9075 Improve ZFS pool import/load process and corrupted pool recovery

[unleashed.git] / usr / src / man / man7p / tcp.7p

'\" te
.\" Copyright (c) 2006, Sun Microsystems, Inc. All Rights Reserved.
.\" Copyright (c) 2011 Nexenta Systems, Inc. All rights reserved.
.\" Copyright 1989 AT&T
.\" The contents of this file are subject to the terms of the Common Development and Distribution License (the "License").  You may not use this file except in compliance with the License.
.\" You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE or http://www.opensolaris.org/os/licensing.  See the License for the specific language governing permissions and limitations under the License.
.\" When distributing Covered Code, include this CDDL HEADER in each file and include the License file at usr/src/OPENSOLARIS.LICENSE.  If applicable, add the following below this CDDL HEADER, with the fields enclosed by brackets "[]" replaced with your own identifying information: Portions Copyright [yyyy] [name of copyright owner]
.TH TCP 7P "Feb 5, 2018"
.SH NAME
tcp, TCP \- Internet Transmission Control Protocol
.SH SYNOPSIS
.LP
.nf
\fB#include <sys/socket.h>\fR
.fi

.LP
.nf
\fB#include <netinet/in.h>\fR
.fi

.LP
.nf
\fBs = socket(AF_INET, SOCK_STREAM, 0);\fR
.fi

.LP
.nf
\fBs = socket(AF_INET6, SOCK_STREAM, 0);\fR
.fi

.LP
.nf
\fBt = t_open("/dev/tcp", O_RDWR);\fR
.fi

.LP
.nf
\fBt = t_open("/dev/tcp6", O_RDWR);\fR
.fi

.SH DESCRIPTION
.LP
\fBTCP\fR is the virtual circuit protocol of the Internet protocol family. It
provides reliable, flow-controlled, in order, two-way transmission of data. It
is a byte-stream protocol layered above the Internet Protocol (\fBIP\fR), or
the Internet Protocol Version 6 (\fBIPv6\fR), the Internet protocol family's
internetwork datagram delivery protocol.
.sp
.LP
Programs can access \fBTCP\fR using the socket interface as a \fBSOCK_STREAM\fR
socket type, or using the Transport Level Interface (\fBTLI\fR) where it
supports the connection-oriented (\fBT_COTS_ORD\fR) service type.
.sp
.LP
\fBTCP\fR uses \fBIP\fR's host-level addressing and adds its own per-host
collection of "port addresses." The endpoints of a \fBTCP\fR connection are
identified by the combination of an \fBIP\fR or IPv6 address and a \fBTCP\fR
port number. Although other protocols, such as the User Datagram Protocol
(UDP), may use the same host and port address format, the port space of these
protocols is distinct. See \fBinet\fR(7P) and \fBinet6\fR(7P) for details on
the common aspects of addressing in the Internet protocol family.
.sp
.LP
Sockets utilizing \fBTCP\fR are either "active" or "passive." Active sockets
initiate connections to passive sockets. Both types of sockets must have their
local \fBIP\fR or IPv6 address and \fBTCP\fR port number bound with the
\fBbind\fR(3SOCKET) system call after the socket is created. By default,
\fBTCP\fR sockets are active. A passive socket is created by calling the
\fBlisten\fR(3SOCKET) system call after binding the socket with \fBbind()\fR.
This establishes a queueing parameter for the passive socket. After this,
connections to the passive socket can be received with the
\fBaccept\fR(3SOCKET) system call. Active sockets use the
\fBconnect\fR(3SOCKET) call after binding to initiate connections.
.sp
.LP
By using the special value \fBINADDR_ANY\fR with \fBIP\fR, or the unspecified
address (all zeroes) with IPv6, the local \fBIP\fR address can be left
unspecified in the \fBbind()\fR call by either active or passive \fBTCP\fR
sockets. This feature is usually used if the local address is either unknown or
irrelevant. If left unspecified, the local \fBIP\fR or IPv6 address will be
bound at connection time to the address of the network interface used to
service the connection.
.sp
.LP
Note that no two TCP sockets can be bound to the same port unless the bound IP
addresses are different.  IPv4 \fBINADDR_ANY\fR and IPv6 unspecified addresses
compare as equal to any IPv4 or IPv6 address. For example, if a socket is bound
to \fBINADDR_ANY\fR or unspecified address and port X, no other socket can bind
to port X, regardless of the binding address. This special consideration of
\fBINADDR_ANY\fR and unspecified address can be changed using the socket option
\fBSO_REUSEADDR\fR. If \fBSO_REUSEADDR\fR is set on a socket doing a bind, IPv4
\fBINADDR_ANY\fR and IPv6 unspecified address do not compare as equal to any IP
address. This means that as long as the two sockets are not both bound to
\fBINADDR_ANY\fR/unspecified address or the same IP address, the two sockets
