Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant’s submission filed on February 2, 2026 has been entered.
Claim Rejections - 35 USC § 102
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
Claims 1, 4, 11, and 18 are rejected under 35 U.S.C 102(a)(1) as being anticipated by Kfir et al. (Patent No: US 2022/0060417 A1), hereinafter, Kfir.
Regarding Claim 1, Kfir teaches
A system for ethernet communication over an InfiniBand (IB) network, comprising: -Fig. 1B; Paragraph [0047] (Fig. 1B shows ethernet communication (from 1st Port, 136, 124) to over IB network (140) via second port (128). [0047] recites,” When an RDMA over Ethernet packet (or RoCE packet) is received at the first port 124, the processing circuitry 132 may be configured to strip content from the UDP header (if applicable), the IP header, and the MAC header and convert the stripped content into content for a GRH and LRH header, thereby enabling RDMA connectivity between the Ethernet network 136 and IB network 140. “)
at least one gateway comprising at least one processor to modify a network access layer of an ethernet communication which comprises at least a layer 3 protocol and a layer 4 protocol in addition to the network access layer, the modification to include a local route header (LRH) for an IB communication with the network access layer -Fig. 2; Paragraph [0045-0047][0040][0050] ([0040] recites,”… the processing circuitry 132 may include an Application Specific Integrated Circuit (ASIC), a Central Processing Unit (CPU), a General Processing Unit (GPU), a microprocessor, a Field Programmable Gate Array (FPGA), a collection of logic gates or transistors, resistors, capacitors, inductors, diodes, or the like. Some or all of the processing circuitry 132 may be provided on a Printed Circuit Board (PCB) or collection of PCBs.” [0050] recites, “…The gateway 204, first L2 bridge 208, first L3 router block 212, second L3 router block 216, and second L2 bridge 220 may all be provided as part of the switch device 104” [0045-0047] recites, “In the configuration of FIG. 1B, packets originating from the first host 108 may be formatted as RDMA over Ethernet packets. In this configuration, RDMA may correspond to the transport layer (L4) and may use the same headers as in the IB network 140. The packet may further include a User Datagram Protocol (UDP) header, an IP header, and a Media Access Control (MAC) header, where MAC is the Ethernet data-link layer (L2) and IP is the Ethernet network layer (L3)……… When an RDMA over Ethernet packet (or RoCE packet) is received at the first port 124, the processing circuitry 132 may be configured to strip content from the UDP header (if applicable), the IP header, and the MAC header and convert the stripped content into content for a GRH and LRH header, thereby enabling RDMA connectivity between the Ethernet network 136 and IB network 140. Conversely, when an RDMA over IB packet is received at the second port 128, the processing circuitry 132 may be configured to strip content from a GRH and LRH header and convert the stripped content into content for a MAC header and IP header, and optionally an UDP header, thereby enabling RDMA connectivity between the IB network 140 and Ethernet network 136.”)
and further to retain ethernet information of all layers of the ethernet communication
or to remove at least one of the layers of the ethernet communication for the IB communication. -Paragraph [0047] ([0047] recites, “When an RDMA over Ethernet packet (or RoCE packet) is received at the first port 124, the processing circuitry 132 may be configured to strip content from the UDP header (if applicable), the IP header, and the MAC header and convert the stripped content into content for a GRH and LRH header, thereby enabling RDMA connectivity between the Ethernet network 136 and IB network 140.”)
Regarding Claim 4, Kfir teaches the limitations of Claim 1.
Kfir further teaches,
The system of claim 1, further comprising the at least one gateway as a hardware component, a firmware component, or a software component that is associated with a host machine. -Fig. 1B, 2; Paragraph [0039][0050] ([0039] recites, “One or both of the first data port 124 and second data port 128 may include any appropriate type of socket, hardware, firmware, and/or software that is used to enable data communications between the switch device 104 and the hosts 108, 112.” [0050] recites, “The gateway 204, first L2 bridge 208, first L3 router block 212, second L3 router block 216, and second L2 bridge 220 may all be provided as part of the switch device 104 and, in some embodiments, may be provided as part of the processing circuitry 132.”)
