GB2364468A - Buffer management - Google Patents
Buffer management Download PDFInfo
- Publication number
- GB2364468A GB2364468A GB0024735A GB0024735A GB2364468A GB 2364468 A GB2364468 A GB 2364468A GB 0024735 A GB0024735 A GB 0024735A GB 0024735 A GB0024735 A GB 0024735A GB 2364468 A GB2364468 A GB 2364468A
- Authority
- GB
- United Kingdom
- Prior art keywords
- packet
- buffer
- cells
- list
- allocated
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Withdrawn
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L49/00—Packet switching elements
- H04L49/90—Buffering arrangements
- H04L49/9084—Reactions to storage capacity overflow
- H04L49/9089—Reactions to storage capacity overflow replacing packets in a storage arrangement, e.g. pushout
- H04L49/9094—Arrangements for simultaneous transmit and receive, e.g. simultaneous reading/writing from/to the storage element
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L49/00—Packet switching elements
- H04L49/90—Buffering arrangements
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L49/00—Packet switching elements
- H04L49/90—Buffering arrangements
- H04L49/9047—Buffering arrangements including multiple buffers, e.g. buffer pools
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04Q—SELECTING
- H04Q11/00—Selecting arrangements for multiplex systems
- H04Q11/04—Selecting arrangements for multiplex systems for time-division multiplexing
- H04Q11/0428—Integrated services digital network, i.e. systems for transmission of different types of digitised signals, e.g. speech, data, telecentral, television signals
- H04Q11/0478—Provisions for broadband connections
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L12/00—Data switching networks
- H04L12/54—Store-and-forward switching systems
- H04L12/56—Packet switching systems
- H04L12/5601—Transfer mode dependent, e.g. ATM
- H04L2012/5638—Services, e.g. multimedia, GOS, QOS
- H04L2012/5646—Cell characteristics, e.g. loss, delay, jitter, sequence integrity
- H04L2012/5652—Cell construction, e.g. including header, packetisation, depacketisation, assembly, reassembly
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L12/00—Data switching networks
- H04L12/54—Store-and-forward switching systems
- H04L12/56—Packet switching systems
- H04L12/5601—Transfer mode dependent, e.g. ATM
- H04L2012/5638—Services, e.g. multimedia, GOS, QOS
- H04L2012/5665—Interaction of ATM with other protocols
- H04L2012/5667—IP over ATM
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L12/00—Data switching networks
- H04L12/54—Store-and-forward switching systems
- H04L12/56—Packet switching systems
- H04L12/5601—Transfer mode dependent, e.g. ATM
- H04L2012/5678—Traffic aspects, e.g. arbitration, load balancing, smoothing, buffer management
- H04L2012/5681—Buffer or queue management
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
Described herein is a method for reassembling variable length packets from fixed length cells. When a variable length packet, for example, an Internet Protocol (IP) packet, is transmitted between routers over a link which transmits data as fixed length cells, for example, an asynchronous transfer mode (ATM) link, the packet must be segmented into compatible fixed length cells. The receiving router must reassemble the original packet from the cells as they arrive. A packet buffer free pool (300) is provided which is maintained as a linked list, known as a 'free list', and which comprises a plurality of buffer elements (302, 304, 306, 308, 310, 312, 314, 316, 318, 320, 322, 324, 326, 328, 330, 332, 334, 336, 338, 340). When a first cell for a given packet arrives, a buffer element (302) is taken from the head of the free list and allocated to that packet. The packet data from the first and subsequent cells is copied into the allocated buffer element and each time the buffer element is moved to the end of the free list. Upon the arrival of the last cell for the given packet, the complete packet is processed. Failed reassemblies are automatically abandoned as the buffer element is reused when it reaches the head of the free list. This technique is relevant to protection against "denial of service" attacks upon computer networks.
Description
IMPROVEMENTS IN OR RELATING TO BUFFER MANAGEMENT The present invention
relates to improvement in or relating to buffer 5 management, and is more particularly concerned with reassembly buffer management.
