore resource map, in which each contiguous block of available semaphores needs an entry. When a semaphore set is freed it is either added to an existing entry that is adjacent to the freed block or it is registered under a new map entry. If the map is full, the freed semaphores get lost (until reboot). Fragmentation of the semaphore space could over time lead to fewer available semaphores than there should be.

Various other settings related to semaphore undo, such as SEMMNU and SEMUME, do not affect Postgres Pro.

When using POSIX semaphores, the number of semaphores needed is the same as for System V, that is one semaphore per allowed connection (max_connections), allowed autovacuum worker process (autovacuum_max_workers) and allowed background process (max_worker_processes). On the platforms where this option is preferred, there is no specific kernel limit on the number of POSIX semaphores.

AIX

At least as of version 5.1, it should not be necessary to do any special configuration for such parameters as SHMMAX, as it appears this is configured to allow all memory to be used as shared memory. That is the sort of configuration commonly used for other databases such as DB/2.

It might, however, be necessary to modify the global ulimit information in /etc/security/limits, as the default hard limits for file sizes (fsize) and numbers of files (nofiles) might be too low.

FreeBSD

The default IPC settings can be changed using the sysctl or loader interfaces. The following parameters can be set using sysctl:

# sysctl kern.ipc.shmall=32768
# sysctl kern.ipc.shmmax=134217728

To make these settings persist over reboots, modify /etc/sysctl.conf.

These semaphore-related settings are read-only as far as sysctl is concerned, but can be set in /boot/loader.conf:

kern.ipc.semmni=256
kern.ipc.semmns=512

After modifying that file, a reboot is required for the new settings to take effect.

You might also want to configure your kernel to lock shared memory into RAM and prevent it from being paged out to swap. This can be accomplished using the sysctl setting kern.ipc.shm_use_phys.

If running in FreeBSD jails by enabling sysctl's security.jail.sysvipc_allowed, postmasters running in different jails should be run by different operating system users. This improves security because it prevents non-root users from interfering with shared memory or semaphores in different jails, and it allows the Postgres Pro IPC cleanup code to function properly. (In FreeBSD 6.0 and later the IPC cleanup code does not properly detect processes in other jails, preventing the running of postmasters on the same port in different jails.)

FreeBSD versions before 4.0 work like old OpenBSD (see below).

NetBSD

In NetBSD 5.0 and later, IPC parameters can be adjusted using sysctl, for example:

# sysctl -w kern.ipc.semmni=100

To make these settings persist over reboots, modify /etc/sysctl.conf.

You will usually want to increase kern.ipc.semmni and kern.ipc.semmns, as NetBSD's default settings for these are uncomfortably small.

You might also want to configure your kernel to lock shared memory into RAM and prevent it from being paged out to swap. This can be accomplished using the sysctl setting kern.ipc.shm_use_phys.

NetBSD versions before 5.0 work like old OpenBSD (see below), except that kernel parameters should be set with the keyword options not option.

OpenBSD

In OpenBSD 3.3 and later, IPC parameters can be adjusted using sysctl, for example:

# sysctl kern.seminfo.semmni=100

To make these settings persist over reboots, modify /etc/sysctl.conf.

You will usually want to increase kern.seminfo.semmni and kern.seminfo.semmns, as OpenBSD's default settings for these are uncomfortably small.

In older OpenBSD versions, you will need to build a custom kernel to change the IPC parameters. Make sure that the options SYSVSHM and SYSVSEM are enabled, too. (They are by default.) The following shows an example of how to set the various parameters in the kernel configuration file:

option        SYSVSHM
option        SHMMAXPGS=4096
option        SHMSEG=256

option        SYSVSEM
option        SEMMNI=256
option        SEMMNS=512
option        SEMMNU=256

HP-UX

The default settings tend to suffice for normal installations. On HP-UX 10, the factory default for SEMMNS is 128, which might be too low for larger database sites.

IPC parameters can be set in the System Administration Manager (SAM) under Kernel ConfigurationConfigurable Parameters. Choose Create A New Kernel when you're done.