can be bound to the same port.
.sp
.LP
If an application does not want to allow another socket  using the
\fBSO_REUSEADDR\fR option to bind to a port its socket is bound to, the
application can set the socket level option \fBSO_EXCLBIND\fR on a socket. The
option values of 0 and 1 mean enabling and disabling the option respectively.
Once this option is enabled on a socket, no other socket can be bound to the
same port.
.sp
.LP
Once a connection has been established, data can be exchanged using the
\fBread\fR(2) and \fBwrite\fR(2) system calls.
.sp
.LP
Under most circumstances, \fBTCP\fR sends data when it is presented. When
outstanding data has not yet been acknowledged, \fBTCP\fR gathers small amounts
of output to be sent in a single packet once an acknowledgement has been
received. For a small number of clients, such as window systems that send a
stream of mouse events which receive no replies, this packetization may cause
significant delays. To circumvent this problem, \fBTCP\fR provides a
socket-level boolean option, \fBTCP_NODELAY.\fR \fBTCP_NODELAY\fR is defined in
\fB<netinet/tcp.h>\fR, and is set with \fBsetsockopt\fR(3SOCKET) and tested
with \fBgetsockopt\fR(3SOCKET). The option level for the \fBsetsockopt()\fR
call is the protocol number for \fBTCP,\fR available from
\fBgetprotobyname\fR(3SOCKET).
.sp
.LP
For some applications, it may be desirable for TCP not to send out data unless
a full TCP segment can be sent. To enable this behavior, an application can use
the \fBTCP_CORK\fR socket option. When \fBTCP_CORK\fR is set with a non-zero
value, TCP sends out a full TCP segment only. When \fBTCP_CORK\fR is set to
zero after it has been enabled, all buffered data is sent out (as permitted by
the peer's receive window and the current congestion window). \fBTCP_CORK\fR is
defined in <\fBnetinet/tcp.h\fR>,  and is set with \fBsetsockopt\fR(3SOCKET)
and tested with \fBgetsockopt\fR(3SOCKET). The option level for the
\fBsetsockopt()\fR call is the protocol  number  for TCP, available from
\fBgetprotobyname\fR(3SOCKET).
.sp
.LP
The \fBSO_RCVBUF\fR socket level option can be used to control the window
that \fBTCP\fR advertises to the peer. \fBIP\fR level options may also be used
with \fBTCP.\fR See \fBip\fR(7P) and \fBip6\fR(7P).
.sp
.LP
\fBTCP\fR provides an urgent data mechanism, which may be invoked using the
out-of-band provisions of \fBsend\fR(3SOCKET). The caller may mark one byte as
"urgent" with the \fBMSG_OOB\fR flag to \fBsend\fR(3SOCKET). This sets an
"urgent pointer" pointing to this byte in the \fBTCP\fR stream. The receiver on
the other side of the stream is notified of the urgent data by a \fBSIGURG\fR
signal. The \fBSIOCATMARK\fR \fBioctl\fR(2) request returns a value indicating
whether the stream is at the urgent mark. Because the system never returns data
across the urgent mark in a single \fBread\fR(2) call, it is possible to
advance to the urgent data in a simple loop which reads data, testing the
socket with the \fBSIOCATMARK\fR \fBioctl()\fR request, until it reaches the
mark.
.sp
.LP
Incoming connection requests that include an \fBIP\fR source route option are
noted, and the reverse source route is used in responding.
.sp
.LP
A checksum over all data helps \fBTCP\fR implement reliability. Using a
window-based flow control mechanism that makes use of positive
acknowledgements, sequence numbers, and a retransmission strategy, \fBTCP\fR
can usually recover when datagrams are damaged, delayed, duplicated or
delivered out of order by the underlying communication medium.
.sp
.LP
If the local \fBTCP\fR receives no acknowledgements from its peer for a period
of time, (for example, if the remote machine crashes), the connection is closed
and an error is returned.
.sp
.LP
TCP follows the congestion control algorithm described in \fIRFC 2581\fR, and
also supports the initial congestion window (cwnd) changes in \fIRFC 3390\fR.
The initial cwnd calculation can be overridden by the socket option
TCP_INIT_CWND. An application can use this option to set the initial cwnd to a
specified number of TCP segments. This applies to the cases when the connection
first starts and restarts after an idle period.  The process must have the
PRIV_SYS_NET_CONFIG privilege if it wants to specify a number greater than that
calculated by \fIRFC 3390\fR.
.sp
.LP
illumos supports \fBTCP\fR Extensions for High Performance (\fIRFC 7323\fR)
which includes the window scale and timestamp options, and Protection Against
