1.14. /proc

IDE channel (0 or 1)

Configuration (only for PCI/IDE bridge)

Mate name (onchip partnered controller)

Type/Chipset of IDE controller

Each device connected to a controller has a separate subdirectory in the controllers directory. The following files listed are contained in these directories:

The cache.

Capacity of the medium (in 512Byte blocks)

driver and version

physical and logical geometry

device identify block

media type

device identifier

device setup

IDE disk management thresholds

IDE disk management values

You can, for example, check which interrupts are currently in use and what they are used for by looking in the file /proc/interrupts:

In 2.4 based kernels a couple of lines were added to this file LOC & ERR (this is the output of an SMP machine):

NMI is incremented in this case because every timer interrupt generates a NMI (Non Maskable Interrupt) which is used by the NMI Watchdog to detect lookups.

LOC is the local interrupt counter of the internal APIC of every CPU.

ERR is incremented in the case of errors in the IO-APIC bus (the bus that connects the CPUs in an SMP system. This means that an error has been detected, the IO-APIC automatically retries the transmission, so it should not be a big problem, but you should read the SMP-FAQ.

In this context it could be interesting to note the new irq directory in 2.4. It could be used to set IRQ to CPU affinity, this means that you can "hook" an IRQ to only one CPU, or to exclude a CPU from handling IRQs. The contents of the irq subdir is one subdir for each IRQ, and one file; prof_cpu_mask. For example,

The contents of the prof_cpu_mask file and each smp_affinity file for each IRQ is the same by default:

It's a bitmask, in which you can specify which CPUs can handle the IRQ, you can set it by doing:

This means that only the first CPU will handle the IRQ, but you can also echo 5 which means that only the first and fourth CPU can handle the IRQ. The way IRQs are routed is handled by the IO-APIC, and its Round Robin between all the CPUs which are allowed to handle it. As usual the kernel has more info than you and does a better job than you, so the defaults are the best choice for almost everyone.

UDP sockets (IPv6).

TCP sockets (IPv6).

Raw device statistics (IPv6).

IP multicast addresses, which this host joined (IPv6).

List of IPv6 interface addresses.

Kernel routing table for IPv6.

Global IPv6 routing tables statistics.

Socket statistics (IPv6).

Snmp data (IPv6).

Kernel ARP table.

network devices with statistics.

the Layer2 multicast groups which a device is listening to (interface index, label, number of references, number of bound addresses).

network device status.

Firewall chain linkage.

Firewall chain names.

Directory containing the masquerading tables.

Major masquerading table.

Network statistics.

raw device statistics.

Kernel routing table.

Directory containing rpc info.

Routing cache.

SNMP data.

Socket statistics.

TCP sockets.

Token ring RIF routing table.

UDP sockets.

UNIX domain sockets.

Wireless interface data (Wavelan etc).

IP multicast addresses, which this host joined.

Global packet scheduler parameters.

List of PF_NETLINK sockets.

List of multicast virtual interfaces.

List of multicast routing cache.

You can use this information to see which network devices are available in your system and how much traffic was routed over those devices. In addition, each Channel Bond interface has its own directory. For example, the bond0 device will have a directory called /proc/net/bond0/. It will contain information that is specific to that bond, such as the current slaves of the bond, the link status of the slaves, and how many times the slaves link has failed.

Any IEEE-1284 device ID information that has been acquired.

list of the device drivers using that port. A + will appear by the name of the device currently using the port (it might not appear against any).

Parallel port's base address, IRQ line and DMA channel.

IRQ that parport is using for that port. This is in a separate file to allow you to alter it by writing a new value in (IRQ number or none).

Status of the directory cache. Since directory entries are dynamically allocated and deallocated, this file indicates the current status. It holds six values, in which the last two are not used and are always zero. The others are listed below:

The file dquot-max shows the maximum number of cached disk quota entries.

shows the number of allocated disk quota entries and the number of free disk quota entries. If the number of available cached disk quotas is very low and you have a large number of simultaneous system users, you might want to raise the limit.

