[Mellanox] [SKU] Mellanox-SN4700-C128 SKU added by vivekrnv · Pull Request #24 · vivekrnv/sonic-buildimage

vivekrnv · 2022-07-20T06:51:29Z

Signed-off-by: Vivek Reddy Karri vkarri@nvidia.com

Why I did it

Draft a new SKU for MSN-4700 Platform i.e. Mellanox-SN4700-C128

Requirements:

Breakout: Port 1-32: 4x100G
Downlinks: 120 (1-30)
Uplinks: 8 (31-32)
Shared Headroom: Enabled
Over Subscribe Ratio: 1:8
Default Topology: T2
Default Cable Length for T2: 1500m
VxLAN source port range set: yes
QoS params: The default ones defined in qos_config.j2 will be applied
Small Packet Percentage: Used 50% for traditional buffer model Note: For dynamic model, the value defined in LOSSLESS_TRAFFIC_PATTERN|AZURE|small_packet_percentage is used

Additional Details:

Switch Type has to be programmed as SpineRouter through config_db.json in DEVICE_METADATA|localhost|type field for the buffer values & cable lengths defined in the buffers_defaults_t2.j2 to apply on the device
Cable Lengths Used for generating buffer_defaults_{t0,t1,t2}.j2 values

 {
        "t2": {
            "up": "1500m",
            "down": "1500m",       
         },
        "t1": {
            "up": "300m",
            "down": "40m",
        },
        "t0": {
            "up": "40m",
            "down": "5m",
        }
}

How I did it

How to verify it

Which release branch to backport (provide reason below if selected)

Description for the changelog

Link to config_db schema for YANG module changes

A picture of a cute animal (not mandatory but encouraged)

Signed-off-by: Vivek Reddy Karri <vkarri@nvidia.com>

stephenxs · 2022-07-21T02:30:35Z

don't suggest exposing this Egress Mirroring: Disabled to the description as this is Mellanox's internal logic.
Over Subscribe Ratio: 1:8 this is provided by customer, right?

vivekrnv · 2022-07-21T02:35:04Z

don't suggest exposing this Egress Mirroring: Disabled to the description as this is Mellanox's internal logic.
Over Subscribe Ratio: 1:8 this is provided by customer, right?

Yeah

stephenxs · 2022-07-21T02:48:06Z

device/mellanox/x86_64-mlnx_msn4700-r0/Mellanox-SN4700-C128/pg_profile_lookup.ini

@@ -0,0 +1,41 @@
+##


the numbers are correct but we do not need the numbers for speed 200000 since there is no 200G port on this SKU.

the numbers are correct but we do not need the numbers for speed 200000 since there is no 200G port on this SKU.

customer asked us to add it. Maybe for future use.

let's keep it then. 200G for 4x ports right?

stephenxs · 2022-07-21T03:10:08Z

device/mellanox/x86_64-mlnx_msn4700-r0/Mellanox-SN4700-C128/buffers_defaults_t2.j2

+{% set ingress_lossless_pool_size =  '36405216' %}
+{% set ingress_lossless_pool_xoff  =  '11567088' %}
+{% set egress_lossless_pool_size =  '60817392' %}
+{% set egress_lossy_pool_size =  '36405216' %}


As a convention, we round down by cell size for egress_lossless_pool only. For the rest sizes, we do not round them down.

Suggested change

{% set ingress_lossless_pool_size = '36405216' %}

{% set ingress_lossless_pool_xoff = '11567088' %}

{% set egress_lossless_pool_size = '60817392' %}

{% set egress_lossy_pool_size = '36405216' %}

{% set ingress_lossless_pool_size = '36405248' %}

{% set ingress_lossless_pool_xoff = '11567104' %}

{% set egress_lossless_pool_size = '60817392' %}

{% set egress_lossy_pool_size = '36405248' %}

stephenxs · 2022-07-21T03:11:25Z

don't suggest exposing this Egress Mirroring: Disabled to the description as this is Mellanox's internal logic. Over Subscribe Ratio: 1:8 this is provided by customer, right?

Just to double check, they do not use features like everflow, etc, right

vivekrnv · 2022-07-21T05:47:15Z

don't suggest exposing this Egress Mirroring: Disabled to the description as this is Mellanox's internal logic. Over Subscribe Ratio: 1:8 this is provided by customer, right?

Just to double check, they do not use features like everflow, etc, right

I think no. But Assaf, Can you confirm?

assafho · 2022-07-21T06:18:47Z

don't suggest exposing this Egress Mirroring: Disabled to the description as this is Mellanox's internal logic. Over Subscribe Ratio: 1:8 this is provided by customer, right?

Just to double check, they do not use features like everflow, etc, right

I think no. But Assaf, Can you confirm?

Yes, customer asked for ratio of 1:8. Regarding everflow, i guess not, but its not part of hwsku questions.

stephenxs · 2022-07-21T06:29:43Z

don't suggest exposing this Egress Mirroring: Disabled to the description as this is Mellanox's internal logic. Over Subscribe Ratio: 1:8 this is provided by customer, right?

Just to double check, they do not use features like everflow, etc, right

I think no. But Assaf, Can you confirm?

