wdiff rfc7440.alt-original rfc7440.txt

INTERNET-DRAFT Patrick
Internet Engineering Task Force (IETF) P. Masotta
Intended status: Standard Track
Request for Comments: 7440 Serva
Expires: Apr 16,
Category: Standards Track January 2015 Oct 16, 2014
ISSN: 2070-1721

TFTP Windowsize Option
draft-masotta-tftpexts-windowsize-opt-13.txt

Abstract

The Trivial File Transfer Protocol (RFC1350) (RFC 1350) is a simple, lock-step, lockstep,
file transfer protocol which that allows a client to get or put a file onto
a remote host. One of its primary uses is in the early stages of
nodes booting from a Local Area Network. TFTP has been used for this
application because it is very simple to implement. The Employment employment
of a lock-step lockstep scheme limits throughput when used on a LAN.

This document describes a TFTP option which that allows the client and
server to negotiate a window size of consecutive blocks to send as an
alternative for replacing the single block lock-step single-block lockstep schema. The TFTP
option mechanism employed is described in TFTP "TFTP Option Extension
(RFC2347). Extension"
(RFC 2347).

Status of this This Memo

This Internet-Draft is submitted in full conformance with the
provisions of BCP 78 and BCP 79.

Internet-Drafts are working documents an Internet Standards Track document.

This document is a product of the Internet Engineering Task Force (IETF), its areas, and its working groups. Note that
other groups may also distribute working documents as Internet-
Drafts.

Internet-Drafts are draft documents valid for a maximum
(IETF). It represents the consensus of six months the IETF community. It has
received public review and may be updated, replaced, or obsoleted has been approved for publication by other documents at any
time. It the
Internet Engineering Steering Group (IESG). Further information on
Internet Standards is inappropriate to use Internet-Drafts as reference
material or to cite them other than as "work available in progress."

The list Section 2 of RFC 5741.

Information about the current Internet-Drafts can be accessed at
http://www.ietf.org/ietf/1id-abstracts.txt

The list status of Internet-Draft Shadow Directories can this document, any errata,
and how to provide feedback on it may be accessed obtained at
http://www.ietf.org/shadow.html

This Internet-Draft will expire on Feb 16, 2014.
http://www.rfc-editor.org/info/rfc7440.

This document is subject to BCP 78 and the IETF Trust's Legal
Provisions Relating to IETF Documents
(http://trustee.ietf.org/license-info) in effect on the date of
publication of this document. Please review these documents
carefully, as they describe your rights and restrictions with respect
to this document. Code Components extracted from this document must
include Simplified BSD License text as described in Section 4.e of
the Trust Legal Provisions and are provided without warranty as
described in the Simplified BSD License.

Legal

This documents and the information contained therein are provided on
an "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE
REPRESENTS OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY, THE
IETF TRUST AND THE INTERNET ENGINEERING TASK FORCE DISCLAIM ALL
WARRANTIES, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY
WARRANTY THAT THE USE OF THE INFORMATION THEREIN WILL NOT INFRINGE
ANY RIGHTS OR ANY IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS
FOR A PARTICULAR PURPOSE.

Table of Contents

1. Introduction...................................................3 Introduction ....................................................2
2. Conventions used Used in this document..............................3 This Document ...............................3
3. Windowsize Option Specification................................3 Specification .................................3
4. Traffic Flow and Error Handling................................5 Handling .................................4
5. Proof of Concept and Windowsize Evaluation.....................6 Evaluation ......................6
6. Congestion and Error Control...................................8 Control ....................................7
7. Security Considerations........................................9 Considerations .........................................8
8. IANA Considerations............................................9
9. References.....................................................9
9.1. References ......................................................9
8.1. Normative References......................................9 References .......................................9
Author's Address ...................................................9

1. Introduction

TFTP is virtually unused for internet Internet transfers today, TFTP is still
massively used in network boot/installation scenarios including EFI
(Extensible Firmware Interface). The TFTP protocol's TFTP's inherently low transfer rate
has been been, so far far, partially mitigated by the use of the blocksize
negotiated extension [RFC2348]. This way Using this method, the original
limitation of 512 byte 512-byte blocks are are, in practice practice, replaced in Ethernet
environments by blocks no larger than 1468 Bytes to avoid IP block
fragmentation. This strategy produces insufficient results when
transferring big files, for example example, the initial ramdisk of Linux
distributions or the PE images used in network installations by
Microsoft WDS/MDT/SCCM. Considering TFTP looks today far from extinction
today, this draft document presents a negotiated extension, under the terms
of the TFTP "TFTP Option Extension Extension" [RFC2347], that produces TFTP transfer
rates comparable to those achieved today by modern file transfer protocols.

2. Conventions used Used in this document This Document

The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
"SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and
"OPTIONAL" in this document are to be interpreted as described in
RFC-2119 RFC
2119 [RFC2119].

In this document, these words will appear with that interpretation
only when in ALL CAPS. Lower case Lowercase uses of these words are not to be
interpreted as carrying RFC-2119 significance. the significance given in RFC 2119.

