U.S. patent application number 12/907389 was filed with the patent office on 2011-02-10 for system and method for interfacing with a management system.
This patent application is currently assigned to BROADCOM CORPORATION. Invention is credited to Steven B. Lindsay, Scott S. McDaniel.
Application Number | 20110035489 12/907389 |
Document ID | / |
Family ID | 32965648 |
Filed Date | 2011-02-10 |
United States Patent
Application |
20110035489 |
Kind Code |
A1 |
McDaniel; Scott S. ; et
al. |
February 10, 2011 |
SYSTEM AND METHOD FOR INTERFACING WITH A MANAGEMENT SYSTEM
Abstract
Systems and methods that interface with a management system are
provided. In one embodiment, a system and a method may provide a
command protocol and format for communication between a network
interface card (NIC) and a management device such as, for example,
an intelligent management device (IMD). An interface may be adapted
to allow the management device to merge its traffic with that of
the NIC to provide a fully integrated management solution. The
fully integrated management solution may be implemented, for
example, without additional network connections.
Inventors: |
McDaniel; Scott S.; (Villa
Park, CA) ; Lindsay; Steven B.; (Bend, OR) |
Correspondence
Address: |
MCANDREWS HELD & MALLOY, LTD
500 WEST MADISON STREET, SUITE 3400
CHICAGO
IL
60661
US
|
Assignee: |
BROADCOM CORPORATION
Irvine
CA
|
Family ID: |
32965648 |
Appl. No.: |
12/907389 |
Filed: |
October 19, 2010 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
12190425 |
Aug 12, 2008 |
7817662 |
|
|
12907389 |
|
|
|
|
10797532 |
Mar 10, 2004 |
7411973 |
|
|
12190425 |
|
|
|
|
60453642 |
Mar 11, 2003 |
|
|
|
Current U.S.
Class: |
709/224 ;
370/389; 709/223 |
Current CPC
Class: |
H04L 43/04 20130101;
H04L 43/028 20130101; H04L 41/06 20130101 |
Class at
Publication: |
709/224 ;
709/223; 370/389 |
International
Class: |
G06F 15/173 20060101
G06F015/173; H04L 12/56 20060101 H04L012/56 |
Claims
1. A communications system, comprising: a network interface card
(NIC); and a management device coupled to the NIC, wherein the NIC
is adapted to merge communications traffic of the management device
with the NIC.
2-56. (canceled)
Description
CROSS REFERENCE TO RELATED APPLICATION
[0001] This application makes reference to, claims priority to and
claims benefit from U.S. Provisional Patent Application Ser. No.
60/453,642, entitled "Intelligent Management Device Interface" and
filed on Mar. 11, 2003.
INCORPORATION BY REFERENCE
[0002] The above-referenced United States patent application is
hereby incorporated herein by reference in its entirety.
FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT
[0003] [Not Applicable]
MICROFICHE/COPYRIGHT REFERENCE
[0004] [Not Applicable]
BACKGROUND OF THE INVENTION
[0005] FIG. 1 illustrates a block diagram of a system 10 connected
to a network 20 via switches 30a, 30b. The system 10 includes a
network interface card (NIC) 40 and an intelligent management
device (IMD) 50. The NIC 40 is connected to the switch 30a and to a
host (not shown) of the system 10. The IMD is connected to the
switch 30b and to the host of the system 10. The NIC includes a NIC
media access control (MAC) 60 and a NIC processor 70. The network
20 is connected to the switch 30a which, in turn, is connected to
the NIC MAC 60. The NIC MAC 60 is connected to the NIC processor 70
which, in turn, is connected to the host of the system 10. The IMD
50 includes a MAC 80 and management processor 90. The network 20 is
also connected to the switch 30b which, in turn, is connected to
the MAC 80. The MAC 80 is connected to the management processor 90
which, in turn, is connected to the host of the system 10.
[0006] The IMD 50 provides, for example, monitoring, management
capabilities and remote functionality. For example, the IMD 50 can
provide monitoring and management capabilities for the system 10
and can provide remote functionality to or from a device (e.g., a
remote device) connected to the network 20.
[0007] The IMD 50 can have one or more of the following
disadvantages. For example, as illustrated in FIG. 1, the system 10
includes an additional dedicated connection to the network 20.
Besides the additional cost (e.g., the additional switch 30b) of
implementing another system port, the IMD 50 is susceptible to a
failure, for example, of the switch 30b. Thus, if switch 30b were
to fail, then the IMD 50 would no longer be accessible via the
network 20. Furthermore, because the IMD 50 is connected to the
network 20, the IMD 50 may be needlessly processing some packets
carried on the network 20. On high-speed networks, in particular,
the resources of the IMD 50 can be substantially consumed by such
unnecessary processing, thereby reducing some resources of the IMD
50 that could have been allocated for other tasks. For example,
some packets (e.g., packets that can be forwarded as received) may
be processed by the IMD 50, even though these packets need not be
processed by the IMD 50.
[0008] Further limitations and disadvantages of conventional and
traditional approaches will become apparent to one of ordinary
skill in the art through comparison of such systems with one or
more aspects of the present invention as set forth in the remainder
of the present application with reference to the drawings.
BRIEF SUMMARY OF THE INVENTION
[0009] Aspects of the present invention may be found in, for
example, systems and methods that interface with a management
system such as, for example, an intelligent management device. In
one embodiment according to some aspects of the present invention,
a communications system may include, for example, a network
interface card (NIC) and a management device. The management device
may be coupled to the NIC. The NIC may be adapted, for example, to
merge communications traffic of the management device with the
NIC.