Claim 11 is the method claim corresponding to the apparatus claim of Claim 1. Claim 1 is rejected above. Applicant’s attention is directed towards Claim 1. Claim 11 is rejected under the same rational of Claim 1
Regarding Claim 18, Kfir teaches,
A system comprising: one or more processing units to modify a network access layer of an ethernet communication -Fig. 1B; Paragraph [0047] (Fig. 1B shows ethernet communication (from 1st Port, 136, 124) to over IB network (140) via second port (128). [0047] recites,” When an RDMA over Ethernet packet (or RoCE packet) is received at the first port 124, the processing circuitry 132 may be configured to strip content from the UDP header (if applicable), the IP header, and the MAC header and convert the stripped content into content for a GRH and LRH header, thereby enabling RDMA connectivity between the Ethernet network 136 and IB network 140. “)
which comprises at least a layer 3 protocol and a layer 4 protocol in addition to the network access layer, the modification to include a local route header (LRH) for an InfiniBand (IB) communication for transmission over an IB network -Paragraph [0045-0047] ([0045-0047] recites, “In the configuration of FIG. 1B, packets originating from the first host 108 may be formatted as RDMA over Ethernet packets. In this configuration, RDMA may correspond to the transport layer (L4) and may use the same headers as in the IB network 140. The packet may further include a User Datagram Protocol (UDP) header, an IP header, and a Media Access Control (MAC) header, where MAC is the Ethernet data-link layer (L2) and IP is the Ethernet network layer (L3)……… When an RDMA over Ethernet packet (or RoCE packet) is received at the first port 124, the processing circuitry 132 may be configured to strip content from the UDP header (if applicable), the IP header, and the MAC header and convert the stripped content into content for a GRH and LRH header, thereby enabling RDMA connectivity between the Ethernet network 136 and IB network 140. Conversely, when an RDMA over IB packet is received at the second port 128, the processing circuitry 132 may be configured to strip content from a GRH and LRH header and convert the stripped content into content for a MAC header and IP header, and optionally an UDP header, thereby enabling RDMA connectivity between the IB network 140 and Ethernet network 136.”)
and further to retain ethernet information of all layers of the ethernet communication
or to remove at least one of the layers of the ethernet communication for the IB communication. -Paragraph [0047] ([0047] recites, “When an RDMA over Ethernet packet (or RoCE packet) is received at the first port 124, the processing circuitry 132 may be configured to strip content from the UDP header (if applicable), the IP header, and the MAC header and convert the stripped content into content for a GRH and LRH header, thereby enabling RDMA connectivity between the Ethernet network 136 and IB network 140.”)
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 2-3, 5-9, 12-16, and 19-20 are rejected under 35 U.S.C. 103 as being unpatentable over Kfir in view of Narasimhamurthy et al. (Patent No: US20160274926A1), hereinafter, Narasimhamurthy.
Regarding Claim 2, Kfir teaches the limitations of Claim 1.
Although implicit, Kfir does not explicitly mention,
The system of claim 1, wherein the at least one gateway is configured to perform the modification further by the inclusion of the LRH in at least a first part of the network access layer, by retaining a first portion of the ethernet information in at least one second part of the network access layer, and by retaining a second portion of the ethernet information in their respective layers for the IB communication.
However, in an analogous invention Narashimhamurthy teaches,
The system of claim 1, wherein the at least one gateway is configured to perform the modification further by the inclusion of the LRH in at least a first part of the network access layer, by retaining a first portion of the ethernet information in at least one second part of the network access layer, and by retaining a second portion of the ethernet information in their respective layers for the IB communication. -Fig. 9A; Paragraph [0095][0157] ([0095] recites, “The VXLAN packet is then sent from the first router (gateway 150) to the second router (gateway 160). As the packet is sent from gateway 150 to gateway 160, the MAC addresses (or layer-2 address, in general) for the source and destination in the outer Ethernet header are updated to reflect the corresponding MAC addresses for gateway 150 and gateway 160.” As shown in Fig. 9A ethernet information is retained in the corresponding network access layer (original MAC frame) and modification (addition) of IB header done on the retained ethernet information. [0157] recites, “The IB packet in FIG. 9A shows an IB header and the original MAC frame. The IB header contains a global route header 972 containing the source GID 974 and the destination GID 976. Additionally, local route header 970, base transport header 978, and other headers 980 all form part of the IB header.”)
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the ” GATEWAY FOR REMOTE DIRECT MEMORY ACCESS OVER CONVERGED ETHERNET” proposed by Kfir with the concept of Narasimhamurthy to include “an ethernet interface associated with a first local identifier (LID); an IB interface associated with a second local identifier (LID) to receive the ethernet communication from the ethernet interface and to query a subnet manager (SM) via a subnet management agent (SMA) for a destination LID (DLID) to be provided in a global unique identifier (GUID) table; and a subnet manager (SM) to communicate the GUID table to the SMA to enable the at least one gateway to prepare the LRH for the IB communication from information in the GUID table and to enable the at least one gateway to transmit the IB communication through the IB network.” One of ordinary skill in the art would have been motivated to make this modification in order for efficient allocation of computing resources [0007].
Regarding Claim 3, Kfir teaches the limitations of Claim 1.
Although implicit, Kfir does not explicitly mention,
The system of claim 1, wherein the at least one gateway is to perform the modification further by inclusion, within an application layer of the IB communication, of the ethernet information of the at least one of the layers of the ethernet communication that is removed.