The kernel allocates file handles dynamically, but doesn't free them again at this time. The value in file-max denotes the maximum number of file handles that the Linux kernel will allocate. When you get a lot of error messages about running out of file handles, you might want to raise this limit. The default value is 4096. To change it, just write the new number into the file:

This method of revision is useful for all customizable parameters of the kernel - simply echo the new value to the corresponding file.

The three values in file-nr denote the number of allocated file handles, the number of used file handles, and the maximum number of file handles. When the allocated file handles come close to the maximum, but the number of actually used handles is far behind, you've encountered a peak in your usage of file handles and you don't need to increase the maximum.

As with file handles, the kernel allocates the inode structures dynamically, but can't free them yet.

The value in inode-max denotes the maximum number of inode handlers. This value should be 3 to 4 times larger than the value in file-max, since stdin, stdout, and network sockets also need an inode struct to handle them. If you regularly run out of inodes, you should increase this value.

The file inode-nr contains the first two items from inode-state, so we'll skip to that file...

inode-state contains three actual numbers and four dummy values. The numbers are nr_inodes, nr_free_inodes, and preshrink (in order of appearance).

Denotes the number of inodes the system has allocated. This can be slightly more than inode-max because Linux allocates them one pageful at a time.

Represents the number of free inodes and preshrink is nonzero when nr_inodes is greater than inode-max and the system needs to prune the inode list instead of allocating more.

Again, super block structures are allocated by the kernel, but not freed. The file super-max contains the maximum number of super block handlers, where super-nr shows the number of currently allocated ones. Every mounted file system needs a super block, so if you plan to mount lots of file systems, you may want to increase these numbers.

This handles the kernel support for miscellaneous binary formats. binfmt_misc provides the ability to register additional binary formats to the kernel without compiling an additional module/kernel. Therefore, binfmt_misc needs to know magic numbers at the beginning or the filename extension of the binary. It works by maintaining a linked list of structs that contain a description of a binary format, including a magic with size (or the filename extension), offset and mask, and the interpreter name. On request it invokes the given interpreter with the original program as argument, as binfmt_java and binfmt_em86 and binfmt_mz do. Since binfmt_misc does not define any default binary-formats, you have to register an additional binary-format. There are two general files in binfmt_misc and one file per registered format. The two general files are register and status. To register a new binary format you have to issue the command echo :name:type:offset:magic:mask:interpreter: > /proc/sys/fs/binfmt_misc/register with appropriate name (the name for the /proc-dir entry), offset (defaults to 0, if omitted), magic, mask (which can be omitted, defaults to all 0xff) and last but not least, the interpreter that is to be invoked (for example and testing /bin/echo). Type can be M for usual magic matching or E for filename extension matching (give extension in place of magic). If you do a cat on the file /proc/sys/fs/binfmt_misc/status, you will get the current status (enabled/disabled) of binfmt_misc. Change the status by echoing 0 (disables) or 1 (enables) or -1 (caution: this clears all previously registered binary formats) to status. For example echo 0 > status to disable binfmt_misc (temporarily). Each registered handler has an entry in /proc/sys/fs/binfmt_misc. These files perform the same function as status, but their scope is limited to the actual binary format. By 'cating' this file, you also receive all related information about the interpreter/magic of the binfmt. An example of the usage of binfmt_misc (emulate binfmt_java) follows:

These four lines add support for Java executables and Java applets (like binfmt_java, additionally recognizing the .html extension with no need to put <!--applet> to every applet file). You have to install the JDK and the shell-script /usr/local/java/bin/javawrapper too. It works around the brokenness of the Java filename handling. To add a Java binary, just create a link to the class-file somewhere in the path.

The file contains three values; highwater, lowwater, and frequency. It exists only when BSD-style process accounting is enabled. These values control its behavior. If the free space on the file system where the log lives goes below lowwater percentage, accounting suspends. If it goes above highwater percentage, accounting resumes. Frequency determines how often you check the amount of free space (value is in seconds). Default settings are: 4, 2, and 30. That is, suspend accounting if there is less than 2 percent free; resume it if we have a value of 3 or more percent; consider information about the amount of free space valid for 30 seconds

When the value in this file is 0, ctrl-alt-del is trapped and sent to the init program to handle a graceful restart. However, when the value is greater that zero, Linux's reaction to this key combination will be an immediate reboot, without syncing its dirty buffers. It should be noted that when a program (like dosemu) has the keyboard in raw mode, the ctrl-alt-del is intercepted by the program before it ever reaches the kernel tty layer, and it is up to the program to decide what to do with it.