Wrap Around Sequence Numbers (PAWS). Note that if timestamps are negotiated on
a connection, received segments without timestamps on that connection are
silently dropped per the suggestion in the RFC. illumos also supports Selective
Acknowledgment (SACK) capabilities (RFC 2018) and Explicit Congestion
Notification (ECN) mechanism (\fIRFC 3168\fR).
.sp
.LP
Turn on the window scale option in one of the following ways:
.RS +4
.TP
.ie t \(bu
.el o
An application can set \fBSO_SNDBUF\fR or \fBSO_RCVBUF\fR size in the
\fBsetsockopt()\fR option to be larger than 64K. This must be done \fIbefore\fR
the program calls \fBlisten()\fR or \fBconnect()\fR, because the window scale
option is negotiated when the connection is established. Once the connection
has been made, it is too late to increase the send or receive window beyond the
default \fBTCP\fR limit of 64K.
.RE
.RS +4
.TP
.ie t \(bu
.el o
For all applications, use \fBndd\fR(1M) to modify the configuration parameter
\fBtcp_wscale_always\fR. If \fBtcp_wscale_always\fR is set to \fB1\fR, the
window scale option will always be set when connecting to a remote system. If
\fBtcp_wscale_always\fR is \fB0,\fR the window scale option will be set only if
the user has requested a send or receive window larger than 64K. The default
value of \fBtcp_wscale_always\fR is \fB1\fR.
.RE
.RS +4
.TP
.ie t \(bu
.el o
Regardless of the value of \fBtcp_wscale_always\fR, the window scale option
will always be included in a connect acknowledgement if the connecting system
has used the option.
.RE
.sp
.LP
Turn on \fBSACK\fR capabilities in the following way:
.RS +4
.TP
.ie t \(bu
.el o
Use \fBndd\fR to modify the configuration parameter \fBtcp_sack_permitted\fR.
If \fBtcp_sack_permitted\fR is set to \fB0\fR, \fBTCP\fR will not accept
\fBSACK\fR or send out \fBSACK\fR information. If \fBtcp_sack_permitted\fR is
set to \fB1\fR, \fBTCP\fR will not initiate a connection with \fBSACK\fR
permitted option in the \fBSYN\fR segment, but will respond with \fBSACK\fR
permitted option in the \fBSYN|ACK\fR segment if an incoming connection request
has the \fBSACK \fR permitted option. This means that \fBTCP\fR will only
accept \fBSACK\fR information if the other side of the connection also accepts
\fBSACK\fR information. If \fBtcp_sack_permitted\fR is set to \fB2\fR, it will
both initiate and accept connections with \fBSACK\fR information. The default
for \fBtcp_sack_permitted\fR is \fB2\fR (active enabled).
.RE
.sp
.LP
Turn on \fBTCP ECN\fR mechanism in the following way:
.RS +4
.TP
.ie t \(bu
.el o
Use \fBndd\fR to modify the configuration parameter \fBtcp_ecn_permitted\fR. If
\fBtcp_ecn_permitted\fR is set to \fB0\fR, \fBTCP\fR will not negotiate with a
peer that supports \fBECN\fR mechanism. If \fBtcp_ecn_permitted\fR is set to
\fB1\fR when initiating a connection, TCP will not tell a peer that it supports
ECN mechanism. However, it will tell a peer that it supports \fBECN\fR
mechanism when accepting a new incoming connection request if the peer
indicates that it supports \fBECN\fR mechanism in the SYN segment. If
tcp_ecn_permitted is set to 2, in addition to negotiating with a peer on ECN
mechanism when accepting connections, TCP will indicate in the outgoing SYN
segment that it supports \fBECN\fR mechanism when \fBTCP\fR makes active
outgoing connections. The default for \fBtcp_ecn_permitted\fR is 1.
.RE
.sp
.LP
Turn on the timestamp option in the following way:
.RS +4
.TP
.ie t \(bu
.el o
Use \fBndd\fR to modify the configuration parameter \fBtcp_tstamp_always\fR. If
\fBtcp_tstamp_always\fR is \fB1\fR, the timestamp option will always be set
when connecting to a remote machine. If \fBtcp_tstamp_always\fR is \fB0\fR, the
timestamp option will not be set when connecting to a remote system. The
default for \fBtcp_tstamp_always\fR is \fB0\fR.
.RE
.RS +4
.TP
.ie t \(bu
.el o
Regardless of the value of \fBtcp_tstamp_always\fR, the timestamp option will
always be included in a connect acknowledgement (and all succeeding packets) if
the connecting system has used the timestamp option.
.RE
.sp
.LP
Use the following procedure to turn on the timestamp option only when the
window scale option is in effect:
.RS +4
.TP
.ie t \(bu
.el o
Use \fBndd\fR to modify the configuration parameter \fBtcp_tstamp_if_wscale\fR.
Setting \fBtcp_tstamp_if_wscale\fR to \fB1\fR will cause the timestamp option
to be set when connecting to a remote system, if the window scale option has
been set. If \fBtcp_tstamp_if_wscale\fR is \fB0\fR, the timestamp option will
not be set when connecting to a remote system. The default for