Linux

The default maximum segment size is 32 MB, and the default maximum total size is 2097152 pages. A page is almost always 4096 bytes except in unusual kernel configurations with huge pages (use getconf PAGE_SIZE to verify).

The shared memory size settings can be changed via the sysctl interface. For example, to allow 16 GB:

$ sysctl -w kernel.shmmax=17179869184
$ sysctl -w kernel.shmall=4194304

In addition these settings can be preserved between reboots in the file /etc/sysctl.conf. Doing that is highly recommended.

Ancient distributions might not have the sysctl program, but equivalent changes can be made by manipulating the /proc file system:

$ echo 17179869184 >/proc/sys/kernel/shmmax
$ echo 4194304 >/proc/sys/kernel/shmall

The remaining defaults are quite generously sized, and usually do not require changes.

macOS

The recommended method for configuring shared memory in macOS is to create a file named /etc/sysctl.conf, containing variable assignments such as:

kern.sysv.shmmax=4194304
kern.sysv.shmmin=1
kern.sysv.shmmni=32
kern.sysv.shmseg=8
kern.sysv.shmall=1024

Note that in some macOS versions, all five shared-memory parameters must be set in /etc/sysctl.conf, else the values will be ignored.

Beware that recent releases of macOS ignore attempts to set SHMMAX to a value that isn't an exact multiple of 4096.

SHMALL is measured in 4 kB pages on this platform.

In older macOS versions, you will need to reboot to have changes in the shared memory parameters take effect. As of 10.5 it is possible to change all but SHMMNI on the fly, using sysctl. But it's still best to set up your preferred values via /etc/sysctl.conf, so that the values will be kept across reboots.

The file /etc/sysctl.conf is only honored in macOS 10.3.9 and later. If you are running a previous 10.3.x release, you must edit the file /etc/rc and change the values in the following commands:

sysctl -w kern.sysv.shmmax
sysctl -w kern.sysv.shmmin
sysctl -w kern.sysv.shmmni
sysctl -w kern.sysv.shmseg
sysctl -w kern.sysv.shmall

Note that /etc/rc is usually overwritten by macOS system updates, so you should expect to have to redo these edits after each update.

In macOS 10.2 and earlier, instead edit these commands in the file /System/Library/StartupItems/SystemTuning/SystemTuning.

Solaris 2.6 to 2.9 (Solaris 6 to Solaris 9)

The relevant settings can be changed in /etc/system, for example:

set shmsys:shminfo_shmmax=0x2000000
set shmsys:shminfo_shmmin=1
set shmsys:shminfo_shmmni=256
set shmsys:shminfo_shmseg=256

set semsys:seminfo_semmap=256
set semsys:seminfo_semmni=512
set semsys:seminfo_semmns=512
set semsys:seminfo_semmsl=32

You need to reboot for the changes to take effect. See also http://sunsite.uakom.sk/sunworldonline/swol-09-1997/swol-09-insidesolaris.html for information on shared memory under older versions of Solaris.

Solaris 2.10 (Solaris 10) and later
OpenSolaris

In Solaris 10 and later, and OpenSolaris, the default shared memory and semaphore settings are good enough for most Postgres Pro applications. Solaris now defaults to a SHMMAX of one-quarter of system RAM. To further adjust this setting, use a project setting associated with the postgres user. For example, run the following as root:

projadd -c "Postgres Pro DB User" -K "project.max-shm-memory=(privileged,8GB,deny)" -U postgres -G postgres user.postgres

This command adds the user.postgres project and sets the shared memory maximum for the postgres user to 8GB, and takes effect the next time that user logs in, or when you restart Postgres Pro (not reload). The above assumes that Postgres Pro is run by the postgres user in the postgres group. No server reboot is required.

Other recommended kernel setting changes for database servers which will have a large number of connections are:

project.max-shm-ids=(priv,32768,deny)
project.max-sem-ids=(priv,4096,deny)
project.max-msg-ids=(priv,4096,deny)

Additionally, if you are running Postgres Pro inside a zone, you may need to raise the zone resource usage limits as well. See "Chapter2: Projects and Tasks" in the System Administrator's Guide for more information on projects and prctl.