These files can be controlled to set the NIS domainname and hostname of your box. For the classic darkstar.frop.org a simple: # echo "darkstar" > /proc/sys/kernel/hostname # echo "frop.org" > /proc/sys/kernel/domainname would suffice to set your hostname and NIS domainname. /proc/sys/kernel/osrelease, /proc/sys/kernel/ostype, /proc/sys/kernel/version The names make it pretty obvious what these fields contain: # cat /proc/sys/kernel/osrelease 2.2.12 # cat /proc/sys/kernel/ostype Linux # cat /proc/sys/kernel/version #4 Fri Oct 1 12:41:14 PDT 1999 The files osrelease and ostype should be clear enough. Version needs a little more clarification. The #4 means that this is the 4th kernel built from this source base and the date after it indicates the time the kernel was built. The only way to tune these values is to rebuild the kernel.

The value in this file represents the number of seconds the kernel waits before rebooting on a panic. When you use the software watchdog, the recommended setting is 60. If set to 0, the auto reboot after a kernel panic is disabled, which is the default setting.

The four values in printk denote * console_loglevel, * default_message_loglevel, * minimum_console_level and * default_console_loglevel respectively. These values influence printk() behavior when printing or logging error messages, which come from inside the kernel. See syslog(2) for more information on the different log levels.

Messages with a higher priority than this will be printed to the console.

Messages without an explicit priority will be printed with this priority.

Minimum (highest) value to which the console_loglevel can be set.

Default value for console_loglevel.

This file shows the size of the generic SCSI (sg) buffer. At this point, you can't tune it yet, but you can change it at compile time by editing include/scsi/sg.h and changing the value of SG_BIG_BUFF. If you use a scanner with SANE (Scanner Access Now Easy) you might want to set this to a higher value. Refer to the SANE documentation on this issue.

The location where the modprobe binary is located. The kernel uses this program to load modules on demand.

The values are:

This is the minimum percentage of memory that should be spent on buffer memory.

When Linux is short on memory, and the buffer cache uses more than it has been allotted, the memory management (MM) subsystem will prune the buffer cache more heavily than other memory to compensate.

This is the maximum amount of memory that can be used for buffer memory.

This file contains three values: min, low and high:

When the number of free pages in the system reaches this number, only the kernel can allocate more memory.

If the number of free pages falls below this point, the kernel starts swapping aggressively.

The kernel tries to keep up to this amount of memory free; if memory falls below this point, the kernel starts gently swapping in the hopes that it never has to do really aggressive swapping.

The file contains three numbers:

The maximum number of pages kswapd tries to free in one round is calculated from this number. Usually this number will be divided by 4 or 8 (see mm/vmscan.c), so it isn't as big as it looks. When you need to increase the bandwidth to/from swap, you'll want to increase this number.

This is the minimum number of times kswapd tries to free a page each time it is called. Basically it's just there to make sure that kswapd frees some pages even when it's being called with minimum priority.

This is probably the greatest influence on system performance. swap_cluster is the number of pages kswapd writes in one turn. You'll want this value to be large so that kswapd does its I/O in large chunks and the disk doesn't have to seek as often, but you don't want it to be too large since that would flood the request queue.

This file contains no less than 8 variables. All of these values are used by kswapd. The first four variables sc_max_page_age, sc_page_advance, sc_page_decline and sc_page_initial_age are used to keep track of Linux's page aging. Page ageing is a bookkeeping method to track which pages of memory are often used, and which pages can be swapped out without consequences.

When a page is swapped in, it starts at sc_page_initial_age (default 3) and when the page is scanned by kswapd, its age is adjusted according to the following scheme.

If the page was used since the last time we scanned, its age is increased by sc_page_advance (default 3). Where the maximum value is given by sc_max_page_age (default 20). Otherwise (meaning it wasn't used) its age is decreased by sc_page_decline (default 1).

When a page reaches age 0, it's ready to be swapped out.