However, in an analogous invention Narashimhamurthy teaches,
The system of claim 1, wherein the at least one gateway is to perform the modification further by inclusion, within an application layer of the IB communication, of the ethernet information of the at least one of the layers of the ethernet communication that is removed. -Fig. 5 (550, 560, 570); Paragraph [0095][0097-0098][0102] ([0095] recites, “The VXLAN packet is then sent from the first router (gateway 150) to the second router (gateway 160). As the packet is sent from gateway 150 to gateway 160, the MAC addresses (or layer-2 address, in general) for the source and destination in the outer Ethernet header are updated to reflect the corresponding MAC addresses for gateway 150 and gateway 160. “[0097-0098] recites, “In step 550, bridge 137 removes the VXLAN header from the VXLAN packet to recover the original MAC frame. In step 560, bridge 137 adds a second IB header to the MAC frame to create a second IB packet. The second 113 packet contains the original MAC frame along with the second 113 header. It should be appreciated that the IB header facilitates the packet to be switched for the switched fabric, and thus the MAC frame with the IB header may be referred to as a ‘switched packet’.” [0102] recites, “In particular, in the event of the data payload arriving at the non-switched-fabric based computing system encapsulated in a VXLAN packet, the data payload may be processed using corresponding technology implemented on the external system (140) for forwarding the original MAC frame to the destination VM. For example, in remote farm 140 that represents a non-switched-fabric based computing system such as an Ethernet-based local area network, local network 146 may rely on Ethernet broadcast to send the MAC frame of step 550 to the destination VM, as would be readily understood by those skilled in the relevant arts.”)
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the ” GATEWAY FOR REMOTE DIRECT MEMORY ACCESS OVER CONVERGED ETHERNET” proposed by Kfir with the concept of Narasimhamurthy to include “at least one gateway is to perform the modification further by inclusion, within an application layer of the IB communication, of the ethernet information of the at least one of the layers of the ethernet communication that is removed.” One of ordinary skill in the art would have been motivated to make this modification in order for efficient allocation of computing resources [0007].
Regarding Claim 5, Kfir teaches the limitations of Claim 1.
Although implicit, Kfir does not explicitly mention,
The system of claim 1, further comprising: a subnet management agent (SMA) associated with a data processing unit (DPU) and with individual physical ports of the at least one gateway; anda subnet manager (SM) to communicate destination information with the SMA to enable the at least one gateway to prepare the LRH for the IB communication.
However, in an analogous invention Narashimhamurthy teaches,
The system of claim 1, further comprising: a subnet management agent (SMA) associated with a data processing unit (DPU) and with individual physical ports of the at least one gateway; -Fig. 1, 3, 8C; Paragraph [0062][0131] (As shown in Fig 3. It is easily understandable to an ordinary person with the skill in the art that Fabric Controller (310) may use as subnet management agent (SMA).[0062] recites, “FIG. 3 is a block diagram illustrating the detailed architecture of switched fabric 106, bridge 107, and computing node 105A of appliance 110 in one embodiment. Computing node 105A is shown containing VMs 391-393, and virtual network interface cards (vNIC) 399A-399C. Switched fabric 106 is shown containing crossbar 370, fabric controller 310, subnet manager 320, migration list 330, and forwarding table 340. VM management system 190 is shown with paths 191, 192A and 192B. Bridge 107 is shown containing network manager 380, VTEP table 350, and bridge table 360. Each of the blocks is described below in further detail.” Fig. 8C shows port mapping and forwarding table for example. [0130] recites, “Row 803 specifies that the native VM with a MID of GID6 is mapped to IB Port 11 (within appliance 130) such that all data addressed to VM GID6 will be forwarded by switched fabric 136 to IB Port 11. Row 804 specifies that bridge 137 with a MID of GID40 is accessible via IB Port 20 such that all data sent to the bridge will be forwarded by switched fabric 136 to IB Port 20. Row 805 contains GID and port information for the in-migrated VM GID2A (a new value assigned to the VM by a corresponding subnet manager in appliance 130, upon in-migration from appliance 110). Row 805 shows VM GID2A being connected to IB Port 11. As with the other VM on computing node 135A (i.e., GID6), the newly migrated VM GID2A is also assigned the same port (i.e., IB Port 11) on switched fabric 136. IB Port 11 now accepts all data directed towards the newly migrated VM GID2A.”)