18.4.2. systemd RemoveIPC

If systemd is in use, some care must be taken that IPC resources (including shared memory) are not prematurely removed by the operating system. This is especially of concern when installing Postgres Pro from source. Users of distribution packages of Postgres Pro are less likely to be affected, as the postgres user is then normally created as a system user.

The setting RemoveIPC in logind.conf controls whether IPC objects are removed when a user fully logs out. System users are exempt. This setting defaults to on in stock systemd, but some operating system distributions default it to off.

A typical observed effect when this setting is on is that shared memory objects used for parallel query execution are removed at apparently random times, leading to errors and warnings while attempting to open and remove them, like

WARNING:  could not remove shared memory segment "/PostgreSQL.1450751626": No such file or directory

Different types of IPC objects (shared memory vs. semaphores, System V vs. POSIX) are treated slightly differently by systemd, so one might observe that some IPC resources are not removed in the same way as others. But it is not advisable to rely on these subtle differences.

A user logging out might happen as part of a maintenance job or manually when an administrator logs in as the postgres user or something similar, so it is hard to prevent in general.

What is a system user is determined at systemd compile time from the SYS_UID_MAX setting in /etc/login.defs.

Packaging and deployment scripts should be careful to create the postgres user as a system user by using useradd -r, adduser --system, or equivalent.

Alternatively, if the user account was created incorrectly or cannot be changed, it is recommended to set

RemoveIPC=no

in /etc/systemd/logind.conf or another appropriate configuration file.

Caution

At least one of these two things has to be ensured, or the Postgres Pro server will be very unreliable.

18.4.3. Resource Limits

Unix-like operating systems enforce various kinds of resource limits that might interfere with the operation of your Postgres Pro server. Of particular importance are limits on the number of processes per user, the number of open files per process, and the amount of memory available to each process. Each of these have a hard and a soft limit. The soft limit is what actually counts but it can be changed by the user up to the hard limit. The hard limit can only be changed by the root user. The system call setrlimit is responsible for setting these parameters. The shell's built-in command ulimit (Bourne shells) or limit (csh) is used to control the resource limits from the command line. On BSD-derived systems the file /etc/login.conf controls the various resource limits set during login. See the operating system documentation for details. The relevant parameters are maxproc, openfiles, and datasize. For example:

default:\
...
        :datasize-cur=256M:\
        :maxproc-cur=256:\
        :openfiles-cur=256:\
...

(-cur is the soft limit. Append -max to set the hard limit.)

Kernels can also have system-wide limits on some resources.

  • On Linux the kernel parameter fs.file-max determines the maximum number of open files that the kernel will support. It can be changed with sysctl -w fs.file-max=N. To make the setting persist across reboots, add an assignment in /etc/sysctl.conf. The maximum limit of files per process is fixed at the time the kernel is compiled; see /usr/src/linux/Documentation/proc.txt for more information.

The Postgres Pro server uses one process per connection so you should provide for at least as many processes as allowed connections, in addition to what you need for the rest of your system. This is usually not a problem but if you run several servers on one machine things might get tight.

The factory default limit on open files is often set to socially friendly values that allow many users to coexist on a machine without using an inappropriate fraction of the system resources. If you run many servers on a machine this is perhaps what you want, but on dedicated servers you might want to raise this limit.

On the other side of the coin, some systems allow individual processes to open large numbers of files; if more than a few processes do so then the system-wide limit can easily be exceeded. If you find this happening, and you do not want to alter the system-wide limit, you can set Postgres Pro's max_files_per_process configuration parameter to limit the consumption of open files.

Another kernel limit that may be of concern when supporting large numbers of client connections is the maximum socket connection queue length. If more than that many connection requests arrive within a very short period, some may get rejected before the postmaster can service the requests, with those clients receiving unhelpful connection failure errors such as Resource temporarily unavailable or Connection refused. The default queue length limit is 128 on many platforms. To raise it, adjust the appropriate kernel parameter via sysctl, then restart the postmaster. The parameter is variously named net.core.somaxconn on Linux, kern.ipc.soacceptqueue on newer FreeBSD, and kern.ipc.somaxconn on macOS and other BSD variants.