[0010] In another embodiment according to some aspects of the
present invention, a communications system may include, for
example, a first NIC, a second NIC and a manager. The first NIC and
the second NIC may be coupled to a network. The manager may be
coupled to the first NIC and the second NIC. The manager may
initially be in two-way communications with the network via the
first NIC. However, if the first NIC fails, then the manager may
switch from the first NIC to the second NIC and be in two-way
communications with the network via the second NIC.
[0011] In yet another embodiment according to some aspects of the
present invention, a method of communications may include, for
example, one or more of the following: providing access to and from
a network for a management device via a NIC; configuring one or
more filters of the NIC via one or more commands generated by an
management device; filtering incoming packets via the one or more
filters; and forwarding the filtered packets based upon one or more
matches between information carried by the filtered packets and one
or more filtering parameters.
[0012] In yet another embodiment according to some aspects of the
present invention, a method of communications between a NIC and a
management device may include, for example, one or more of the
following: generating a command in the management device, the
command comprising a particular sequence number; storing the
command in the management device; sending the command to the NIC;
executing the command in the NIC; and generating a response to the
command, the response comprising the particular sequence number. In
some embodiments according to the present invention, the command
may also include, for example, an identifier-type field and a
command structure. The response may also include, for example, an
identifier-type field and a response structure.
[0013] In yet still another embodiment according to some aspects of
the present invention, a method of remote management over a network
may include, for example, one or more of the following: accessing
the network via a plurality of NICs of a local server system;
communicating between a local manager of the local server system
and a remote manager over the network through a NIC selected by the
local manager, the selected NIC being one of the plurality of NICs;
managing the local server system via the local manager; and
responding locally to management commands sent over the network
from the remote manager.
[0014] These and other features and advantages of the present
invention may be appreciated from a review of the following
detailed description of the present invention, along with the
accompanying figures in which like reference numerals refer to like
parts throughout.
BRIEF DESCRIPTION OF THE DRAWING
[0015] FIG. 1 shows a block diagram of a system connected to a
network via switches.
[0016] FIG. 2 shows a block diagram of a system coupled to a
network according to an embodiment of the present invention.
[0017] FIG. 3 shows a block diagram of a system coupled to a
network according to an embodiment of the present invention.
[0018] FIG. 4 shows a flowchart illustrating an embodiment of a
process for receiving and forwarding packets according to the
present invention.
[0019] FIG. 5 shows a flowchart illustrating an embodiment of a
process for handling command packets according to the present
invention.
[0020] FIG. 6 shows a flowchart illustrating an embodiment of a
process for handling response packets according to the present
invention.
DETAILED DESCRIPTION OF THE INVENTION
[0021] Aspects of the present invention may be found, for example,
in systems and methods that interface with a management system such
as, for example, a management system including an intelligent
management device (IMD). In some embodiments according to the
present invention, a command protocol and format for communication
between an interface card (e.g., a network interface card (NIC), a
controller, an adapter, etc.) and a management system may be
provided.
[0022] In some embodiments according to the present invention, an
interface that may allow a management system to merge its traffic
with that NIC (e.g., a standard NIC, a network interface
controller, etc.) to provide a fully integrated management solution
may be provided. The fully integrated management solution may be
implemented, for example, without adding further network
connections.
[0023] In some embodiments according to the present invention, a
separate Ethernet connection port on a NIC may be provided. The
separate Ethernet connection port may be, for example, a universal
management port (UMP). Via this interface, a management system may
see packets that it would see if it were directly connected to the
network as well as many other types of packets (e.g., commands,
responses, etc.)
[0024] In some embodiments according to the present invention, a
system that passes command packets over an Ethernet interface
between a management system and an UMP may be provided.
[0025] FIG. 2 shows an embodiment of a system 100 including an IMD
110 according to the present invention. The system 100 may be, for
example, a server system, a server blade, a desktop system, a
computer system, a network system, a set top box, etc.) The system
100 may include, for example, the IMD 110 and a NIC 120. The system
100 may be coupled to a network 130 via the NIC 120. The IMD 110
may include, for example, a management processor 140 and a media
access control (MAC) 150 (e.g., a 10/100 MAC). The NIC 120 may
include, for example, a NIC processor 160, a set of filters 170, a
MAC 180 (e.g., a 10/100 MAC or UMP) and a main NIC MAC 190. The
management processor 140 may be coupled to the MAC 150 which, in
turn, may be coupled to the MAC 180. The MAC 180 may be coupled to
the NIC processor 160 which, in turn, may be coupled to the main
NIC MAC 190. The NIC processor 160 may also be coupled to the set
of filters 170 which, in turn, may be coupled to the NIC MAC 190.
The main NIC MAC 190 may be coupled to the network 130 via a switch
200.
[0026] The system 100 may also include, for example, system sensors
210, system controls 220, a system interconnect 230, a central
processing unit (CPU) 240, another CPU 250, a system storage device
260 (e.g., a system memory) and peripheral devices 270 (e.g., disk
devices, video devices, etc.) The system sensors 210 and the system
controls 220 may be coupled to the IMD 110 and, in particular, may
be coupled to the management processor 140. The NIC 120, the IMD
110, the CPUs 240, 250, the system storage device 260 and the
peripheral devices 270 may each be coupled to the system
interconnect 230. In particular, the NIC processor 160 and the
management processor 140 may be coupled to the system interconnect
230 via a first host connection 280 and a second host connection
290. The host connections 280, 290 may be, for example, peripheral
component interconnects (PCIs).