The variables sc_age_cluster_fract, sc_age_cluster_min, sc_pageout_weight and sc_bufferout_weight, can be used to control kswapd's aggressiveness in swapping out pages.

Sc_age_cluster_fract is used to calculate how many pages from a process are to be scanned by kswapd. The formula used is

(sc_age_cluster_fract divided by 1024) times resident set size

So if you want kswapd to scan the whole process, sc_age_cluster_fract needs to have a value of 1024. The minimum number of pages kswapd will scan is represented by sc_age_cluster_min, which is done so that kswapd will also scan small processes. The values of sc_pageout_weight and sc_bufferout_weight are used to control how many tries kswapd will make in order to swap out one page/buffer. These values can be used to fine-tune the ratio between user pages and buffer/cache memory. When you find that your Linux system is swapping out too many process pages in order to satisfy buffer memory demands, you may want to either increase sc_bufferout_weight, or decrease the value of sc_pageout_weight.

This directory contains four files, which enable or disable debugging for the RPC functions NFS, NFS-daemon, RPC and NLM. The default values are 0. They can be set to one to turn debugging on. (The default value is 0 for each)

The interface to the networking parts of the kernel is located in /proc/sys/net. The following table shows all possible subdirectories. You may see only some of them, depending on your kernel's configuration. Our main focus will be on IP networking since AX15, X.25, and DEC Net are only minor players in the Linux world. Should you wish review the online documentation and the kernel source to get a detailed view of the parameters for those protocols not covered here. In this section we'll discuss the subdirectories listed above. As default values are suitable for most needs, there is no need to change these values.

Network core options

The default setting of the socket receive buffer in bytes.

The maximum receive socket buffer size in bytes.

The default setting (in bytes) of the socket send buffer.

The maximum send socket buffer size in bytes.

These parameters are used to limit the warning messages written to the kernel log from the networking code. They enforce a rate limit to make a denial-of-service attack impossible. A higher message_cost factor, results in fewer messages that will be written. Message_burst controls when messages will be dropped. The default settings limit warning messages to one every five seconds.

Maximum number of packets, queued on the INPUT side, when the interface receives packets faster than kernel can process them.

Maximum ancillary buffer size allowed per socket. Ancillary data is a sequence of struct cmsghdr structures with appended data.

Parameters for Unix domain sockets

There are only two files in this subdirectory. They control the delays for deleting and destroying socket descriptors.

IPV4 settings. IP version 4 is still the most used protocol in Unix networking. It will be replaced by IP version 6 in the next couple of years, but for the moment it's the de facto standard for the internet and is used in most networking environments around the world. Because of the importance of this protocol, we'll have a deeper look into the subtree controlling the behavior of the Ipv4 subsystem of the Linux kernel.

Let's start with the entries in /proc/sys/net/ipv4.

Turn on (1) or off (0), if the kernel should ignore all ICMP ECHO requests, or just those to broadcast and multicast addresses.

Please note that if you accept ICMP echo requests with a broadcast/multi\-cast destination address your network may be used as an exploder for denial of service packet flooding attacks to other hosts.

Sets limits for sending ICMP packets to specific targets. A value of zero disables all limiting. Any positive value sets the maximum package rate in hundredth of a second (on Intel systems).

This file contains the number one if the host received its IP configuration by RARP, BOOTP, DHCP or a similar mechanism. Otherwise it is zero.

TTL (Time To Live) for IPv4 interfaces. This is simply the maximum number of hops a packet may travel.

Enable dynamic socket address rewriting on interface address change. This is useful for dialup interface with changing IP addresses.

Enable or disable forwarding of IP packages between interfaces. Changing this value resets all other parameters to their default values. They differ if the kernel is configured as host or router.

Range of ports used by TCP and UDP to choose the local port. Contains two numbers, the first number is the lowest port, the second number the highest local port. Default is 1024-4999. Should be changed to 32768-61000 for high-usage systems.

Global switch to turn path MTU discovery off. It can also be set on a per socket basis by the applications or on a per route basis.

Enable/disable debugging of IP masquerading.

Maximum memory used to reassemble IP fragments. When ipfrag_high_thrash bytes of memory is allocated for this purpose, the fragment handler will toss packets until ipfrag_low_thrash is reached.