18.4.4. Linux Memory Overcommit

In Linux 2.4 and later, the default virtual memory behavior is not optimal for Postgres Pro. Because of the way that the kernel implements memory overcommit, the kernel might terminate the Postgres Pro postmaster (the master server process) if the memory demands of either Postgres Pro or another process cause the system to run out of virtual memory.

If this happens, you will see a kernel message that looks like this (consult your system documentation and configuration on where to look for such a message):

Out of Memory: Killed process 12345 (postgres).

This indicates that the postgres process has been terminated due to memory pressure. Although existing database connections will continue to function normally, no new connections will be accepted. To recover, Postgres Pro will need to be restarted.

One way to avoid this problem is to run Postgres Pro on a machine where you can be sure that other processes will not run the machine out of memory. If memory is tight, increasing the swap space of the operating system can help avoid the problem, because the out-of-memory (OOM) killer is invoked only when physical memory and swap space are exhausted.

If Postgres Pro itself is the cause of the system running out of memory, you can avoid the problem by changing your configuration. In some cases, it may help to lower memory-related configuration parameters, particularly shared_buffers and work_mem. In other cases, the problem may be caused by allowing too many connections to the database server itself. In many cases, it may be better to reduce max_connections and instead make use of external connection-pooling software.

On Linux 2.6 and later, it is possible to modify the kernel's behavior so that it will not overcommit memory. Although this setting will not prevent the OOM killer from being invoked altogether, it will lower the chances significantly and will therefore lead to more robust system behavior. This is done by selecting strict overcommit mode via sysctl:

sysctl -w vm.overcommit_memory=2

or placing an equivalent entry in /etc/sysctl.conf. You might also wish to modify the related setting vm.overcommit_ratio. For details see the kernel documentation file https://www.kernel.org/doc/Documentation/vm/overcommit-accounting.

Another approach, which can be used with or without altering vm.overcommit_memory, is to set the process-specific OOM score adjustment value for the postmaster process to -1000, thereby guaranteeing it will not be targeted by the OOM killer. The simplest way to do this is to execute

echo -1000 > /proc/self/oom_score_adj

in the postmaster's startup script just before invoking the postmaster. Note that this action must be done as root, or it will have no effect; so a root-owned startup script is the easiest place to do it. If you do this, you should also set these environment variables in the startup script before invoking the postmaster:

export PG_OOM_ADJUST_FILE=/proc/self/oom_score_adj
export PG_OOM_ADJUST_VALUE=0

These settings will cause postmaster child processes to run with the normal OOM score adjustment of zero, so that the OOM killer can still target them at need. You could use some other value for PG_OOM_ADJUST_VALUE if you want the child processes to run with some other OOM score adjustment. (PG_OOM_ADJUST_VALUE can also be omitted, in which case it defaults to zero.) If you do not set PG_OOM_ADJUST_FILE, the child processes will run with the same OOM score adjustment as the postmaster, which is unwise since the whole point is to ensure that the postmaster has a preferential setting.

Older Linux kernels do not offer /proc/self/oom_score_adj, but may have a previous version of the same functionality called /proc/self/oom_adj. This works the same except the disable value is -17 not -1000.

Note

Some vendors' Linux 2.4 kernels are reported to have early versions of the 2.6 overcommit sysctl parameter. However, setting vm.overcommit_memory to 2 on a 2.4 kernel that does not have the relevant code will make things worse, not better. It is recommended that you inspect the actual kernel source code (see the function vm_enough_memory in the file mm/mmap.c) to verify what is supported in your kernel before you try this in a 2.4 installation. The presence of the overcommit-accounting documentation file should not be taken as evidence that the feature is there. If in any doubt, consult a kernel expert or your kernel vendor.