and a subnet manager (SM) to communicate destination information with the SMA to enable the at least one gateway to prepare the LRH for the IB communication. -Paragraph [0070][0095] ([0075] recites, “Fabric controller 310 controls the switching operation of crossbar 370 by issuing the appropriate control signals on path 317. Fabric controller 310 receives data signals on path 303 indicating the MID of source VM (from which packet is being received) and the MID of destination VM (to which packet is to be sent). Fabric controller switches the packets to the appropriate output port based on data stored in forwarding table 340.” As shown in Fig. 3, subnet manager (SM) updates the forwarding table and Fabric Controller (SMA in this case) through appropriate bridge and gateway help prepare gateway to add local route header (LRH) and for IB communication)
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the ” GATEWAY FOR REMOTE DIRECT MEMORY ACCESS OVER CONVERGED ETHERNET” proposed by Kfir with the concept of Narasimhamurthy to include “a subnet management agent (SMA) associated with a data processing unit (DPU) and with individual physical ports of the at least one gateway; and a subnet manager (SM) to communicate destination information with the SMA to enable the at least one gateway to prepare the LRH for the IB communication.” One of ordinary skill in the art would have been motivated to make this modification in order for efficient allocation of computing resources [0007].
Regarding Claim 6, Kfir and Narashimhamurthy teach the limitations of Claim 5.
Although implicit, Kfir does not explicitly mention,
The system of claim 5, wherein the at least one gateway is further configured to: query the SM for the destination information associated with a destination media access control (MAC) that is associated with a destination of the IB communication, wherein the SM communicates the destination information in response to the query.
However, in an analogous invention Narashimhamurthy teaches,
a) The system of claim 5, wherein the at least one gateway is further configured to: query the SM for the destination information associated with a destination media access control (MAC) that is associated with a destination of the IB communication, wherein the SM communicates the destination information in response to the query. -Fig. 1; Paragraph [0144-0146] ([0144-0146] recites, “After the re-set of internal routing tables, and prior to sending a data packet for the first time, an Address Resolution Protocol (ARP) request is issued by VM GID2A to determine the address parameters (MID and MAC address) of the destination VM (i.e., the intended recipient of the data packet). The ARP request is processed by bridge 137, which creates a multicast IP packet with the ARP request in it and forwards the multicast to all VTEPs in the VNI over which VM GID2A communicates. Thereafter, bridge 107 receives the multicast packet and sends the ARP request to VM GID1. VM GID1 then creates a unicast ARP reply and sends the reply to bridge 107, which embeds the reply in a VxLAN packet and sends it to bridge 137. Upon receiving the unicast packet, bridge 137 creates an entry in its VTEP table (i.e., entry 806 in FIG. 8D) and stores the MAC address of VM GID1 and the IP address of the bridge 107 (IP-VTEP1) in the VTEP table. Bridge 137 then sends the MAC address of VM GID1 to VM GID2A as a response to the ARP request. The ARP response may further specify that the MID of bridge 137 (i.e., GID40) as being the MID of GID1 (even though VM GID1 has some other MID within appliance 110). In general, the MID of all VMs in native appliance 110 (from which VMs have migrated in to appliance 130) is mapped to GID40 in appliance 130.”)
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the ” GATEWAY FOR REMOTE DIRECT MEMORY ACCESS OVER CONVERGED ETHERNET” proposed by Kfir with the concept of Narasimhamurthy to include “at least one gateway is further configured to: query the SM for the destination information associated with a destination media access control (MAC) that is associated with a destination of the IB communication, wherein the SM communicates the destination information in response to the query.” One of ordinary skill in the art would have been motivated to make this modification in order for efficient allocation of computing resources [0007].
Regarding Claim 7, Kfir and Narashimhamurthy teach the limitations of Claim 6.
Although implicit, Kfir does not explicitly mention,
The system of claim 6, wherein the query comprises a destination local identifier (DLID) of a destination IB device associated with the destination MAC.
However, in an analogous invention Narashimhamurthy teaches,
The system of claim 6, wherein the query comprises a destination local identifier (DLID) of a destination IB device associated with the destination MAC identifier. -Paragraph [0063-0064] ([0063-0064] recites, “Each VM 391-393 is shown communicating via a respective vNIC of vNICs 399A-399C (on a respective path of paths 394A-394C), and may thus have a medium access control (MAC) address and IP address. As described below, the MAC address is retained even after migration, and used as the basis for migrating the corresponding VM to external systems. Each VM is assumed to be connected to a corresponding port of switched fabric 106, and identified by a respective machine identifier (MID), which is a unique number within the corresponding appliance. A MID may be represented by a global ID (GID) and/or a local ID (LID) depending on the particular implementation of the VMs in the IB fabric of appliance 110. To illustrate, the IB fabric of appliance 110 may be implemented using one or more IB subnets. An IB subnet is a group of ports available in an IB fabric, where the combination of such one or more IB subnets defines all the available ports in that particular IB fabric.” As explained above all appliances (device) of IB fabric may have an associated local identifier (LID), i.e., destination local identifier (DLID) corresponding to the MAC address of the destination device. )
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the ” GATEWAY FOR REMOTE DIRECT MEMORY ACCESS OVER CONVERGED ETHERNET” proposed by Kfir with the concept of Narasimhamurthy to include “the query comprises a destination local identifier (DLID) of a destination IB device associated with the destination MAC.” One of ordinary skill in the art would have been motivated to make this modification in order for efficient allocation of computing resources [0007].