\fBtcp_tstamp_if_wscale\fR is \fB1\fR.
.RE
.sp
.LP
Protection Against Wrap Around Sequence Numbers (PAWS) is always used when the
timestamp option is set.
.sp
.LP
SunOS also supports multiple methods of generating initial sequence numbers.
One of these methods is the improved technique suggested in \fBRFC\fR 1948. We
\fBHIGHLY\fR recommend that you set sequence number generation parameters as
close to boot time as possible. This prevents sequence number problems on
connections that use the same connection-ID as ones that used a different
sequence number generation. The \fBsvc:/network/initial:default\fR service
configures the initial sequence number generation.  The service reads the value
contained in the configuration file \fB/etc/default/inetinit\fR to determine
which method to use.
.sp
.LP
The \fB/etc/default/inetinit\fR file is an unstable interface, and may change
in future releases.
.sp
.LP
\fBTCP\fR may be configured to report some information on connections that
terminate by means of an \fBRST\fR packet. By default, no logging is done. If
the \fBndd\fR(1M) parameter \fItcp_trace\fR is set to 1, then trace data is
collected for all new connections established after that time.
.sp
.LP
The trace data consists of the \fBTCP\fR headers and \fBIP\fR source and
destination addresses of the last few packets sent in each direction before RST
occurred. Those packets are logged in a series of \fBstrlog\fR(9F) calls. This
trace facility has a very low overhead, and so is superior to such utilities as
\fBsnoop\fR(1M) for non-intrusive debugging for connections terminating by
means of an \fBRST\fR.
.sp
.LP
SunOS supports the keep-alive mechanism described in \fIRFC 1122\fR. It is
enabled using the socket option SO_KEEPALIVE. When enabled, the first
keep-alive probe is sent out after a TCP is idle for two hours If the peer does
not respond to the probe within eight minutes, the TCP connection is aborted.
You can alter the interval for sending out the first probe using the socket
option TCP_KEEPALIVE_THRESHOLD. The option value is an unsigned integer in
milliseconds. The system default is controlled by the TCP ndd parameter
tcp_keepalive_interval. The minimum value is ten seconds. The maximum is ten
days, while the default is two hours. If you receive no response to the probe,
you can use the TCP_KEEPALIVE_ABORT_THRESHOLD socket option to change the time
threshold for aborting a TCP connection. The option value is an unsigned
integer in milliseconds. The value zero indicates that TCP should never time
out and abort the connection when probing. The system default is controlled by
the TCP ndd parameter tcp_keepalive_abort_interval. The default is eight
minutes.
.sp
.LP
socket options TCP_KEEPIDLE, TCP_KEEPCNT and TCP_KEEPINTVL are also supported
for compatibility with other Unix Flavors. TCP_KEEPIDLE option specifies the
interval in seconds for sending out the first keep-alive probe. TCP_KEEPCNT
specifies the number of keep-alive probes to be sent before aborting the
connection in the event of no response from peer. TCP_KEEPINTVL specifies the
interval in seconds between successive keep-alive probes.
.SH SEE ALSO
.LP
\fBsvcs\fR(1), \fBndd\fR(1M), \fBioctl\fR(2), \fBread\fR(2), \fBsvcadm\fR(1M),
\fBwrite\fR(2), \fBaccept\fR(3SOCKET), \fBbind\fR(3SOCKET),
\fBconnect\fR(3SOCKET), \fBgetprotobyname\fR(3SOCKET),
\fBgetsockopt\fR(3SOCKET), \fBlisten\fR(3SOCKET), \fBsend\fR(3SOCKET),
\fBsmf\fR(5), \fBinet\fR(7P), \fBinet6\fR(7P), \fBip\fR(7P), \fBip6\fR(7P)
.sp
.LP
Ramakrishnan, K., Floyd, S., Black, D., RFC 3168, \fIThe Addition of Explicit
Congestion Notification (ECN) to IP\fR, September 2001.
.sp
.LP
Mathias, M. and Hahdavi, J. Pittsburgh Supercomputing Center; Ford, S. Lawrence
Berkeley National Laboratory; Romanow, A. Sun Microsystems, Inc. \fIRFC 2018,
TCP Selective Acknowledgement Options\fR, October 1996.
.sp
.LP
Bellovin, S., \fIRFC 1948, Defending Against Sequence Number Attacks\fR, May
1996.
.sp
.LP
D. Borman, B. Braden, V. Jacobson and R. Scheffenegger, Ed., \fIRFC 7323, TCP
Extensions for High Performance\fR, September 2014.
.sp
.LP
Postel, Jon, \fIRFC 793, Transmission Control Protocol - DARPA Internet Program
Protocol Specification\fR, Network Information Center, SRI International, Menlo
Park, CA., September 1981.
.SH DIAGNOSTICS
.LP
A socket operation may fail if:
.sp
.ne 2
.na
\fB\fBEISCONN\fR\fR
.ad
.RS 17n
A \fBconnect()\fR operation was attempted on a socket on which a
\fBconnect()\fR operation had already been performed.
.RE