[0027] Other devices may be coupled to the network 130. For
example, a management console 300 may be coupled to the network 130
via a switch 310. Other systems 320 may be coupled to the network
via respective switches 330. Although illustrated as a single
switch 330, each respective system 320 may include its own
respective switch or switches 330. Some of the other systems 320
may be identical to the system 100 illustrated in FIG. 2 and some
of the other systems 320 may be identical to the system 100
illustrated in FIG. 3. However, the other systems 320 need not be
so limited in scope. In some embodiments, some of the other systems
320 and the system 100 may be remotely controlled (e.g., remotely
monitored, remotely activated, remotely managed, remotely accessed,
etc.) by the management console 300.
[0028] FIG. 3 shows an embodiment of the present invention in which
the system 10 is coupled to the network 130 via a plurality of NICs
120 (i.e., NIC 120(1), NIC 120(2), . . . , NIC 120(n), where n is
an integer value). Each NIC 120 may include, for example, a
respective NIC processor 160, a set of filters 170, a first MAC 180
and a second MAC 190. Each NIC 120 may be coupled to the network
130 via a respective switch 200. The various components may be
coupled together as described above with respect to the NIC 120 of
FIG. 2. In addition, each NIC 120 may be coupled to the IMD 110. In
one embodiment, the second MACs 190 of the respective NICs 120 each
may be coupled to the MAC 150 of the IMD 110.
[0029] In operation and with reference to FIG. 2, the management
console 300 may be in two-way communications with the system 100
and some of the other systems 320 coupled to the network 130. The
management console 300 may provide, for example, management
services (e.g., remote management services) for the system 100 and
some of the other systems 320. Thus, the management console 300 may
be able to pull up graphical user interfaces (e.g., windows) on its
display for each of the systems that the management console 300
manages. In some embodiments, the activities performed via the
management console 300 may appear seemless even though the systems
being managed may be far way.
[0030] The IMD 110 may be in two-way communications with the
management console 300 (e.g., a remote management console) via
network 130. The management console 300 may be in two-way
communications with not only the IMD 110 of the system 100, but
also some of the other systems 320 coupled to the network 130. In
some embodiments according to the present invention, the IMD 110
may be in two-way communications with the management console 300
only through the NIC 120. In various embodiments according to the
present invention, the IMD 110 may not have its own direct
connection to the network 130, but instead may use the NIC 120 to
access the network. The IMD 110 may provide management services for
the system 100. For example, the IMD 110 may monitor the system
sensors 142 or adjust the system controls 144. Thus, for example,
the IMD 110 may monitor power supply parameters, voltage
parameters, current parameters, temperature parameters, status
parameters, failure parameters of various components and circuits.
The system sensors 142 may also provide alerts to the IMD 110 such
as, for example, that the cover of the system 100 has not been
replaced, that components have been removed or failed, or that the
temperature in the system 100 has exceeded particular thresholds.
The IMD 110 may also adjust, activate or set system controls 144 in
response to monitored parameters or in response to particular
commands or requests. For example, the IMD 110 may reset power
settings or parameters settings, power up the system 100 or a
particular component of the system 100, or power down the system or
a particular component of the system 100.
[0031] In addition to responding to current conditions, the IMD 110
may also respond to requests received from the host of the system
100 or from the network 130. In some embodiments according to the
present invention, the IMD 110 may receive requests or commands
from the management console 300. In various embodiments according
to the present invention, the management console 300 may provide,
for example, a user input device (e.g., a keyboard, a mouse, etc.)
and a user output device (e.g., a display, a graphical user
interface, a video output, a graphical output, an audio output,
etc.) By opening up a window in a display, a user located at the
management console 300 may monitor information that the IMD 110 may
be receiving from the system sensors 210 or the system controls 220
or other components of the system 100. The user may then send
commands or requests to the IMD 110 to which the IMD 110 may
respond. For example, the management console 300 may send a request
for information to the IMD 110. The request may be sent via the
network 130 to the NIC 120. The request may be routed through the
filters 170 which may determine whether or not the request is
destined for the IMD 110. If the request had not been destined for
the IMD 110, then, for example, it may have been forwarded to the
NIC processor 160 or to the host of the system 100 or elsewhere for
further processing. If the request is destined for the IMD 110,
then the NIC 120 may route the request to the IMD 110. Thus, in
some embodiments according to the present invention, the IMD 110
may only receive requests or commands or data packets that are
destined for the IMD 110.
[0032] The IMD 110 may then analyze the request and may perform the
request. For example, the request may include commands for powering
down an overheated component. The IMD 110 may thus adjust one or
more of the system controls 144 of the system 100. The IMD 110 may
then send a response to the management console 300 through the NIC
120 and the network 130. The response may include, for example,
graphical information for display on a monitor that the user is
viewing. Thus, user seemlessly manages the IMD 110 as if the user
were located at the system 100. In some embodiments, the management
console 300 may act as a dummy terminal in which keyboard strokes
or other input commands (e.g., mouse interface information) are
relayed to the IMD 110 which, in turn, then may send graphical
information to the management console 300 which may be displayed on
the monitor as alphanumeric symbols (or, for example, cursor
movements) corresponding to the keystrokes (or, for example, mouse
movements).