In the Internet, data is transferred over a global network of heterogeneous computers by means of a plurality of routing devices in accordance with a standard protocol known as Internet Protocol (IP). IP is 10 a protocol based on the transfer of data in variable sized portions known as packets. All network traffic involves the transportation of packets of data.
In Asynchronous Transfer Mode (ATM) networks, data is transferred in small cells of a fixed length, typically carrying 48 bytes of data. ATM allows high transmission rates by keeping the overheads due to 15 communication protocols to a minimum and by implementing the majority of the conimunication protocols in hardware. In particular, ATM routing is achieved entirely in hardware. In ATM, virtual circuits between senders and destinations called virtual channels are established, the set- up and the maintenance of the virtual channels being implemented in hardware to 20 minimise switching delays.
Routers are devices for accepting incoming packets; temporarily storing each packet; and then forwarding the packets to another part of the network. For the purposes of the following description the term 'routing device' refers to any device which performs the function of a router or a
25 circuit switch. One relevant example of a routing device is an ATM to IP switch.
There is an urgent requirement for routing devices that can route IP traffic at extremely large aggregate bandwidths in the order of several terabits per second. Such routing devices are termed "terabit routers".
When an IP packet is transmitted between routers over an ATM link, 5 the packet must be segmented into fixed length ATM cells. The receiving router must reassemble the original packet from the cells as they arrive.
Conventional reassembly proceeds as follows:
First, a free pool of packet buffers (or reassembly buffers) is maintained. Secondly, on arrival of the first cell for a given packet, a 10 packet buffer is allocated from the free pool. Packet data is copied from the cell into the buffer and a timer is started. The timer is known as a reassembly timer whose fiinction is to protect the system from lost cells.
Upon arrival of each subsequent cell for the given packet, except the last, packet data is copied from the cell into the buffer. After each new 15 copy event, the reassembly timer is restarted. On arrival of the last cell for the given packet, packet data is again copied from the cell into the buffer and the reassembly timer is stopped. The new complete packet is processed and transmitted to its intended destination or destinations. The buffer is then returned to the free pool.
20 If the reassembly timer expires, it is assumed that one or more cells have been lost or corrupted. In this case, the reassembly is abandoned and the buffer is returned to the free pool.
It is important to note, however, that the router must perform multiple concurrent reassernblies. Typically, the router will have a number 25 of ATM virtual circuits open, each carrying data from IP packets. Within any one virtual circuit, the cells for a given packet will arrive contiguously.
However, the cells for the given packet arriving on different virtual circuits will be interspersed relative to one another, which also means that cells from different packets will be interspersed. It is possible that concurrent reassemblies be required for each virtual circuit, each requiring its own timer. For high capacity routers with large numbers of virtual circuits,, large numbers of timers are required.
5 It is therefore an object of the invention to obviate or at least mitigate the aforementioned problems.
In accordance with one aspect of the present invention, there is provided a method of operating a reassembly buffer function, the method comprising the steps ofi 10 a) receiving a first fragment from a new packet; b) allocating a buffer location to the new packet; c) moving the allocated buffer location to the end of a buffer list; d) receiving subsequent fragments and passing them to the allocated buffer location and repeating step c); 15 e) transmitting reassembled packet from the allocated buffer location when the last fragment has been received; f) allowing the allocated buffer location to reach the top of the buffer list if no further fragments are received; and g) reusing the allocated buffer location when it reaches the top of 20 the buffer list.
A fragment is defined as a part of a packet of data which is transmitted separately due to the constraints of a network. A fragment may be a cell or an IP fragment.
An advantage of the present invention is that it allows the reassembly 25 of variable length packets from fixed length cells in the absence of reassembly timers.
In one embodiment of the present invention, a method for reassembling variable length Internet Protocol (IP) packets from fixed length Asynchronous Transfer Mode (ATM) cells in the absence of reassembly timers is provided.
