Chapter 15. Minimizing system latency by isolating interrupts and user processes

Real-time environments need to minimize or eliminate latency when responding to various events. To do this, you can isolate interrupts (IRQs) from user processes from one another on different dedicated CPUs.

15.1. Interrupt and process binding

Isolating interrupts (IRQs) from user processes on different dedicated CPUs can minimize or eliminate latency in real-time environments.

Interrupts are generally shared evenly between CPUs. This can delay interrupt processing when the CPU has to write new data and instruction caches. These interrupt delays can cause conflicts with other processing being performed on the same CPU.

It is possible to allocate time-critical interrupts and processes to a specific CPU (or a range of CPUs). In this way, the code and data structures for processing this interrupt will most likely be in the processor and instruction caches. As a result, the dedicated process can run as quickly as possible, while all other non-time-critical processes run on the other CPUs. This can be particularly important where the speeds involved are near or at the limits of memory and available peripheral bus bandwidth. Any wait for memory to be fetched into processor caches will have a noticeable impact in overall processing time and determinism.

In practice, optimal performance is entirely application-specific. For example, tuning applications with similar functions for different companies, required completely different optimal performance tunings.

  • One firm saw optimal results when they isolated 2 out of 4 CPUs for operating system functions and interrupt handling. The remaining 2 CPUs were dedicated purely for application handling.
  • Another firm found optimal determinism when they bound the network related application processes onto a single CPU which was handling the network device driver interrupt.

To bind a process to a CPU, you usually need to know the CPU mask for a given CPU or range of CPUs. The CPU mask is typically represented as a 32-bit bitmask, a decimal number, or a hexadecimal number, depending on the command you are using.

Table 15.1. Example of the CPU Mask for given CPUs









0, 1




15.2. Disabling the irqbalance daemon

The irqbalance daemon is enabled by default and periodically forces interrupts to be handled by CPUs in an even manner. However in real-time deployments, irqbalance is not needed, because applications are typically bound to specific CPUs.


  1. Check the status of irqbalance.

    # systemctl status irqbalance
    irqbalance.service - irqbalance daemon
       Loaded: loaded (/usr/lib/systemd/system/irqbalance.service; enabled)
       Active: active (running) …
  2. If irqbalance is running, disable it, and stop it.

    # systemctl disable irqbalance
    # systemctl stop irqbalance


  • Check that the irqbalance status is inactive.

    # systemctl status irqbalance

15.3. Excluding CPUs from IRQ balancing

You can use the IRQ balancing service to specify which CPUs you want to exclude from consideration for interrupt (IRQ) balancing. The IRQBALANCE_BANNED_CPUS parameter in the /etc/sysconfig/irqbalance configuration file controls these settings. The value of the parameter is a 64-bit hexadecimal bit mask, where each bit of the mask represents a CPU core.


  1. Open /etc/sysconfig/irqbalance in your preferred text editor and find the section of the file titled IRQBALANCE_BANNED_CPUS.

    # 64 bit bitmask which allows you to indicate which cpu's should
    # be skipped when reblancing irqs. Cpu numbers which have their
    # corresponding bits set to one in this mask will not have any
    # irq's assigned to them on rebalance
  2. Uncomment the IRQBALANCE_BANNED_CPUS variable.
  3. Enter the appropriate bitmask to specify the CPUs to be ignored by the IRQ balance mechanism.
  4. Save and close the file.

If you are running a system with up to 64 CPU cores, separate each group of eight hexadecimal digits with a comma. For example: IRQBALANCE_BANNED_CPUS=00000001,0000ff00

Table 15.2. Examples





8 - 15


8 - 15, 33



In RHEL 7.2 and higher, the irqbalance utility automatically avoids IRQs on CPU cores isolated via the isolcpus kernel parameter if IRQBALANCE_BANNED_CPUS is not set in /etc/sysconfig/irqbalance.

15.4. Manually assigning CPU affinity to individual IRQs

Assigning CPU affinity enables binding and unbinding processes and threads to a specified CPU or range of CPUs. This can reduce caching problems.


  1. Check the IRQs in use by each device by viewing the /proc/interrupts file.

    # cat /proc/interrupts

    Each line shows the IRQ number, the number of interrupts that happened in each CPU, followed by the IRQ type and a description.

             CPU0       CPU1
    0:   26575949         11         IO-APIC-edge  timer
    1:         14          7         IO-APIC-edge  i8042
  2. Write the CPU mask to the smp_affinity entry of a specific IRQ. The CPU mask must be expressed as a hexadecimal number.

    For example, the following command instructs IRQ number 142 to run only on CPU 0.

    # echo 1 > /proc/irq/142/smp_affinity

    The change only takes effect when an interrupt occurs.

Verification steps

  1. Perform an activity that will trigger the specified interrupt.
  2. Check /proc/interrupts for changes.

    The number of interrupts on the specified CPU for the configured IRQ increased, and the number of interrupts for the configured IRQ on CPUs outside the specified affinity did not increase.

15.5. Binding processes to CPUs with the taskset utility

The taskset utility uses the process ID (PID) of a task to view or set its CPU affinity. You can use the utility to run a command with a chosen CPU affinity.

To set the affinity, you need to get the CPU mask to be as a decimal or hexadecimal number. The mask argument is a bitmask that specifies which CPU cores are legal for the command or PID being modified.


The taskset` utility works on a NUMA (Non-Uniform Memory Access) system, but it does not allow the user to bind threads to CPUs and the closest NUMA memory node. On such systems, taskset is not the preferred tool, and the numactl utility should be used instead for its advanced capabilities.

For more information, see the numactl(8) man page.


  • Run taskset with the necessary options and arguments.

    • You can specify a CPU list using the -c parameter instead of a CPU mask. In this example, my_embedded_process is being instructed to run only on CPUs 0,4,7-11.

      # taskset -c 0,4,7-11 /usr/local/bin/my_embedded_process

      This invocation is more convenient in most cases.

    • To set the affinity of a process that is not currently running, use taskset and specify the CPU mask and the process.

      In this example, my_embedded_process is being instructed to use only CPU 3 (using the decimal version of the CPU mask).

      # taskset 8 /usr/local/bin/my_embedded_process
    • You can specify more than one CPU in the bitmask. In this example, my_embedded_process is being instructed to execute on processors 4, 5, 6, and 7 (using the hexadecimal version of the CPU mask).

      # taskset 0xF0 /usr/local/bin/my_embedded_process
    • You can set the CPU affinity for processes that are already running by using the -p (--pid) option with the CPU mask and the PID of the process you want to change. In this example, the process with a PID of 7013 is being instructed to run only on CPU 0.

      # taskset -p 1 7013

You can combine the listed options.

Additional resources

  • taskset(1) man page
  • numactl(8) man page