Time in seconds to keep an IP fragment in memory.

This file controls the use of the ECN bit in the IPv4 headers, this is a new feature about Explicit Congestion Notification, but some routers and firewalls block traffic that has this bit set, so it could be necessary to echo 0 to /proc/sys/net/ipv4/tcp_ecn, if you want to talk to this sites. For more info you could read RFC2481.

Bug-to-bug compatibility with some broken printers. On retransmit, try to send larger packets to work around bugs in certain TCP stacks. Can be turned off by setting it to zero.

Number of keep alive probes TCP sends out, until it decides that the connection is broken.

How often TCP sends out keep alive messages, when keep alive is enabled. The default is 2 hours.

Number of times initial SYNs for a TCP connection attempt will be retransmitted. Should not be higher than 255. This is only the timeout for outgoing connections, for incoming connections the number of retransmits is defined by tcp_retries1.

Enable select acknowledgments after RFC2018.

Enable timestamps as defined in RFC1323.

Enable the strict RFC793 interpretation of the TCP urgent pointer field. The default is to use the BSD compatible interpretation of the urgent pointer pointing to the first byte after the urgent data. The RFC793 interpretation is to have it point to the last byte of urgent data. Enabling this option may lead to interoperability problems. Disabled by default.

Only valid when the kernel was compiled with CONFIG_SYNCOOKIES. Send out syncookies when the syn backlog queue of a socket overflows. This is to ward off the common 'syn flood attack'. Disabled by default. Note that the concept of a socket backlog is abandoned. This means the peer may not receive reliable error messages from an over loaded server with syncookies enabled.

Enable window scaling as defined in RFC1323.

The length of time in seconds it takes to receive a final FIN before the socket is always closed. This is strictly a violation of the TCP specification, but required to prevent denial-of-service attacks.

Indicates how many keep alive probes are sent per slow timer run. Should not be set too high to prevent bursts.

Length of the per socket backlog queue. Since Linux 2.2 the backlog specified in listen(2) only specifies the length of the backlog queue of already established sockets. When more connection requests arrive Linux starts to drop packets. When syncookies are enabled the packets are still answered and the maximum queue is effectively ignored.

Defines how often an answer to a TCP connection request is retransmitted before giving up.

Defines how often a TCP packet is retransmitted before giving up.

Here you'll find one subdirectory for each interface the system knows about and one directory called all. Changes in the all subdirectory affect all interfaces, whereas changes in the other subdirectories affect only one interface. All directories have the same entries:

This switch decides if the kernel accepts ICMP redirect messages or not. The default is 'yes' if the kernel is configured for a regular host and 'no' for a router configuration.

Should source routed packages be accepted or declined. The default is dependent on the kernel configuration. It's 'yes' for routers and 'no' for hosts.

Accept packets with source address 0.b.c.d with destinations not to this host as local ones. It is supposed that a BOOTP relay daemon will catch and forward such packets. The default is 0.

Enable or disable IP forwarding on this interface.

Log packets with source addresses with no known route to kernel log.

Do multicast routing. The kernel needs to be compiled with CONFIG_MROUTE and a multicast routing daemon is required.

Does (1) or does not (0) perform proxy ARP.

Integer value determines if a source validation should be made. 1 means yes, 0 means no. Disabled by default, but local/broadcast address spoofing is always on. If you set this to 1 on a router that is the only connection for a network to the net, it will prevent spoofing attacks against your internal networks (external addresses can still be spoofed), without the need for additional firewall rules.

Accept ICMP redirect messages only for gateways, listed in default gateway list. Enabled by default.

If it is not set the kernel does not assume that different subnets on this device can communicate directly. Default setting is 'yes'.

Determines whether to send ICMP redirects to other hosts.

The directory /proc/sys/net/ipv4/route contains several file to control routing issues.

These parameters are used to limit the warning messages written to the kernel log from the routing code. The higher the error_cost factor is, the fewer messages will be written. Error_burst controls when messages will be dropped. The default settings limit warning messages to one every five seconds.

Writing to this file results in a flush of the routing cache.