For a better understanding of the present invention, reference will now be made, by way of example only, to the accompanying drawings in which:Figure I illustrates an ATM network; Figure 2 illustrates a device for reassembling packets of data in accordance with the present invention; and Figure 3 illustrates a buffer comprising a part of the Figure 2 device.
10 Figure 1 illustrates an ATM network 100 to which are connected a plurality of packet switches or routers 102, 104, 106, 108, 110, 112.
Although only six packet switches or routers are shown, it will be appreciated that any number of such switches or routers may be connected to the network 100 as required by a particular application.
15 Each packet switch 102, 104, 106, 108, 110, 112 is connected to each of the packet switches via the network 100. Although the network 100 is described as an ATM network, it may also be an internet protocol (IP) network.
Each packet switch 102, 104, 106, 108, 110, 112 can be considered 20 to be an interface unit for a terabit router (not shown). Such a router, for example, RipCore (Registered Trade Mark), comprises a plurality of interface units, each interface unit having to support interface speeds of 2.5, 10 and 40 Gigabits per second. Therefore, packet handling has to be as simple as possible to allow the higher levels of hardware integration 25 required and reduce development risk.
The present invention will now be described with reference to a terabit router, but it will readily understood that it is equally applicable to packet switches or any device where packet data reassembly needs to take place. One particular instance where packet data reassembly is required is at the ingress to a packet switch or a terabit router.
Figure 2 illustrates a terabit router 200 which comprises an input 202 for receiving packets of data, in the form of cells, from a network (not 5 shown). The router 200 includes a cell receive function 204 for receiving individual cells from the network and forwarding the cells to a reassembly buffer 206 where the cells are collected and reassembled into their original packets of data. The reassembled cells are output from the router 200 on output 208. The buffer 206 is described in more detail with reference to 10 Figure 3.
In Figure 3, the buffer 300 comprises a plurality of buffer elements 302 3045 306, 3085 310, 312, 314, 316, 318, 320, 3225 324, 326, 328, 330, 3325 334, 336, 338, 340 arranged in a list. It will be appreciated that, although twenty buffer elements are shown, any suitable number can be 15 employed in accordance with a particular application.
As shown in Figure 3, element 302 is at the top of the list and is therefore free for use, element 340 contains at least one cell from a packet, and element 318 may contain a substantially reassembled packet. This is given by way of example. It will be appreciated that element 318 may also 20 be free as it is in the middle of the list. Furthermore element 340 may also be free if the packet reassembly has only just begun.
In an embodiment of the present invention, the following steps are implemented on a terabit router: - First, a packet buffer free pool, buffer 300, is maintained as a linked 25 list. The linked list is known as a 'free list'.
When the first cell for a given packet arrives, a buffer element is taken from the head of the free list,, for example, buffer element 302, and the packet data from the first cell is copied into that buffer element. The buffer element is then moved to the end of the free list, as shown by arrow 342. Buffer element 340 then moves off the end of the list in the direction indicated by arrow 344.
On arrival of subsequent cells for the given packet, excluding the 5 last, the packet data is copied into the relevant buffer element and the buffer element is moved to the end of the free list.
On arrival of the last cell for the given packet, the packet data from the last cell is copied to the buffer element and the complete packet is processed, and passed to output 208 as shown in Figure 2. Once the 10 reassembly has taken place, the buffer element moves up the list in the direction indicated by arrow 344 until it is at the top of the list and the process re-starts for a new packet.
If cells for a packet are lost so that the complete packet is never received, the buffer element will eventually, as a result of buffer allocations 15 for other packets, reappear at the head of the free list, as indicated by arrow 344) and be re-used for a new packet. The failed reassembly is automatically abandoned.
This is repeated for each individual packet of data so that only one buffer element collects cells relating to a particular packet of data, and 20 there is an effective time out when the buffer element reaches the top of the list.
It will be readily appreciated that this technique could also be used for protection against certain so-called "denial of service" attacks upon computer networks.