Regarding Claim 8, Kfir teaches the limitations of Claim 1.
Although implicit, Kfir does not explicitly mention,
The system of claim 1, further comprising: an IB interface of the at least one gateway to receive a second IB communication; and an SMA associated with the IB interface to enable removal of an associated LRH from the second IB communication, wherein removal of the associated LRH provides a second ethernet communication, and wherein the second ethernet communication is to be forwarded through one or more ethernet devices to a destination ethernet device according to a MAC identifier associated with the second ethernet communication.
However, in an analogous invention, Narasimhamurthy teaches,
The system of claim 1, further comprising: an IB interface of the at least one gateway to receive a second IB communication; -Fig. 1; Paragraph [0041] ([0041] recites, “Each of the gateways 150, 160, and 170 represents a router that routes each IP datagram from a source IP system to a destination IP system via Internet 120, based on the IP address in the IP destination field. Specifically, a gateway processes an incoming datagram by examining the datagram for a destination IP address, and forwards the datagram towards that destination system based on the IP address and routing information. “
and an SMA associated with the IB interface to enable removal of an associated LRH from the second IB communication, wherein removal of the associated LRH provides a second ethernet communication, and wherein the second ethernet communication is to be forwarded through one or more ethernet devices to a destination ethernet device according to a MAC identifier associated with the second ethernet communication. -Fig. 5 (550, 560, 570, 580); Paragraph [0096-0100] ([0096-0100] recites, “In step 540, the VXLAN packet is transmitted to the second bridge 137 as the end point of the VXLAN connection from the second router, gateway 160. Second bridge 137 is part of the external system, i.e., appliance 130, hosting the migrated VM GID2. Second bridge 137 is in communication with a corresponding fabric controller in appliance 130, which in turn communicates with the destination VM GID2 through a corresponding crossbar framework. In step 550, bridge 137 removes the VXLAN header from the VXLAN packet to recover the original MAC frame. In step 560, bridge 137 adds a second IB header to the MAC frame to create a second IB packet. The second 113 packet contains the original MAC frame along with the second 113 header. It should be appreciated that the IB header facilitates the packet to be switched for the switched fabric, and thus the MAC frame with the IB header may be referred to as a ‘switched packet’. In step 570, bridge 137 sends the second IB packet to a switched fabric 136 connected to the second VM GID2. The corresponding fabric controller in appliance 130 switches the packet to VM GID2 based on the content of a corresponding forwarding table. In step 580, the second 113 header is removed from the second IB packet (after completion of switching by the switch fabric 106) and the resulting MAC frame is transmitted to the second VM GID2. The flow chart ends in step 599.” Fig. 5 shows the flow chart where bridge together with fabric controller removes the LRH and sends to the destination based on the MAC address)
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the ” GATEWAY FOR REMOTE DIRECT MEMORY ACCESS OVER CONVERGED ETHERNET” proposed by Kfir with the concept of Narasimhamurthy to include “an IB interface of the at least one gateway to receive a second IB communication; and an SMA associated with the IB interface to enable removal of an associated LRH from the second IB communication, wherein removal of the associated LRH provides a second ethernet communication, and wherein the second ethernet communication is to be forwarded through one or more ethernet devices to a destination ethernet device according to a MAC identifier associated with the second ethernet communication.” One of ordinary skill in the art would have been motivated to make this modification in order for efficient allocation of computing resources [0007].
Regarding Claim 9, Kfir and Narasimhamurthy teach the limitations of Claim 8
Although implicit, Kfir does not explicitly teach,
The system of claim 8, wherein the at least one gateway is further configured to: enable the second ethernet communication to include a source MAC of the IB interface and a destination MAC of the destination ethernet device, wherein the source MAC and the destination MAC are obtained by a query to a subnet manager (SM) or is included in an application layer of the IB communication.
However, in an analogous invention Narasimhamurthy teaches,
The system of claim 8, wherein the at least one gateway is further configured to: enable the second ethernet communication to include a source MAC of the IB interface and a destination MAC of the destination ethernet device, -Paragraph [0090] ([0090] recites, “In step 510, vNIC 399A on appliance 110 sends a first IB packet to VM GID2 that has been migrated to the external system appliance 130. It is assumed that the data payload of the IB packet is received on path 394A from VM GID1. The first IB packet contains the data payload, a MAC header, and a first IB header. The data payload and the MAC header together may be viewed as a MAC frame. The MAC header specifies the MAC addresses for the source VM and the destination VM…”
wherein the source MAC and the destination MAC are obtained by a query to a subnet manager (SM)
or is included in an application layer of the IB communication. -Paragraph [0090] ([0090] recites, “…The IB header specifies GID1 as the source MID and GID2 as the destination MID….”)