[0033] Referring to FIG. 3, the IMD 110 may be coupled to a
plurality of NICs 120. In some embodiments according to the present
invention, only one of the plurality of NICs 120 provides access to
and from the network 130 for the IMD 110. The IMD 110 may select
which of the NICs 120 to use as a sole connection to and from the
network 130. In some embodiments according to the present
invention, the IMD 110 may have the ability to switch between using
one NIC or another NIC as the sole connection to and from the
network 130, for example, at different times. This ability may
enhance throughput as well as promote fault tolerance. For example,
if the selected NIC is being overused by other resources of the
system 100 or the network 130, then the IMD 110 may select another
NIC (e.g., a NIC that is not being overused by other resources of
the system 100 or the network 130). In another example, if the
selected NIC has failed or has been removed from the system 100,
then the IMD may select another NIC to provide a connection (e.g.,
a sole connection) to and from the network 130. Furthermore, if the
management console 300 determines that the selected NIC has failed
or has been removed from the system 100, the management console 300
may wait and try again or the management console 300 may broadcast
the information to the other NICs of the system 100. The IMD 110
may then be notified via the broadcasted information received from
one or more of the other NICs of the system and switch to another
NIC. The IMD 110 may then use the newly selected NIC to inform the
management console 300 of the switch in NIC selection.
[0034] In some embodiments according to the present invention, the
system 100 may allow, for example, the IMD 110 to merge its traffic
with that of the NIC 120. The system 100 may provide, for example,
a fully integrated management solution that does not necessitate
additional network connections. The NIC 120 may transmit, for
example, non-command packets via the main NIC MAC 190. The NIC 120
may receive, for example, packets via the main NIC MAC 190 and send
the received packets through the set of filters 170. The set of
filters 170 may allow received packets to be passed to the IMD 110
if, for example, the received packets meet (e.g., match) at least
some of the filter requirements (e.g., programmed filter
parameters). The NIC 120 may receive packets from the IMD 110 that
meet a particular encapsulation format and may process the packets
locally. The NIC 120 may receive other types of packets from the
IMD 110 that may be transmitted exactly as they were received from
the IMD 110 including, for example, virtual local area network
(VLAN) tag information.
[0035] FIG. 4 shows a flowchart illustrating an embodiment of a
process for receiving and forwarding packets according to the
present invention. In step 340, the system 100 may receive packets
from the network 130 via the main NIC MAC 190. In query 350,
whether the set of filters 170 are operable may be based upon
whether the set of filters 170 have been configured. The set of
filters 170 may be configured, for example, via commands generated
by the NIC processor 160 of the NIC 120 or the management processor
140 of the IMD 110. If the set of filters 170 have been configured,
then, in query 360, it is determined whether the information
carried by the packet satisfies (e.g., matches) one or more of the
configured filter parameters. If the information carried by the
packet does satisfy one or more of the configured filter
parameters, then, in step 390, the packet may be processed by the
NIC processor 160 and the management processor 140 and forwarded to
a destination according to the matched filter parameters. In one
embodiment, with regard to layer 2 (L2) address values, the NIC 120
may forward data that may meet, for example, any of the selected
perfect match filters, for example, for the L2 address. In another
embodiment, the NIC 120 may limit forwarded traffic up to a
particular number of VLAN networks. In yet another embodiment, the
NIC 120 may filter L2 multicast or broadcast traffic and forward
such traffic accordingly. In one embodiment, a broadcast packet
that is determined to be an address resolution protocol (ARP)
packet or other specific types of broadcast packets may be
forwarded accordingly.
[0036] If the set of filters are not configured (query 350) or if
the information carried by the packet does not satisfy any of the
configured filter parameters (query 360), then, in step 370, the
packet may not reach the IMD 110. In step 380, the packet may be
forwarded as received by the main NIC MAC 190 to the rest of the
system 100. In one embodiment, if the packets are not filtered, the
receive (RX) packets including, for example, VLAN tag information
may be forwarded as received via the main NIC MAC port 190.
[0037] In some embodiments, the present invention may provide
command packets and/or response packets transmitted and/or received
by the NIC 120 having a format as set forth below. For example, the
commands may be from the IMD 110 to the NIC 120 and the responses
may be from the NIC 120 to the IMD 110.
TABLE-US-00001 Byte 31 23 15 7 0 XXXX XXXX DA DA 4 DA DA DA DA 8 SA
SA SA SA 12 SA SA BRCM Num. 16 0x5706 Cmd. Seq. Num. 20 Cmd. Type
Data Length 24 Data0 Data1 Data2 Data3 28 Data4 Data5 Data6 Data7 .
. . . . . . . . . . . . . . . . . DataN Optional Padding to 32 bits
. . . 2s Complement Checksum Compensation . . . Optional Zero Pad
to 64 Byte Legal Ethernet Frame
[0038] In some embodiments, the present invention may provide for
fields with definitions as set forth below. [0039] XXXX Padding for
the purposes of this table to make protocol values align on the
32-bit rows. These bytes may not be a part of any packet. [0040] DA
Ethernet Destination Address. Value is configurable, but typically
is equal to the address the receiver represents on the main
interface. [0041] SA Ethernet Source Address. Value is
configurable, but typically is equal to the address the sender
represents on the main interface. [0042] BRCM Number [0043]
Well-known IEEE packet type number. [0044] 0x5706 Packet type
number used to define a packet sub-type for the BRCM number. [0045]
Cmd Type Command Type. When bit 15 of the type is set, it indicates
a response packet sent from the NIC to the IMD. [0046] Data Length
[0047] This value is the length in bytes of the data payload of the
command packet. This is the number of "Data" bytes in the packet.