25 IP supports packet fragmentation to allow large packets to be transmitted over networks which contain links with physical limits on their packet sizes. Accordingly, a large packet may be broken into a number of small packets to be reassembled at their ultimate destination. This makes the network vulnerable to attack. A hostile agent may send to its target a large number of single fragments each identified as belonging to larger packets, but for which no subsequent fragments are sent. The target (using a conventional reassembly scheme as described above) will reserve 5 resources for each reassembly, resulting in buffer exhaustion. It is difficult to combat this sort of attack through the use of reassembly timers since if the timers were to be short enough to be effective, they would not be long enough to accommodate the arrival of real fragmented packets.
A target using a reassembly scheme in accordance with the present 10 invention is much less likely to suffer buffer exhaustion under such denial of service attacks. Bogus fragments do waste bandwidth but have no ultimate effect on the free pool.
Claims (3)
1. A method of operating a reassembly buffer Rinction, the method comprising the steps of-- a) receiving a first fragment from a new packet; b) allocating a buffer location to the new packet; c) moving the allocated buffer location to the end of a buffer list; d) receiving subsequent fragments and passing them to the allocated buffer location and repeating step c); e) transmitting reassembled packet from the allocated buffer location when the last fragment has been received; f) allowing the allocated buffer location to reach the top of the buffer list if no further fragments are received; and g) reusing the allocated buffer location when it reaches the top of the buffer list.
2. A method according to claim 1, further comprising the step of using the position of the allocated buffer location in the buffer list to effect timing for receipt of fragments from each packet.
3. A method of operating a reassembly buffer function substantially as hereinbefore described with reference to the accompanying drawings.
Priority Applications (7)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP01202199A EP1170975B1 (en) | 2000-07-05 | 2001-06-08 | Method of operating a packet reassembly buffer and network router |
ES01202199T ES2220661T3 (en) | 2000-07-05 | 2001-06-08 | METHOD OF OPERATION OF A TEMPORARY MEMORY OF PACKAGE REASSEMBLY. |
AT01202199T ATE267502T1 (en) | 2000-07-05 | 2001-06-08 | METHOD FOR OPERATING A PACKET REASSEMBLY BUFFER AND NETWORK ROUTER |
DE60103331T DE60103331T2 (en) | 2000-07-05 | 2001-06-08 | Method of operating a packet reassembly buffer and network router |
CA002350778A CA2350778C (en) | 2000-07-05 | 2001-06-14 | Improvements in or relating to buffer management |
JP2001204460A JP4606657B2 (en) | 2000-07-05 | 2001-07-05 | How the reconfiguration buffer works |
US09/898,485 US6888835B2 (en) | 2000-07-05 | 2001-07-05 | Buffer management |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
GBGB0016474.9A GB0016474D0 (en) | 2000-07-05 | 2000-07-05 | Reassembly buffer management |
Publications (2)
Publication Number | Publication Date |
---|---|
GB0024735D0 GB0024735D0 (en) | 2000-11-22 |
GB2364468A true GB2364468A (en) | 2002-01-23 |
Family
ID=9895043
Family Applications (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
GBGB0016474.9A Ceased GB0016474D0 (en) | 2000-07-05 | 2000-07-05 | Reassembly buffer management |
GB0024735A Withdrawn GB2364468A (en) | 2000-07-05 | 2000-10-10 | Buffer management |
Family Applications Before (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
GBGB0016474.9A Ceased GB0016474D0 (en) | 2000-07-05 | 2000-07-05 | Reassembly buffer management |
Country Status (1)
Country | Link |
---|---|
GB (2) | GB0016474D0 (en) |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP0404078A2 (en) * | 1989-06-19 | 1990-12-27 | Hitachi, Ltd. | Communication apparatus for reassembling packets received from a network into a message |
US5809024A (en) * | 1995-07-12 | 1998-09-15 | Bay Networks, Inc. | Memory architecture for a local area network module in an ATM switch |