.sp
.ne 2
.na
\fB\fBETIMEDOUT\fR\fR
.ad
.RS 17n
A connection was dropped due to excessive retransmissions.
.RE

.sp
.ne 2
.na
\fB\fBECONNRESET\fR\fR
.ad
.RS 17n
The remote peer forced the connection to be closed (usually because the remote
machine has lost state information about the connection due to a crash).
.RE

.sp
.ne 2
.na
\fB\fBECONNREFUSED\fR\fR
.ad
.RS 17n
The remote peer actively refused connection establishment (usually because no
process is listening to the port).
.RE

.sp
.ne 2
.na
\fB\fBEADDRINUSE\fR\fR
.ad
.RS 17n
A \fBbind()\fR operation was attempted on a socket with a network address/port
pair that has already been bound to another socket.
.RE

.sp
.ne 2
.na
\fB\fBEADDRNOTAVAIL\fR\fR
.ad
.RS 17n
A \fBbind()\fR operation was attempted on a socket with a network address for
which no network interface exists.
.RE

.sp
.ne 2
.na
\fB\fBEACCES\fR\fR
.ad
.RS 17n
A \fBbind()\fR operation was attempted with a "reserved" port number and the
effective user \fBID\fR of the process was not the privileged user.
.RE

.sp
.ne 2
.na
\fB\fBENOBUFS\fR\fR
.ad
.RS 17n
The system ran out of memory for internal data structures.
.RE

.SH NOTES
.LP
The \fBtcp\fR service is managed by the service management facility,
\fBsmf\fR(5), under the service identifier:
.sp
.in +2
.nf
svc:/network/initial:default
.fi
.in -2
.sp

.sp
.LP
Administrative actions on this service, such as enabling, disabling, or
requesting restart, can be performed using \fBsvcadm\fR(1M). The service's
status can be queried using the \fBsvcs\fR(1) command.