This count may not include this value or any bytes before this
value. This count may not include any Padding or Checksum bytes.
The checksum may sum bytes not in this count if this count is not a
multiple of four. [0048] Data . . . . Command payload. [0049]
Optional Padding to 32-bits [0050] If the Data portion of the
command is not a multiple of four in length (i.e., Bottom two bits
of Data Length field are not zero) then from one to three bytes of
zero value padding may be added so that the 2s complement checksum
may be calculated over full 32-bit values. [0051] 2s Complement
Checksum Compensation [0052] When this value is added to the 2's
complement sum of all the 32-bit words in the "Data" and "Optional
Padding" areas, the resulting value may be 0xffffffff. The checksum
checking may optionally be disabled at each receiver. The sender
may then take advantage of this by setting this value to
0xffffffff. [0053] Optional Zero Pad to 64 Byte Legal Ethernet
Frame [0054] For frames to be compliant with IEEE 820.3, all frames
are padded up to a total of 64 bytes. This is typically done
automatically by most Ethernet MAC devices.
[0055] In some embodiments, the present invention may provide that
command frames may not use additional L2 encapsulation techniques
such as, for example, VLAN, SNAP, etc.
[0056] In some embodiments, the present invention may employ a
transport mechanism by which reliable reception of control frames
may be tracked via sequence numbers. Each command packet may be
associated with a respective response packet by each carrying a
common sequence number in a particular sequence number space.
Accordingly, the same sequence number space may be used for
command/response packets in either direction between the NIC 120
and the IMD 110.
[0057] FIG. 5 shows a flowchart illustrating an embodiment of a
process for handling command packets according to the present
invention, for example, as executed by the IMD 110. In step 400, a
sender of a command (e.g., the IMD 110) may generate a command
packet. The command packet may include, for example, a sequence
number with which the command packet is associated. In step 410,
the sender of the command may then store the generated command. In
step 420, the sequence space counter of the sender may then be
incremented. In one embodiment, the step 420 may proceed step 400
and step 410. In another embodiment, the command is sent from the
IMD 110 to the NIC 120. The NIC 120 may respond to the command with
a response. In step 422, the NIC 120 may receive a packet from the
IMD 110. In query 424, a BRCM number may be detected in a
particular field (e.g., an identifier-type field). If the BRCM is
not detected, then, in step 426, the packet may be processed as a
management protocol frame for the management console 300. If the
BRCM is detected, then query 430 follows. In query 430, if sender
of command receives a response, then, in step 450, the command
corresponding to the received response may be deleted from storage.
In one embodiment, the received response may include, for example,
a particular sequence number associated with a particular command,
thereby identifying the stored command which can be deleted.
[0058] If a response to a particular command is not received, then,
in step 440, the particular command, which has been stored, may be
retransmitted. In one embodiment, the command sender may be able to
retransmit any command until a response with the same sequence
number is received and is processed. Accordingly, a retransmission
capability is provided in the case a command packet is lost,
corrupted or dropped. In another embodiment, if the sender of
commands detects a particular sequence number of a response that
has already been received, then the sender of commands may resend
all the commands that have not had responses received that follow
the latest received sequence number. In yet another embodiment, via
a periodic timer, a requestor may verify commands to which have
been responded. If a command has not been responded to with a
response, then a retransmit of the command may be commenced. The
requestor may time-out on any request without a response and may
start retransmission after that timer has expired. This may be
done, for example, by setting the timeout each time a command is
transmitted and each time a response is received. When the timer
expires, retransmit may be started if all the outstanding commands
have not been responded to with corresponding responses.
[0059] FIG. 6 shows a flowchart illustrating an embodiment of a
process for handling response packets according to the present
invention as executed, for example, by the NIC 120. In step 460, a
packet may be received. In one embodiment, a command packet may be
received by a responder to commands (e.g., the NIC 120) from the
sender of commands (e.g., the IMD 110). In query 462, a BRCM number
may be detected in a particular field (e.g., an identifier-type
field). If the BRCM number is not detected, then, in step 464, the
packet may be forwarded to a main transmission port. If the BRCM
number is detected, then, in step 470, the sequence number carried
by the command packet may be determined. For example, the NIC
processor 160 may parse the command packet to determine the
sequence number carried by the command packet. In query 480, it may
be determined whether the determined sequence number is the
expected sequence number. If the determined sequence number is the
expected sequence number, then, in step 490, the command may be
executed and a response packet may be generated that may include,
for example, the expected sequence number. In one embodiment, the
NIC processor 160 may configure (e.g., program), for example, one
or more filters in the set of filters 170 in light of the received
command packet. In step 500, the generated response packet may be
stored and returned to the sender of the command packet. In one
embodiment, the responder to commands may store only the last
response packet sent to the sender of commands. Thus, the latest
response packet may be written over the previous response packet or
the previous response packet may be deleted or invalidated. In
another embodiment, the responder to commands may store one or more
response packets sent to the sender of commands.
[0060] If the determined sequence number is not the expected
sequence number (query 480), then, in step 510, the stored response
packet (e.g., the previously stored response packet) may be resent.