3. Windowsize Option Specification

The TFTP Read Request or Write Request packet is modified to include
the windowsize option as follows. Note that all fields except "opc"
MUST be ASCII strings followed by a single-byte NULL character.

2B string 1B string 1B string 1B string 1B
+-------+---~~---+----+---~~---+----+-----~~-----+----+---~~---+----+
| opc |filename| 0 | mode | 0 | windowsize | 0 | #blocks| 0 |
+-------+---~~---+----+---~~---+----+-----~~-----+----+---~~---+----+

opc
The opcode field either contains either a 1, 1 for Read Requests, Requests or 2, a 2 for
Write Requests, as defined in [RFC1350].

filename
The name of the file to be read or written, as defined in
[RFC1350].

mode
The mode of the file transfer: "netascii", "octet", or "mail", as
defined in [RFC1350].

windowsize
The Windowsize windowsize option, "windowsize" (case in-sensitive). insensitive).

#blocks
The base-10 ASCII string representation of the number of blocks in
a window. Valid The valid values range MUST be between "1" 1 and "65535" 65535
blocks, inclusive. The windowsize refers to the number of consecutives
consecutive blocks transmitted before stop stopping and wait waiting for the
reception of the acknowledgment of the last block transmitted.

For example:

+------+--------+----+-------+----+------------+----+----+----+
|0x0001| foobar |0x00| octet |0x00| windowsize |0x00| 16 |0x00|
+------+--------+----+-------+----+------------+----+----+----+

is a Read Request, Request for the file named "foobar", "foobar" in octet transfer
mode, mode
with a window-size windowsize of 16 blocks (option blocksize is not negotiated in
this example, the default of 512 Bytes per block default applies).

If the server is willing to accept the windowsize option, it sends an
Option Acknowledgment (OACK) to the client. The specified value MUST
be less than or equal to the value specified by the client. The
client MUST then either use the size specified in the OACK, OACK or send an
ERROR packet, with error code 8, to terminate the transfer.

The rules for determining the final packet are unchanged from
[RFC1350] and [RFC2348].

The reception of a data window with a number of blocks less than the
negotiated windowsize is the final window. If the windowsize is
greater than the amount of data to be transferred, the first window
is the final window.

4. Traffic Flow and Error Handling

The next diagram depicts a section of the traffic flow between the
Data Sender (DSND) and the Data Receiver (DRCV) parties on a generic
windowsize TFTP file transfer.

The DSND MUST cyclically send to the DRCV the agreed windowsize
consecutives
consecutive data blocks before to normally stop stopping and wait waiting for the
ACK of the transferred window. The DRCV MUST send to the DSND the
ACK of the last data block of the window in order to confirm a
successful data block window reception.

In the case of an expected ACK not timely reaching the DSND (timeout)
(timeout), the last received ACK SHALL set the beginning of the next
windowsize data block window to send. be sent.

In the case of a data block sequence error error, the DRCV SHOULD notify
the DSND by sending an ACK corresponding to the last data block
correctly received. The notified DSND SHOULD send a new data block
window which whose beginning MUST be set based on the ACK received out of sequence ACK.
sequence.

Traffic with windowsize = 1 MUST be equivalent to traffic specified
by RFC1350 [RFC1350].

For traffic normative not specifically addressed in this section section,
please refer to RFC1350 [RFC1350] and its updates.

[ DRCV ] <---traffic---> [ DSND ]
ACK# -> <- Data Block# window block#
...
<- |DB n+01| 1
<- |DB n+02| 2
<- |DB n+03| 3
<- |DB n+04| 4
|ACK n+04| ->
<- |DB n+05| 1
Error |<- |DB n+06| 2
<- |DB n+07| 3
|ACK n+05| ->
<- |DB n+06| 1
<- |DB n+07| 2
<- |DB n+08| 3
<- |DB n+09| 4
|ACK n+09| ->
<- |DB n+10| 1
Error |<- |DB n+11| 2
<- |DB n+12| 3
|ACK n+10| ->| Error
<- |DB n+13| 4
- timeout -
<- |DB n+10| 1
<- |DB n+11| 2
<- |DB n+12| 3
<- |DB n+13| 4
|ACK n+13| ->
...

Section of a windowsize Windowsize = 4 TFTP transfer including
errors Transfer
Including Errors and error recovery Error Recovery

5. Proof of Concept and Windowsize Evaluation

Performance tests were run on the prototype implementation using a
variety of windowsizes and a fixed blocksize of 1456 bytes. The
tests were run on a lightly loaded Gigabit Ethernet, between two
Toshiba Tecra Core 2 Duo 2.2 Ghz laptops, in "octet" mode,
transferring a 180 MByte file.