US5870394A (en) * | 1996-07-23 | 1999-02-09 | Northern Telecom Limited | Method and apparatus for reassembly of data packets into messages in an asynchronous transfer mode communications system |
US5898688A (en) * | 1996-05-24 | 1999-04-27 | Cisco Technology, Inc. | ATM switch with integrated system bus |
US5920561A (en) * | 1996-03-07 | 1999-07-06 | Lsi Logic Corporation | ATM communication system interconnect/termination unit |
US6148001A (en) * | 1997-06-12 | 2000-11-14 | Nokia Telecommunications, Oy | Multipoint-to-point system which multiplexes complete packets comprised of ATM cells on to a single virtual channel connection |
-
2000
- 2000-07-05 GB GBGB0016474.9A patent/GB0016474D0/en not_active Ceased
- 2000-10-10 GB GB0024735A patent/GB2364468A/en not_active Withdrawn
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP0404078A2 (en) * | 1989-06-19 | 1990-12-27 | Hitachi, Ltd. | Communication apparatus for reassembling packets received from a network into a message |
US5809024A (en) * | 1995-07-12 | 1998-09-15 | Bay Networks, Inc. | Memory architecture for a local area network module in an ATM switch |
US5920561A (en) * | 1996-03-07 | 1999-07-06 | Lsi Logic Corporation | ATM communication system interconnect/termination unit |
US5898688A (en) * | 1996-05-24 | 1999-04-27 | Cisco Technology, Inc. | ATM switch with integrated system bus |
US5870394A (en) * | 1996-07-23 | 1999-02-09 | Northern Telecom Limited | Method and apparatus for reassembly of data packets into messages in an asynchronous transfer mode communications system |
US6148001A (en) * | 1997-06-12 | 2000-11-14 | Nokia Telecommunications, Oy | Multipoint-to-point system which multiplexes complete packets comprised of ATM cells on to a single virtual channel connection |
Also Published As
Publication number | Publication date |
---|---|
GB0016474D0 (en) | 2000-08-23 |
GB0024735D0 (en) | 2000-11-22 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US6888835B2 (en) | Buffer management | |
US6781994B1 (en) | Distributing ATM cells to output ports based upon destination information using ATM switch core and IP forwarding | |
US6873796B1 (en) | Node device and optical path setting method | |
EP0274793B1 (en) | Packet switching network | |
KR100308593B1 (en) | How to Route Packet-Switched Traffic | |
JP4547339B2 (en) | Packet relay device having transmission control function | |
US7912069B2 (en) | Virtual segmentation system and method of operation thereof | |
US20020085567A1 (en) | Metro switch and method for transporting data configured according to multiple different formats | |
US20050152355A1 (en) | Reducing transmission time for data packets controlled by a link layer protocol comprising a fragmenting/defragmenting capability | |
US6212185B1 (en) | Multiple network address resolution | |
JP2002300193A (en) | Router | |
US20020085548A1 (en) | Quality of service technique for a data communication network | |
US5434855A (en) | Method and apparatus for selective interleaving in a cell-switched network | |
US7197051B1 (en) | System and method for efficient packetization of ATM cells transmitted over a packet network | |
US6785239B1 (en) | Reducing delays in a communication network using a re-fragmentation pipeline | |
CN111602376A (en) | Method and apparatus for managing transmission of delay sensitive packets | |
Box et al. | Architecture and design of connectionless data service for a public ATM network | |
US7139253B2 (en) | Packet switches | |
WO1994014263A1 (en) | A method for congestion management in a frame relay network and a node in a frame relay network | |
US6515998B1 (en) | Table data retrieving apparatus retrieving table in which reference data is stored by using retrieval key | |
GB2364468A (en) | Buffer management | |
Moors et al. | ATM receiver implementation issues | |
US7561800B2 (en) | Optical burst switching system and method using duplicate burst transmission | |
EP1011291A2 (en) | VC-merge-capable label switching routers for IP communications over ATM networks | |
EP1052808A1 (en) | Communications network using priority classes |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
WAP | Application withdrawn, taken to be withdrawn or refused ** after publication under section 16(1) |