In one embodiment, if the responder receives a command with a
sequence number other than the one expected (e.g., the last one
plus one), then the NIC 120 may not execute that command, but
instead may send the response for the previously executed command.
In another embodiment, the responder to commands may re-execute a
command without any adverse effects (e.g., side effects).
[0061] In one embodiment, no more than approximately 2.sup.15-1
commands may be outstanding at any time, for example, due to
protocol limitations. Memory limitations in the sender may become
evident long before the protocol limit is reached. The responder
may only keep the last response that was sent so that it might not
have any retransmit memory limitations for outstanding packets.
However, other limitations may become factors for
consideration.
[0062] Since the requestor may be the only station saving the
outstanding commands, the responder may be able to execute all
commands more than once without any adverse effects. Accordingly,
the retransmission of a command that exhibits modal effects may
cause problems. These modal effects can be avoided by the design of
the command packets such that repeated execution has no side
effects.
[0063] The following are some examples of commands and responses
according to an embodiment of the present invention.
[0064] Hello--0x0001. The Hello command may solicit a presence
response from the NIC 120. The NIC 120 may respond as long as the
message is received without error. The command does not make a
commitment to work further. The Hello command may be intended for
use as a flush command, if needed, possibly during initial
negotiation. Error Codes may include, for example, OK.
[0065] ID Request--0x0002. The ID Request command may indicate the
IMD type and version as a string up to, for example, forty
characters long. The response to the ID Request command may return
the NIC type and version as a string up to, for example, forty
characters long. Error Codes may include, for example, OK; UNAVAIL;
BAD_ID; or FATAL
[0066] Reset To Default--0x0003. The Reset To Default command may
request that defaults be set for settings relating to, for example,
filters, flow controls, etc. This may be an equivalent state to a
reset NIC. Error Codes may include, for example, OK or FATAL.
[0067] Set NIC<->IMD Flow Control Method--0x0101. The Set
NIC<->IMD Flow Control Method command may set, for example,
the flow control method used between, for example, the IMD 110 and
the NIC 120. In some embodiments, only symmetrical flow control may
be supported in which both ends use the same method. An OK response
may indicate that the NIC 120 has set itself to the same flow
control method. Settings may include, for example, NONE or PAUSE.
Error Codes may include, for example, OK; NOT SUPPORTED; or
FATAL.
[0068] Set NIC Drop Policy--0x0102. The Set NIC Drop Policy command
may set the drop policy used by the NIC 120 when passing frames to
the IMD 110. Settings may include, for example, DROP or
BACKPRESSURE. Responses may include, for example, DROP or
BACKPRESSURE--setting now in; or DRIVER_CTL--bit set when driver is
running and BACKPRESSURE mode could not be selected. Error Codes
may include, for example, OK or FATAL.
[0069] Set NIC Port--0x0103. The Set NIC Port command may request,
for example, that the NIC 120 set the main port to a particular
speed. If the main port speed is already at the specified speed,
then no action may be taken and the link might not be dropped. The
speed may be limited, for example, by power management limitations.
Setting the requested speed may jeopardize exceeding the power
supply for a particular design. This may not be returned on a card
or LAN-on-motherboard (LOM) that may provide enough power in all
modes. The speed change may be denied due to the OS driver being
loaded. The speed change may be denied because of, for example,
interface limits such as Fiber or because an external "in-line"
device connected that may not be controlled for some reason.
Settings may include, for example, GET, 10, 100, 1000--GET requests
that current state be returned with no effect; HD_FLAG--half duplex
flag, invalid with 1000 setting; PAUSE--Pause Flow Control Enable
Flag; or AUTO--flag indicating that a setting is a "maximum
setting" to advertise for auto-negotiation, when set, the link may
be dropped as the link is re-negotiated. Responses may include, for
example, 10, 100, 1000--speed of link; HD_FLAG--set if link is half
duplex; PAUSE--set if pause flow control is enabled; LINK UP--flag
set if link was attained in new mode; FIBER--flag set if fiber
connection; PWR_LIMIT--flag set if requested speed was denied due
to particular card power limitations; DRV_LIMIT--flag set if
request speed was denied due to driver being loaded and forcing
link type; INTF_LIMIT--flag is set if requested speed was denied
due to limitations in the PHY device (e.g., Fiber that cannot
support 10/100); AUTO_NEG--flag is set if link was attained using
auto-negotiation; AUTO_NEG_PAR--flag is set if link was attained
using auto-negotiation and parallel detection was used to detect
link (AUTO_NEG may also be set if this flag is set); CAP10, CAP100,
CAP1000--capability advertised by link partner (valid only if
AUTO_NEG=1 && AUTO_NEG_PAR=0); CAP_HD_FLAG--capability
advertised by link partner (valid only if AUTO_NEG=1 &&
AUTO_NEG_PAR==0); or CAP PAUSE--capability advertised by link
partner (valid only if AUTO_NEG=1&& AUTO NE_PAR=0). Error
Codes may include, for example, OK or FATAL.
[0070] Drop NIC Port Link--0x0104. The Drop NIC Port Link command
may drop a link on the main port if it has not already been
dropped. The link may be re-established, for example, by using the
Set NIC Port command. Error Codes may include, for example, OK or
FATAL.
[0071] Filter All Packets--0x0201.
TABLE-US-00002 Byte 31 23 15 7 16 0x5706 Cmd. Seq. Num. 20 0x0201 4
24 0xFDFEFFFB 28-63 0
[0072] The Filter All Packets command may provide that packets
received on the main port may be filtered and may not be delivered
to the IMD 110. The command may be used, for example, as a power-up
setting and may be used with any filters (e.g., time filters) when
reset or re-loaded.