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the ” GATEWAY FOR REMOTE DIRECT MEMORY ACCESS OVER CONVERGED ETHERNET” proposed by Kfir with the concept of Narasimhamurthy to include “at least one gateway is further configured to: enable the second ethernet communication to include a source MAC of the IB interface and a destination MAC of the destination ethernet device, wherein the source MAC and the destination MAC are obtained by a query to a subnet manager (SM) or is included in an application layer of the IB communication.” One of ordinary skill in the art would have been motivated to make this modification in order for efficient allocation of computing resources [0007].
Claim 12 is the method claim corresponding to the apparatus claim of Claim 2. Claim 2 is rejected above. Applicant’s attention is directed towards Claim 2. Claim 12 is rejected under the same rational of Claim 2
Claim 13 is the method claim corresponding to the apparatus claim of Claim 3. Claim 3 is rejected above. Applicant’s attention is directed towards Claim 3. Claim 13 is rejected under the same rational of Claim 3
Claim 14 is the method claim corresponding to the apparatus claim of Claim 6. Claim 6 is rejected above. Applicant’s attention is directed towards Claim 6. Claim 14 is rejected under the same rational of Claim 6
Claim 15 is the method claim corresponding to the apparatus claim of Claim 8. Claim 8 is rejected above. Applicant’s attention is directed towards Claim 8. Claim 15 is rejected under the same rational of Claim 8
Claim 16 is the method claim corresponding to the apparatus claim of Claim 9. Claim 9 is rejected above. Applicant’s attention is directed towards Claim 9. Claim 16 is rejected under the same rational of Claim 9
Claim 19 is same as Claim 2 except Claim 19 is dependent claim of independent Claim 18 instead of Claim 1. Claim 2 has been rejected above. Applicant’s attention is directed towards Claim 2. Claim 19 is rejected under the same rational as Claim 2
Claim 20 is same as Claim 3 except Claim 20 is dependent claim of independent Claim 18 instead of Claim 1. Claim 3 has been rejected above. Applicant’s attention is directed towards Claim 3. Claim 20 is rejected under the same rational as Claim 3
Claim 10 and 17 are rejected under 35 U.S.C. 103 as being unpatentable over Kfir in view of JOHNSEN et al. (Patent No: 20220263892A1), hereinafter, JOHNSEN.
Regarding Claim 10, Kfir teaches the limitations of Claim 1.
Although implicit, Kfir does not explicitly teach,
The system of claim 1, further comprising: an ethernet interface associated with a first local identifier (LID); an IB interface associated with a second local identifier (LID) to receive the ethernet communication from the ethernet interface and to query a subnet manager (SM) via a subnet management agent (SMA) for a destination LID (DLID) to be provided in a global unique identifier (GUID) table; and a subnet manager (SM) to communicate the GUID table to the SMA to enable the at least one gateway to prepare the LRH for the IB communication from information in the GUID table and to enable the at least one gateway to transmit the IB communication through the IB network.
However, in an analogous invention JOHNSEN teaches,
The system of claim 1, further comprising: an ethernet interface associated with a first local identifier (LID); -Paragraph [0056] ([0056] recites, “ IB uses three different types of addresses. A first type of address is the 16 bits Local Identifier (LID). At least one unique LID is assigned to each HCA port and each switch by the SM..”
an IB interface associated with a second local identifier (LID) to receive the ethernet communication from the ethernet interface and to query a subnet manager (SM) via a subnet management agent (SMA) for a destination LID (DLID) to be provided in a global unique identifier (GUID) table; -Paragraph [0042-0045] ([0042-0043] recites, “Through the subnet management interface, the subnet manager exchanges control packets, which are referred to as subnet management packets (SMPs), with subnet management agents (SMAs). The subnet management agents reside on every IB subnet device. By using SMPs, the subnet manager is able to discover the fabric, configure end nodes and switches, and receive notifications from SMAs. In accordance with an embodiment, intra-subnet routing in an IB network can be based on linear forwarding tables (LFTs) stored in the switches. The LFTs are calculated by the SM according to the routing mechanism in use. In a subnet, Host Channel Adapter (HCA) ports on the end nodes and switches are addressed using local identifiers (LIDs). Each entry in a linear forwarding table (LFT) consists of a destination LID (DLID) and an output port. Only one entry per LID in the table is supported. When a packet arrives at a switch, its output port is determined by looking up the DLID in the forwarding table of the switch. The routing is deterministic as packets take the same path in the network between a given source-destination pair (LID pair).”[0045] recites,” Furthermore, hosts and switches within a subnet can be addressed using local identifiers (LIDs), and a single subnet can be limited to 49151 unicast LIDs. Besides the LIDs, which are the local addresses that are valid within a subnet, each IB device can have a 64-bit global unique identifier (GUID). A GUID can be used to form a global identifier (GID), which is an IB layer three (L3) address.”)