Yes, customer asked for ratio of 1:8. Regarding everflow, i guess not, but its not part of hwsku questions.

Thanks @assafho . Better to know it. this is something we missed in hwsku questions.

Signed-off-by: Vivek Reddy Karri <vkarri@nvidia.com>

vivekrnv · 2022-07-21T07:32:49Z

@stephenxs, Apart from clarification on mirroring, is anything left for buffers?

stephenxs · 2022-07-21T09:54:57Z

@stephenxs, Apart from clarification on mirroring, is anything left for buffers?

No. all the rest LGTM.

Signed-off-by: Vivek Reddy Karri <vkarri@nvidia.com>

…to sku_4700_c128

vivekrnv · 2022-07-23T01:38:33Z

Mirroring is required and updated buffer defaults

dgsudharsan · 2022-07-23T03:55:23Z

device/mellanox/x86_64-mlnx_msn4700-r0/Mellanox-SN4700-C128/sai.profile

@@ -0,0 +1,4 @@
+SAI_VXLAN_SRCPORT_RANGE_ENABLE=1


Why do we need SAI_VXLAN_SRCPORT_RANGE_ENABLE? I believe this is removed in latest branches. Please check

Yes, it was removed. Will update

dgsudharsan · 2022-07-23T03:56:50Z

device/mellanox/x86_64-mlnx_msn4700-r0/Mellanox-SN4700-C128/sai_4700_128x100g.xml

+                <module>17</module>
+                <breakout-modes>3</breakout-modes>
+                <port-speed>1536</port-speed>
+            <split>4</split></port-info>


The split and closing port info are not properly aligned. Can you please align?

Signed-off-by: Vivek Reddy Karri <vkarri@nvidia.com>

Changes: Update submodule branch to 202012 [sonic-linkmgrd][202012] submodule update a8ddff5 Jing Zhang Fri Feb 25 11:38:28 2022 Incrementing tolerance on mux state inconsistency (#27) a3f78a3 Jing Zhang Thu Feb 17 17:23:56 2022 Update log level for mux probing and mux state chance (#23) 05156fb Jing Zhang Thu Feb 17 17:21:01 2022 Handle xcvrd crashing scenarios (#22) 74529ef Longxiang Lyu Mon Feb 14 13:26:07 2022 [make] Enable make extra includes (#24) sign-off: Jing Zhang zhangjing@microsoft.com

Signed-off-by: Vivek Reddy Karri <vkarri@nvidia.com> Co-authored-by: Vivek Reddy Karri <vkarri@nvidia.com>

…sonic-net#21269) #### Why I did it src/dhcpmon ``` * e003522 - (HEAD -> master, origin/master, origin/HEAD) [Build] Update to buijld bookworm debian package (#24) (21 hours ago) [Yaqiang Zhu] ``` #### How I did it #### How to verify it #### Description for the changelog

… automatically (sonic-net#635) #### Why I did it src/sonic-platform-common ``` * d9de488 - (HEAD -> 202412, origin/202412) [code sync] Merge code from sonic-net/sonic-platform-common:202411 to 202412 (#28) (7 hours ago) [mssonicbld] * 30112ca - [code sync] Merge code from sonic-net/sonic-platform-common:202411 to 202412 (#26) (31 hours ago) [mssonicbld] * a36263b - [code sync] Merge code from sonic-net/sonic-platform-common:202411 to 202412 (#24) (2 days ago) [mssonicbld] ``` #### How I did it #### How to verify it #### Description for the changelog

…tically (sonic-net#723) #### Why I did it src/sonic-sairedis ``` * 5ee5610 - (HEAD -> 202412, origin/HEAD, origin/202412) [code sync] Merge code from sonic-net/sonic-sairedis:202411 to 202412 (#24) (21 hours ago) [mssonicbld] ``` #### How I did it #### How to verify it #### Description for the changelog

…tomatically (sonic-net#754) #### Why I did it src/sonic-linux-kernel ``` * 3f0c0de - (HEAD -> 202412, origin/HEAD, origin/202412) [code sync] Merge code from sonic-net/sonic-linux-kernel:202411 to 202412 (#24) (20 hours ago) [mssonicbld] ``` #### How I did it #### How to verify it #### Description for the changelog

…omatically (sonic-net#770) #### Why I did it src/sonic-swss-common ``` * 9a7a61a - (HEAD -> 202412, origin/HEAD, origin/202412) [FC] remove FLEX_COUNTER_DELAY_STATUS_FIELD (sonic-net#982) (#24) (21 hours ago) [mssonicbld] ``` #### How I did it #### How to verify it #### Description for the changelog

