Chapter 2. Important Changes to External Kernel Parameters

This chapter provides system administrators with a summary of significant changes in the kernel shipped with Red Hat Enterprise Linux 6.5. These changes include added or updated procfs entries, sysfs default values, boot parameters, kernel configuration options, or any noticeable behavior changes.
This RW file contains a number of reserved blocks in the file system which are used in specific situations to avoid unexpected No space left on device (ENOSPC) errors or possible data loss.
proc/<pid>/comm and /proc/<pid>/task/<tid>/comm files
These files provide a method to access a task's comm value. It also allows for a task to set its own or one of its thread siblings' comm values. The comm value is limited in size compared to the cmdline value, so writing anything longer then the kernel's TASK_COMM_LEN macro (currently 16 chars) will result in a truncated comm value.
This parameter allows users to enable power limit notification interrupts.
The default value of this parameter is 0. When set to 1, NFSv4 server returns only numeric user IDs (UIDs) and group IDs (GIDs) to clients using AUTH_SYS mode, and will accept numeric UIDs and GIDs from such clients. This facilitates migration from NFS version 2 to NFS version 3.
PCI Subsystem Options
The following options for the pci kernel parameter can be used in Red Hat Enterprise Linux 6.5:
  • pcie_bus_tune_off—disables PCIe maximum payload size (MPS) tuning and uses the BIOS-configured MPS default values.
  • pcie_bus_safe—sets every device MPS to the largest value supported by all devices below the root complex.
  • pcie_bus_perf—sets the device MPS to the largest allowable MPS based on its parent bus.
  • pcie_bus_peer2peer— sets every device's MPS to 128B, which every device is guaranteed to support.
With this parameter, universally unique identifiers (UUIDs) are displayed in the System Management BIOS (SMBIOS) 2.6 format.
With this parameter, additional information about the Time Stamp Counter (TSC) is displayed during system boot.
This option displays memory limit in MB for buffers allowated by USB device file system (usbfs).
tcp_limit_output_bytes controls TCP Small Queue limit per TCP socket.
tcp_challenge_ack_limit limits the number of challenge acknowledgements sent per second, as recommended in RFC 5961 (Improving TCP's Robustness to Blind In-Window Attacks).
The accept_ra boolean allows for accepting router discovery messages (also known as router advertisements).
cookie_hmac_alg is used to select the keyed-hash message authentication code (HMAC) algorithm used when generating the cookie value sent by a listening SCTP socket to a connecting client in the INIT-ACK chunk. Valid values are:
  • md5
  • sha1
  • none
The nf_conntrack_acct boolean enables connection tracking flow accounting.
nf_conntrack_buckets determines the size of a hash table. If it is not specified as parameter during module loading, the default size is calculated by dividing total memory by 16384 to determine the number of buckets but the hash table will never have fewer than 32 nor more than 16384 buckets.
This parameter is used to verify the checksums of incoming packets. Packets with invalid checksums are in INVALID state. If this is enabled, such packets will not be considered for connection tracking.
This option is only relevant when "reliable connection tracking events" are used. Normally, ctnetlink is "lossy", that is, events are normally dropped when userspace listeners cannot keep up. Userspace can request "reliable event mode". When this mode is active, the connection tracking will only be destroyed after the event was delivered. If event delivery fails, the kernel periodically re-tries to send the event to userspace. The default value 15 is the maximum interval the kernel should use when re-trying to deliver the destroy event. A higher number means there will be fewer delivery retries and it will take longer for a backlog to be processed.
The merge_across_nodes parameter specifies if pages from different NUMA nodes can be merged. When set to 0, Kernel SamePage Merging (KSM) merges only pages which physically reside in the memory area of the same NUMA node. 1 is the default value and merging across nodes is performed as in earlier releases.