and a subnet manager (SM) to communicate the GUID table to the SMA to enable the at least one gateway to prepare the LRH for the IB communication from information in the GUID table and to enable the at least one gateway to transmit the IB communication through the IB network. -Paragraph [0056][0118-0119] ([0056] recites, “The SM may assign additional subnet unique GUIDs to an HCA port, which is useful when SR-IOV is used…” And communicates to SMA the GUID table. [0118-0119] recites, “..a dual port router abstraction can provide a simple way for enabling subnet-to-subnet router functionality to be defined based on a switch hardware implementation that has the ability to do GRH (global route header) to LRH (local route header) conversion in addition to performing normal LRH based switching. In accordance with an embodiment, a virtual dual-port router can logically be connected outside a corresponding switch port. This virtual dual-port router can provide an InfiniBand specification compliant view to a standard management entity, such as a Subnet Manager.” Router acts as a gateway.
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the ” GATEWAY FOR REMOTE DIRECT MEMORY ACCESS OVER CONVERGED ETHERNET” proposed by Kfir with the concept of JOHNSEN to include “an ethernet interface associated with a first local identifier (LID); an IB interface associated with a second local identifier (LID) to receive the ethernet communication from the ethernet interface and to query a subnet manager (SM) via a subnet management agent (SMA) for a destination LID (DLID) to be provided in a global unique identifier (GUID) table; and a subnet manager (SM) to communicate the GUID table to the SMA to enable the at least one gateway to prepare the LRH for the IB communication from information in the GUID table and to enable the at least one gateway to transmit the IB communication through the IB network.” One of ordinary skill in the art would have been motivated to make this modification in order to improve performance, scalability and flexibility of the (virtualized) I/O resources to match the level of performance seen in modern CPU virtualization [0066]
Claim 17 is the method claim corresponding to the apparatus claim of Claim 10. Claim 10 is rejected above. Applicant’s attention is directed towards Claim 10. Claim 17 is rejected under the same rational of Claim 10
Response to Argument(s)
Applicant's argument(s) filed on February 28, 2025 have been fully considered but they are not persuasive. Therefore, the Examiner regretfully maintains the rejection.
Regarding Claim 1, the Applicant argues,
Applicant's claim 1 is under a section 102 rejection and recites:
A system for ethernet communication over an InfiniBand (IB) network, comprising:
at least one gateway comprising at least one processor to modify a network access layer of an ethernet communication which comprises at least a layer 3 protocol and a layer 4 protocol in addition to the network access layer, the modification to include a local route header (LRH) for an IB communication with the network access layer and further to retain ethernet information of all layers of the ethernet communication or to remove at least one of the layers of the ethernet communication for the IB communication.
Such subject matter is not found in the cited reference asserted in the Office Action. For example, as appreciated in the Examiner Interview, the cited reference under section 102 at least fails to disclose:
at least one gateway comprising at least one processor to modify a network access layer of an ethernet communication which comprises at least a layer 3 protocol and a layer 4 protocol in addition to the network access layer, the modification to include a local route header (LRH) for an IB communication with the network access layer and further to retain ethernet information of all layers of the ethernet communication or to remove at least one of the layers of the ethernet communication for the IB communication.
See claim 1.
Instead, as appreciated in the Examiner Interview, Kfir reports a process to strip content from the MAC header which is alleged as the network access layer and to convert the stripped content to a GRH and LRH header:
[T]he processing circuitry [is] configured to strip content from the
UDP header (if applicable), the IP header, and the MAC header
and convert the stripped content into content for a GRH and LRH
header, thereby enabling RDMA connectivity between the Ethernet
network 136 and IB network 140.
Kfir, 0047. The plain and ordinary interpretation available in Kfir is that the MAC no longer exists when the reference expressly strips and converts the MAC header (and other headers) into content for a GRH and LRH header. This is contrary to Applicant's slim communication by reduced-byte Ethernet communication which specifically includes the LRH with the network access layer and, additionally, removes a layer or joins the remaining layers with other parts of a header for IB communication ("the modification to include a local route header (LRH) for an IB communication with the network access layer," as in claim 1). -Page (8-9)
The Examiner’s response is the following:
The Examiner respectfully totally disagree with the Applicant and believes all parts of claim 1 are completely addressed with wording as close as possible and with the similarity easily understandable to an ordinary person with the skill in the art. The Examiner broke down the limitations and mapped each limitation with recitation from the prior-art (Kfir) as described above and also, noted down in the following for convenience.