…t back to back Paladin ports up with Arista-7060X6-16PE-384C-O128S2 (sonic-net#22719)  #### Why I did it Currently when we loaded HWSKU `Arista-7060X6-16PE-384C-O128S2` on two moby devices and connect their Paladin ports back to back, we can't get link up. It may help if we can get these links up and run the tests. ##### Work item tracking - Microsoft ADO **(number only)**: #### How I did it Created a new `FANOUT` HWSKU containing special lanemap and polarity configs so that we can load `Arista-7060X6-16PE-384C-O128S2` on one Moby and `Arista-7060X6-16PE-384C-O128S2-FANOUT` and get Paladin ports up when connecting them back to back with the following setup: ``` Moby1 Moby2 HWSKU: Arista-7060X6-16PE-384C-O128S2 HWSKU: Arista-7060X6-16PE-384C-O128S2-FANOUT #17 <-> #18 #19 <-> #20 #21 <-> #22 #23 <-> #24 #18 <-> #17 #20 <-> #19 #22 <-> #21 #24 <-> #23 ``` #### How to verify it Verified that all the Paladin ports can link up with the above setup.  #### Which release branch to backport (provide reason below if selected)  - [ ] 201811 - [ ] 201911 - [ ] 202006 - [ ] 202012 - [ ] 202106 - [ ] 202111 - [ ] 202205 - [ ] 202211 - [ ] 202305 - [x] msft-202412 #### Tested branch (Please provide the tested image version)  - [ ]  - [ ]  - [x] msft-202412 #### Description for the changelog  Created `Arista-7060X6-16PE-384C-O128S2-FANOUT` based on `Arista-7060X6-16PE-384C-O128S2` and only update lanemap and polarity settings in bcm config.  #### Link to config_db schema for YANG module changes  #### A picture of a cute animal (not mandatory but encouraged)

…ly (sonic-net#24550) #### Why I did it src/sonic-bmp ``` * db23384 - (HEAD -> master, origin/master, origin/HEAD) Merge pull request #24 from yijingyan2/migrate-agent-pool-master (3 days ago) [StormLiangMS] * b4e1bc5 - Automated agent pool migration (7 days ago) [yijingyan2] ``` #### How I did it #### How to verify it #### Description for the changelog

vivekrnv added 5 commits July 20, 2022 06:48

SKU first draft added

7872b21

Signed-off-by: Vivek Reddy Karri <vkarri@nvidia.com>

Removev 200G,1500m tuple and added ports2cable dict in t2 defaults

193aa60

Signed-off-by: Vivek Reddy Karri <vkarri@nvidia.com>

Updated buffer_defaults and rounded off to nearest cell size multiple

dcbfc26

Signed-off-by: Vivek Reddy Karri <vkarri@nvidia.com>

Removed fec settings from the port_config.ini

2026d73

Signed-off-by: Vivek Reddy Karri <vkarri@nvidia.com>

Updated the copyright

fa67b5d

Signed-off-by: Vivek Reddy Karri <vkarri@nvidia.com>

stephenxs reviewed Jul 21, 2022

View reviewed changes

Cell roundoff removed and updated mistakes in t0,t1 buffer defaults

ad1b4b3

Signed-off-by: Vivek Reddy Karri <vkarri@nvidia.com>

vivekrnv added 2 commits July 22, 2022 10:22

adjust sizes for mirroring

f27c1ee

Signed-off-by: Vivek Reddy Karri <vkarri@nvidia.com>

Merge branch 'master' of https://github.com/Azure/sonic-buildimage in…

0731d35

…to sku_4700_c128

dgsudharsan requested changes Jul 23, 2022

View reviewed changes

vivekrnv added 2 commits July 25, 2022 17:31

Updated port related config

63acd3e

Signed-off-by: Vivek Reddy Karri <vkarri@nvidia.com>

Minor Indent change

039c82a

Signed-off-by: Vivek Reddy Karri <vkarri@nvidia.com>

dgsudharsan approved these changes Jul 26, 2022

View reviewed changes

assafho approved these changes Jul 26, 2022

View reviewed changes

stephenxs approved these changes Jul 28, 2022

View reviewed changes

vivekrnv closed this Jul 28, 2022

vivekrnv added a commit that referenced this pull request Mar 18, 2024

Fixed OpenIPMI service failure in the runtime (#24)

18ee2c6

Signed-off-by: Vivek Reddy Karri <vkarri@nvidia.com> Co-authored-by: Vivek Reddy Karri <vkarri@nvidia.com>

Conversation

vivekrnv commented Jul 20, 2022 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Why I did it

Requirements:

Additional Details:

How I did it

How to verify it

Which release branch to backport (provide reason below if selected)

Description for the changelog

Link to config_db schema for YANG module changes

A picture of a cute animal (not mandatory but encouraged)

Uh oh!

stephenxs commented Jul 21, 2022 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

vivekrnv commented Jul 21, 2022

Uh oh!

stephenxs Jul 21, 2022 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

stephenxs commented Jul 21, 2022

Uh oh!

vivekrnv commented Jul 21, 2022

Uh oh!

assafho commented Jul 21, 2022

Uh oh!

stephenxs commented Jul 21, 2022

Uh oh!

vivekrnv commented Jul 21, 2022

Uh oh!

stephenxs commented Jul 21, 2022

Uh oh!

vivekrnv commented Jul 23, 2022

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

vivekrnv commented Jul 20, 2022 •

edited

Loading

stephenxs commented Jul 21, 2022 •

edited

Loading

stephenxs Jul 21, 2022 •

edited

Loading