[0073] Filter All Packets Response--0x8201.
TABLE-US-00003 Byte 31 23 15 7 16 0x5706 Res. Seq. Num. 20 0x8201 4
24 0x7DFEFFFB 28-63 0
[0074] Filter based on settings--0x0202. Packets may be filtered
according to particular filter settings (e.g., current filter
settings).
[0075] Set Perfect Match Filter--0x0203. It may set the filters
defined by Fnum to the value specified in DA. Legal values for Fnum
are 0 and 1. If the "OFF" bit is set, then the specified perfect
match filter may be disabled.
TABLE-US-00004 Byte 31 23 15 7 16 0x5706 Cmd. Seq. Num. 20 0x0203 8
24 DA 28 DA OFF FNum 28 Checksum 32-63 0
[0076] Set Perfect Match Filter Response--0x8204.
[0077] Set Broadcast Filter--0x0205. The Set Broadcast Filter
command may be set if broadcast frames should be forwarded to the
IMD 110. Settings may include, for example, ON. Error Code may
include, for example, OK or FATAL.
[0078] Set ARP Filter--0x0206. The Set ARP Filter command may be
set if ARP frames should be forwarded to the IMD 110. Settings may
include, for example, ON. Error Code may include, for example, OK
or FATAL.
[0079] Get Statistics--0x0301. The Get Statistics command may
request a response with the requested statistics in it. Results may
include, for example:
TABLE-US-00005 u64_t IfHCInOctets; /* Collected from
emac_rx_stat_ifhcinoctets. */ /* This is the number of octets
received on the interface, including framing characters. */ u64_t
IfHCOutOctets; /* Collected from emac_tx_stat_ifhcoutoctets. */ /*
This is the number of octets that have been transmitted on the
interface. */ u64_t IfHCInUcastPkts; /* Collected from
emac_rx_stat_ifhcinucastpkts. */ /* This is the number of frames
received on the wire that were not dropped due to errors that have
Unicast Ethernet destination addresses. */ u64_t
IfHCInMulticastPkts; /* Collected from
emac_rx_stat_ifhcinmulticastpkts. */ /* This is the number of
frames received on the wire that were not dropped due to errors
that have multicast Ethernet destination addresses. */ u64_t
IfHCInBroadcastPkts; /* Collected from
emac_rx_stat_ifhcinbroadcastpkts. */ /* This is the number of
frames received on the wire that were not dropped due to errors
that have the broadcast Ethernet destination addresses. */ u64_t
IfHCOutUcastPkts; /* Collected from emac_tx_stat_ifhcoutucastpkts.
*/ /* This is the number of packets transmitted that have unicast
destination addresses. */ u64_t IfHCOutMulticastPkts; /* Collected
from emac_tx_stat_ifhcoutmulticastpkts. */ /* This is the number of
packets transmitted that have multicast destination addresses. */
u64_t IfHCOutBroadcastPkts; /* Collected from
emac_tx_stat_ifhcoutbroadcastpkts. */ /* This is the number of
packets transmitted that have the broadcast destination address. */
u32_t Dot3StatsCarrierSenseErrors; /* Collected from
emac_rx_stat_dot3statscarriersenseerrors. */ /* This is the number
of times a false carrier has been detected on the internal PHY
device. This is indicated from the PHY by asserting RXER while RXDV
is low when the RXD pins are at a state of 0x0e. */ u32_t
Dot3StatsFCSErrors; /* Collected from
emac_rx_stat_dot3statsfcserrors. */ /* A IfInErrors value is the
sum of this value and <b>Dot3StatsFCSErrors</b> and
<b>Dot3StatsAlignmentErrors</b>. */ u32_t
Dot3StatsAlignmentErrors; /* Collected from
emac_rx_stat_dot3statsalignmenterrors. */ /* This is the number of
frames received on the wire that have an odd number of nibbles and
fail FCS check and are of legal length. */ u32_t
Dot3StatsSingleCollisionFrames; /* Collected from
emac_tx_stat_dot3statssinglecollisionframes. */ /* This is the
number of collisions that were followed by successful packet
transmits. This is the same as the number of packets that were
transmitted with only one collision. */ u32_t
Dot3StatsMultipleCollisionFrames; /* Collected from
emac_tx_stat_dot3statsmultiplecollisionframes. */ /* This is the
number of packets that have transmitted with more that one
collision. */ u32_t Dot3StatsDeferredTransmissions; /* Collected
from emac_tx_stat_dot3statsdeferredtransmissionsl. */ /* This is
the number of packets that were delayed in transmission because
they had to wait for a RX packet to complete. */ u32_t
Dot3StatsExcessiveCollisions; /* Collected from
emac_tx_stat_dot3statsexcessivecollisions. */ /* This is the number
of packets that have been dropped due to having 16 collisions in a
row. */ u32_t Dot3StatsLateCollisions; /* Collected from
emac_tx_stat_dot3statslatecollisions. */ /* This is the number of
packets that have been dropped due to having a collision received
after the 64-byte collision window. */ u32_t EtherStatsCollisions;
/* Collected from emac_tx_stat_etherstatscollisions. */ /* This is
the number of collisions that have been detected on the interface.