Values to control the frequency and behavior of the garbage collection algorithm for the routing cache.

Maximum size of the routing cache. Old entries will be purged once the cache reached has this size.

Delays for flushing the routing cache.

Factors which determine if more ICPM redirects should be sent to a specific host. No redirects will be sent once the load limit or the maximum number of redirects has been reached.

Timeout for redirects. After this period redirects will be sent again, even if this has been stopped, because the load or number limit has been reached.

Network Neighbor handling. It contains settings about how to handle connections with direct neighbors (nodes attached to the same link). As we saw it in the conf directory, there is a default subdirectory which holds the default values, and one directory for each interface. The contents of the directories are identical, with the single exception that the default settings contain additional options to set garbage collection parameters.

In the interface directories you'll find the following entries:

A base value used for computing the random reachable time value as specified in RFC2461.

The time, expressed in jiffies (1/100 sec), between retransmitted Neighbor Solicitation messages. Used for address resolution and to determine if a neighbor is unreachable.

Maximum queue length for a pending arp request - the number of packets which are accepted from other layers while the ARP address is still resolved.

Maximum for random delay of answers to neighbor solicitation messages in jiffies (1/100 sec). Not yet implemented (Linux does not have anycast support yet).

Maximum number of retries for unicast solicitation.

Maximum number of retries for multicast solicitation.

Delay for the first time probe if the neighbor is reachable. (see gc_stale_time)

An ARP/neighbor entry is only replaced with a new one if the old is at least locktime old. This prevents ARP cache thrashing.

Maximum time (real time is random [0..proxytime]) before answering to an ARP request for which we have an proxy ARP entry. In some cases, this is used to prevent network flooding.

Maximum queue length of the delayed proxy arp timer. (see proxy_delay).

Determines the number of requests to send to the user level ARP daemon. Use 0 to turn off.

Determines how often to check for stale ARP entries. After an ARP entry is stale it will be resolved again (which is useful when an IP address migrates to another machine). When ucast_solicit is greater than 0 it first tries to send an ARP packet directly to the known host When that fails and mcast_solicit is greater than 0, an ARP request is broadcasted.

Holds the Appletalk configuration data when Appletalk is loaded. The configurable parameters are:

The amount of time we keep an ARP entry before expiring it. Used to age out old hosts.

The amount of time we will spend trying to resolve an Appletalk address.

The number of times we will retransmit a query before giving up.

Controls the rate at which expires are checked.

Holds the list of active Appletalk sockets on a machine. The fields indicate the DDP type, the local address (in network:node format) the remote address, the size of the transmit pending queue, the size of the received queue (bytes waiting for applications to read) the state and the uid owning the socket.

lists all the interfaces configured for appletalk. It shows the name of the interface, its Appletalk address, the network range on that address (or network number for phase 1 networks), and the status of the interface.

lists each known network route. It lists the target (network) that the route leads to, the router (may be directly connected), the route flags, and the device the route is using.

The IPX protocol has no tunable values in proc/sys/net, it does, however, provide proc/net/ipx. This lists each IPX socket giving the local and remote addresses in Novell format (that is network:node:port). In accordance with the strange Novell tradition, everything but the port is in hex. Not_Connected is displayed for sockets that are not tied to a specific remote address. The Tx and Rx queue sizes indicate the number of bytes pending for transmission and reception. The state indicates the state the socket is in and the uid is the owning uid of the socket.

Lists all IPX interfaces. For each interface it gives the network number, the node number, and indicates if the network is the primary network. It also indicates which device it is bound to (or Internal for internal networks) and the Frame Type if appropriate. Linux supports 802.3, 802.2, 802.2 SNAP and DIX (Blue Book) ethernet framing for IPX.

Table holding a list of IPX routes. For each route it gives the destination network, the router node (or Directly) and the network address of the router (or Connected) for internal networks.

To see which tty's are currently in use, you can simply look into the file /proc/tty/drivers:

Note that while the above files tend to be easily readable text files, they can sometimes be formatted in a way that is not easily digestible. There are many commands that do little more than read the above files and format them for easier understanding. For example, the free program reads /proc/meminfo and converts the amounts given in bytes to kilobytes (and adds a little more information, as well).