18.4.5. Linux Huge Pages

Using huge pages reduces overhead when using large contiguous chunks of memory, as Postgres Pro does, particularly when using large values of shared_buffers. To use this feature in Postgres Pro you need a kernel with CONFIG_HUGETLBFS=y and CONFIG_HUGETLB_PAGE=y. You will also have to adjust the kernel setting vm.nr_hugepages. To estimate the number of huge pages needed, start Postgres Pro without huge pages enabled and check the postmaster's anonymous shared memory segment size, as well as the system's huge page size, using the /proc file system. This might look like:

$ head -1 $PGDATA/postmaster.pid
4170
$ pmap 4170 | awk '/rw-s/ && /zero/ {print $2}'
6490428K
$ grep ^Hugepagesize /proc/meminfo
Hugepagesize:       2048 kB

6490428 / 2048 gives approximately 3169.154, so in this example we need at least 3170 huge pages, which we can set with:

$ sysctl -w vm.nr_hugepages=3170

A larger setting would be appropriate if other programs on the machine also need huge pages. Don't forget to add this setting to /etc/sysctl.conf so that it will be reapplied after reboots.

Sometimes the kernel is not able to allocate the desired number of huge pages immediately, so it might be necessary to repeat the command or to reboot. (Immediately after a reboot, most of the machine's memory should be available to convert into huge pages.) To verify the huge page allocation situation, use:

$ grep Huge /proc/meminfo

It may also be necessary to give the database server's operating system user permission to use huge pages by setting vm.hugetlb_shm_group via sysctl, and/or give permission to lock memory with ulimit -l.

The default behavior for huge pages in Postgres Pro is to use them when possible and to fall back to normal pages when failing. To enforce the use of huge pages, you can set huge_pages to on in postgresql.conf. Note that with this setting Postgres Pro will fail to start if not enough huge pages are available.

For a detailed description of the Linux huge pages feature have a look at https://www.kernel.org/doc/Documentation/vm/hugetlbpage.txt.

18.4.6. Resource Prioritization

Postgres Pro Enterprise provides an experimental feature for resource prioritization.

On systems with limited resources or under heavy load, you may need to prioritize transaction execution, so that some transactions are executed more quickly than the other. For example, you may want to execute simple user queries as fast as possible, even if it delays less urgent tasks, such as complex OLAP queries that may be running at the same time. Postgres Pro Enterprise enables you to configure resource prioritization policy, which can slow down a particular session based on the amount of CPU, I/O read, and I/O write resources this session consumes as compared to other sessions.

By default, resource prioritization is disabled, so all backends have equal access to all the available resources. You can assign weight to each backend to control the amount of resources each session can use within the specified time interval. Depending on the current resource consumption, Postgres Pro Enterprise may suspend backends with lower weight from time to time to ensure that high-priority sessions have more resources available.

To enable prioritization in your database cluster:

  1. Configure the time interval for collecting usage statistics for all active backends by setting the usage_tracking_interval parameter in the postgresql.conf file and reload the server configuration.

    Once the usage_tracking_interval parameter is set, Postgres Pro Enterprise starts collecting statistics on resource usage at the specified interval.

    Tip

    Avoid setting usage_tracking_interval to small values as frequent statistics collection can cause overhead.

  2. Depending on the resources you need to control, modify one or more of the following parameters for the sessions you would like to prioritize:

    These parameters can take weight values 1, 2, 4, and 8. The higher the value, the more resources the session can use. Sessions with the same weight have the same priority for resource usage, so if you assign equal weights to all sessions, performance is not affected, regardless of the weight value. By default, all sessions have weight 4 for all types of resources.

    For all possible ways of modifying configuration for a particular session, see Section 19.1.

Once you change the weight of one or more sessions, Postgres Pro Enterprise enables prioritization policy based on the assigned weight values and the usage statistics measured for the previous usage_tracking_interval. Thus, session activity is adjusted for each usage tracking interval, if required.

Note

Even though weights for each resource are assigned separately, prioritizing a session by one resource can indirectly affect the session performance with regard to other resources.