*/ u32_t EtherStatsFragments; /* Collected from
emac_rx_stat_etherstatsfragments. */ /* This is the count of frames
less than 64 bytes with bad FCS. */ u32_t EtherStatsJabbers; /*
Collected from emac_rx_stat_etherstatsjabbers. */ /* This is the
number of frames received that exceed the programmed MTU size and
have bad FCS. */ u32_t EtherStatsUndersizePkts; /* Collected from
emac_rx_stat_etherstatsundersizepkts. */ /* This is the number of
frames received that are less than 64 bytes in length. */ u32_t
EtherStatsOverrsizePkts; /* Collected from
emac_rx_stat_dot3statsframestoolong. */ /* This is the number of
frames received that exceed the programmed MTU size. */ u32_t
EtherStatsPktsRx64Octets; /* Collected form
emac_rx_stat_etherstatspkts64octets. */ /* This is the number of
good frames received of 64 bytes in size. */ u32_t
EtherStatsPktsRx65Octetsto127Octets; /* Collected from
emac_rx_stat_etherstatspkts65octetsto127octets. */ /* This is the
number of good frames received of 65 bytes to 127 bytes in size. */
u32_t EtherStatsPktsRx128Octetsto255Octets; /* Collected from
emac_rx_stat_etherstatspkts128octetsto255octets. */ /* This is the
number of good frames received of 128 bytes to 255 bytes in size.
*/ u32_t EtherStatsPktsRx256Octetsto511Octets; /* Collected from
emac_rx_stat_etherstatspkts256octetsto511octets. */ /* This is the
number of good frames received of 256 bytes to 511 bytes in size.
*/ u32_t EtherStatsPktsRx512Octetsto1023Octets; /* Collected from
emac_rx_stat_etherstatspkts512octetsto1023octets. */ /* This is the
number of good frames received of 512 bytes to 1023 bytes in size.
*/ u32_t EtherStatsPktsRx1024Octetsto1522Octets; /* Collected from
emac_rx_stat_etherstatspkts1024octetsto1522octets. */ /* This is
the number of good frames received of 1024 bytes to 1522 bytes in
size. */ u32_t EtherStatsPktsRx1523Octetsto9022Octets; /* Collected
from emac_rx_stat_etherstatspkts1523octetsto9022octets. */ /* This
is the number of good frames received of 1523 bytes to 9022 bytes
in size. */ u32_t EtherStatsPktsTx64Octets; /* Collected form
emac_tx_stat_etherstatspkts64octets. */ /* This is the number of
good frames transmitted of 64 bytes in size. */ u32_t
EtherStatsPktsTx65Octetsto127Octets; /* Collected from
emac_tx_stat_etherstatspkts65octetsto127octets. */ /* This is the
number of good frames transmitted of 65 bytes to 127 bytes in size.
*/ u32_t EtherStatsPktsTx128Octetsto255Octets; /* Collected from
emac_tx_stat_etherstatspkts128octetsto255octets. */ /* This is the
number of good frames transmitted of 128 bytes to 255 bytes in
size. */ u32_t EtherStatsPktsTx256Octetsto511Octets; /* Collected
from emac_tx_stat_etherstatspkts256octetsto511octets. */ /* This is
the number of good frames transmitted of 256 bytes to 511 bytes in
size. */ u32_t EtherStatsPktsTx512Octetsto1023Octets; /* Collected
from emac_tx_stat_etherstatspkts512octetsto1023octets. */ /* This
is the number of good frames transmitted of 512 bytes to 1023 bytes
in size. */ u32_t EtherStatsPktsTx1024Octetsto1522Octets; /*
Collected from emac_tx_stat_etherstatspkts1024octetsto1522octets.
*/ /* This is the number of good frames transmitted of 1024 bytes
to 1522 bytes in size. */ u32_t
EtherStatsPktsTx1523Octetsto9022Octets; /* Collected from
emac_tx_stat_etherstatspkts1523octetsto9022octets. */ /* This is
the number of good frames transmitted of 1523 bytes to 9022 bytes
in size. */ u32_t XonPauseFramesReceived; /* Collected from
emac_rx_stat_xonpauseframesreceived. */ /* This is the number of
good MAC control frames received of pause type with a back-off
value of zero. This register increments regardless of flow control
state. */ u32_t XoffPauseFramesReceived; /* Collected from
emac_rx_stat_xoffpauseframesreceived. */ /* This is the number of
good MAC control frames received of pause type with a back-off
value other than zero. This register increments regardless of flow
control state. */ u32_t OutXonSent; /* Collected from
emac_tx_stat_outxonsent. */ /* This is the number of MAC Control
pause packets of value 0xffff that have be transmitted. */ u32_t
OutXoffSent; /* Collected from emac_tx_stat_outxoffsent. */ /* This
is the number of MAC Control pause packets of value 0 that have be
transmitted. */ u32_t MacControlFramesReceived; /* Collected from
emac_rx_stat_maccontrolframesreceived. */ /* This is the number of
good MAC control frames received that are not of pause type. */
Error Codes may include, for example, OK or FATAL.
[0080] While the present invention has been described with
reference to certain embodiments, it will be understood by those
skilled in the art that various changes may be made and equivalents
may be substituted without departing from the scope of the present
invention. In addition, many modifications may be made to adapt a
particular situation or material to the teachings of the present
invention without departing from its scope. Therefore, it is
intended that the present invention not be limited to the
particular embodiments disclosed, but that the present invention
will include all embodiments falling within the scope of the
appended claims.
* * * * *