A system for ethernet communication over an InfiniBand (IB) network, comprising: -Fig. 1B; Paragraph [0047] (Fig. 1B shows ethernet communication (from 1st Port, 136, 124) to over IB network (140) via second port (128). [0047] recites,” When an RDMA over Ethernet packet (or RoCE packet) is received at the first port 124, the processing circuitry 132 may be configured to strip content from the UDP header (if applicable), the IP header, and the MAC header and convert the stripped content into content for a GRH and LRH header, thereby enabling RDMA connectivity between the Ethernet network 136 and IB network 140. “)
at least one gateway comprising at least one processor to modify a network access layer of an ethernet communication which comprises at least a layer 3 protocol and a layer 4 protocol in addition to the network access layer, the modification to include a local route header (LRH) for an IB communication with the network access layer -Fig. 2; Paragraph [0045-0047][0040][0050] ([0040] recites,”… the processing circuitry 132 may include an Application Specific Integrated Circuit (ASIC), a Central Processing Unit (CPU), a General Processing Unit (GPU), a microprocessor, a Field Programmable Gate Array (FPGA), a collection of logic gates or transistors, resistors, capacitors, inductors, diodes, or the like. Some or all of the processing circuitry 132 may be provided on a Printed Circuit Board (PCB) or collection of PCBs.” [0050] recites, “…The gateway 204, first L2 bridge 208, first L3 router block 212, second L3 router block 216, and second L2 bridge 220 may all be provided as part of the switch device 104” [0045-0047] recites, “In the configuration of FIG. 1B, packets originating from the first host 108 may be formatted as RDMA over Ethernet packets. In this configuration, RDMA may correspond to the transport layer (L4) and may use the same headers as in the IB network 140. The packet may further include a User Datagram Protocol (UDP) header, an IP header, and a Media Access Control (MAC) header, where MAC is the Ethernet data-link layer (L2) and IP is the Ethernet network layer (L3)……… When an RDMA over Ethernet packet (or RoCE packet) is received at the first port 124, the processing circuitry 132 may be configured to strip content from the UDP header (if applicable), the IP header, and the MAC header and convert the stripped content into content for a GRH and LRH header, thereby enabling RDMA connectivity between the Ethernet network 136 and IB network 140. Conversely, when an RDMA over IB packet is received at the second port 128, the processing circuitry 132 may be configured to strip content from a GRH and LRH header and convert the stripped content into content for a MAC header and IP header, and optionally an UDP header, thereby enabling RDMA connectivity between the IB network 140 and Ethernet network 136.” In the above, it is clearly states that the communication comprises layer 3, 4 protocols in addition to network access layer and convert/modify the stripped content for a GRH and LRH for IB communication. It is understandable to any ordinary person that convert and modify are synonym here. )
and further to retain ethernet information of all layers of the ethernet communication
or to remove at least one of the layers of the ethernet communication for the IB communication. -Paragraph [0047] ([0047] recites, “When an RDMA over Ethernet packet (or RoCE packet) is received at the first port 124, the processing circuitry 132 may be configured to strip content from the UDP header (if applicable), the IP header, and the MAC header and convert the stripped content into content for a GRH and LRH header, thereby enabling RDMA connectivity between the Ethernet network 136 and IB network 140.” It is easily understandable to an ordinary person to understand stripping and removing are synonyms.)
As was explained during the interview and the examiner explained again above that all the limitations of Claim 1 are addressed by the prior-art used and the examiner regretfully maintains the rejection status.
The Applicant Argues that independent claims 11 and 18 are allowable for the same reason as the applicant thinks Claim 1 should be allowable (Page-12)
The Examiner’s response is as the examiner explained above why he thinks Claim 1 is rejected with explanation, independent Claim 11 and 18 are rejected under the same rational as Claim 1.
The Applicant argues that dependent Claims 2-10, 12-17, and 19-20 are allowable as they directly or indirectly depend on one of independent claims 1, 11 and 18. (Page-12)
The Examiner’s response is as the examiner believes that claims 1, 11, and 18 are rejected for reasons explained above, the logic of the applicant does not hold and the examiner maintains the rejection status.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to AHMED SAIFUDDIN whose telephone number is (703)756-4581. The examiner can normally be reached Monday-Friday 8:30am-6:00pm.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, KHALED M KASSIM can be reached on 571-270-3770. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/AHMED SAIFUDDIN/Examiner, Art Unit 2475
/ABDULLAHI AHMED/Examiner, Art Unit 2475