^
|
300 +
Seconds | windowsize | time (s)
| ---------- ------
| x 1 257
250 + 2 131
| 4 76
| 8 54
| 16 42
200 + 32 38
| 64 35
|
|
150 +
|
| x
|
100 +
|
| x
|
50 + x
| x
| x x
|
0 +-//--+-----+-----+-----+-----+-----+-----+-->
1 2 4 8 16 32 64
windowsize

Windowsize (in blocks Blocks of 1456 bytes)

Comparatively Bytes)

Comparatively, the same 180 MB transfer performed over an SMB/CIFS
mapped a drive mapped
on Session Management Broker (SMB) / Common Internet File System
(CIFS) on the same scenario took 23 seconds.

The comparison of transfer times (without a gateway) between the
standard lock-step lockstep schema and the negotiated windowsizes are:

Windowsize | Time Reduction (%)
---------- -----------------
1 -0%
2 -49%
4 -70%
8 -79%
16 -84%
32 -85%
64 -86%

The transfer time decreases with the use of a windowed schema. The
reason for the reduction in time is the reduction in the number of
the required synchronous acknowledgements exchanged.

The choice of appropriate windowsize values on a particular scenario
depends on the underlying networking technology and topology, and
likely other factors as well. Operators SHOULD test various values
and SHOULD be conservative when selecting a windowsize value because
as the former table and chart shows, there is a point where the
benefit of continuing to increase the windowsize is subject to
diminishing returns.

6. Congestion and Error Control

From a congestion control (CC) standpoint standpoint, the number of blocks in a
window does not pose an intrinsic threat to the ability of
intermediate devices to signal congestion through drops. The rate at
which TFTP UDP datagrams are sent SHOULD follow the CC guidelines in
Section 3.1 of RFC 5405 [RFC5405].

From an error control standpoint standpoint, while RFC 1350 [RFC1350] and
subsequent updates do not specify a circuit breaker (CB), existing
implementations have always chosen to fail under certain
circumstances. Implementations SHOULD always set a maximum number of
retries for datagram retransmissions, imposing an appropriate
threshold on error recovery attempts, after which a transfer SHOULD
always be aborted to prevent pathological retransmission conditions.

An Implementation implementation example scaled for an Ethernet environment (1 Gb/s,
Gbit/s, MTU=1500) would be to set:

windowsize = 8
blksize = 1456
maximum retransmission attempts per block/window = 6
timeout between retransmissions = 1 S
minimum inter-packet delay = 80 uS

Implementations might well choose other values based on expected
and/or tested operating conditions.

7. Security Considerations

TFTP includes no login or access control mechanisms. Care must be
taken when using TFTP for file transfers where authentication, access
control, confidentiality, or integrity checking are needed. Note
that those security services could be supplied above or below the
layer at which TFTP runs. Care must be also be taken in the rights
granted to a TFTP server process so as not to violate the security of
the server's file system. TFTP is often installed with controls such
that only files that have public read access are available via TFTP.
Also listing, deleting, renaming, and writing files via TFTP are
typically disallowed. TFTP file transfers are NOT RECOMMENDED where
the inherent protocol limitations could raise insurmountable
liability concerns.

TFTP includes no protection against an on-path attacker, attacker; care must be
taken in controlling windowsize values according to data sender, data
receiver, and network environment capabilities. TFTP service is
frequently associated with bootstrap and initial provisioning
activities,
activities; servers in such an environment are in a position to
impose device or network specific throughput limitations as
appropriate.

This document does not add any security controls to TFTP; however,
the specified extension does not pose additional security risks
either.

8. IANA Considerations

This document has no actions for IANA.

9. References

9.1.

8.1. Normative References

[RFC1350] Sollins, K., "The TFTP Protocol (Revision 2)", STD 33,
RFC 1350
(STD 33), October 1992. 1350, July 1992,
<http://www.rfc-editor.org/info/rfc1350>.

[RFC2119] Bradner, S., "Key words for use in RFCs to Indicate
Requirement Levels", BCP 14, RFC 2119, March 1997,
<http://www.rfc-editor.org/info/rfc2119>.

[RFC2347] Malkin, G., G. and A. Harkin, A., "TFTP Option Extension", RFC 2347
2347, May 1998. 1998, <http://www.rfc-editor.org/info/rfc2347>.

[RFC2348] Malkin, G., G. and A. Harkin, A., "TFTP Blocksize option", Option", RFC 2348
2348, May 1998. 1998, <http://www.rfc-editor.org/info/rfc2348>.

[RFC5405] Eggert, L. and G. Fairhurst, "Unicast UDP Usage
Guidelines for Application Designers", BCP 145, RFC 5405,
November
2008.

[RFC2119] Bradner, S., "Key words for use in RFCs to Indicate
Requirement Levels", BCP 14, RFC 2119, March 1997.

Authors' Addresses 2008, <http://www.rfc-editor.org/info/rfc5405>.

Author's Address

Patrick Masotta
Serva
300 W 11th Avenue,
Denver, CO 80204

Email: masotta[-at-]vercot[-dot-]com

EMail: masotta@vercot.com
URI: http://www.vercot.com/~serva/