Adjusting test_stress_arp is it can handle cases of ipv6-only topologies . by gshemesh2 · Pull Request #20932 · sonic-net/sonic-mgmt

gshemesh2 · 2025-10-06T08:10:52Z

Description of PR

Summary:
Fixes # (issue)

Type of change

Back port request

Approach

What is the motivation for this PR?

Adjusting test_stress_arp is it can handle cases of ipv6-only topologies .

How did you do it?

In IPv6-only topologies, use 'ip -6 neigh flush all' insted of 'ip -stats neigh flush all' to make it more focus on IPv6 only, and reduce the chance of unsuccessful execution of a command.

Add skip for test_ipv4_arp in case of ipv6 only topology.

How did you verify/test it?

Ran it on Nvidia platforms in IPv6 and IPv4 topologies.

Any platform specific information?

No

Supported testbed topology if it's a new test case?

IPv6 and IPv4 topologies.

Documentation

example

1. adding reeboot before running the test to clean up the db for reset the faults counters 2. adding command that bringup all the ports od the device 3. addding condition for the picked ports, that they are up and connected to optical module.

remove Nvidia platform checks

…ies . (sonic-net#20932) In IPv6-only topologies, use 'ip -6 neigh flush all' insted of 'ip -stats neigh flush all' to make it more focus on IPv6 only, and reduce the chance of unsuccessful execution of a command. Add skip for test_ipv4_arp in case of ipv6 only topology. Signed-off-by: YiFan Wang <yifan@nexthop.ai>

1) Use "enum_rand_one_per_hwsku_frontend_hostname" for "test_ipv6_nd_incomplete" to select the duthost (as being used by other config fixtures like "config_facts", "ip_and_intf_info" etc). [Test was updated recently in PR20780 to use "enum_rand_one_per_hwsku_frontend_hostname" fixture but it seems mistakenly reverted by sonic-net#20932] 2) Most of testcases are sending protocol packets (arp, ipv6_echo requests etc) from server towards ToR device. On dualtor, these protocol packets will be processed by both the ToR duts and thus NEIGH_TABLE in APPL_DB will be populated in both the duts (though hardware/ASIC_DB programming will happen only on active ToR). Fix: It's always better to clear arp and fdb tables on both the ToRs before and after the testcase by updating "arp_cache_fdb_cleanup" fixture to iterate over "duthosts" for cleanup. Signed-off-by: Vinod <vkjammala@arista.com> Signed-off-by: YiFan Wang <yifan@nexthop.ai>

1) Use "enum_rand_one_per_hwsku_frontend_hostname" for "test_ipv6_nd_incomplete" to select the duthost (as being used by other config fixtures like "config_facts", "ip_and_intf_info" etc). [Test was updated recently in PR20780 to use "enum_rand_one_per_hwsku_frontend_hostname" fixture but it seems mistakenly reverted by sonic-net#20932] 2) Most of testcases are sending protocol packets (arp, ipv6_echo requests etc) from server towards ToR device. On dualtor, these protocol packets will be processed by both the ToR duts and thus NEIGH_TABLE in APPL_DB will be populated in both the duts (though hardware/ASIC_DB programming will happen only on active ToR). Fix: It's always better to clear arp and fdb tables on both the ToRs before and after the testcase by updating "arp_cache_fdb_cleanup" fixture to iterate over "duthosts" for cleanup. Signed-off-by: Vinod <vkjammala@arista.com>

1) Use "enum_rand_one_per_hwsku_frontend_hostname" for "test_ipv6_nd_incomplete" to select the duthost (as being used by other config fixtures like "config_facts", "ip_and_intf_info" etc). [Test was updated recently in PR20780 to use "enum_rand_one_per_hwsku_frontend_hostname" fixture but it seems mistakenly reverted by sonic-net#20932] 2) Most of testcases are sending protocol packets (arp, ipv6_echo requests etc) from server towards ToR device. On dualtor, these protocol packets will be processed by both the ToR duts and thus NEIGH_TABLE in APPL_DB will be populated in both the duts (though hardware/ASIC_DB programming will happen only on active ToR). Fix: It's always better to clear arp and fdb tables on both the ToRs before and after the testcase by updating "arp_cache_fdb_cleanup" fixture to iterate over "duthosts" for cleanup. Signed-off-by: Vinod <vkjammala@arista.com> Signed-off-by: Priyansh Tratiya <ptratiya@microsoft.com>

…cases of ipv6-only topologies . #20932 (#923) Manual cherry-pick of PR sonic-net/sonic-mgmt#20932 Adjusting test_stress_arp is it can handle cases of ipv6-only topologies .

…ies . (sonic-net#20932) In IPv6-only topologies, use 'ip -6 neigh flush all' insted of 'ip -stats neigh flush all' to make it more focus on IPv6 only, and reduce the chance of unsuccessful execution of a command. Add skip for test_ipv4_arp in case of ipv6 only topology.

mssonicbld · 2026-01-23T21:14:18Z

Cherry-pick PR to 202511: #22112

…ies . (sonic-net#20932) In IPv6-only topologies, use 'ip -6 neigh flush all' insted of 'ip -stats neigh flush all' to make it more focus on IPv6 only, and reduce the chance of unsuccessful execution of a command. Add skip for test_ipv4_arp in case of ipv6 only topology. Signed-off-by: Guy Shemesh <gshemesh@nvidia.com>

…ies . (sonic-net#20932) In IPv6-only topologies, use 'ip -6 neigh flush all' insted of 'ip -stats neigh flush all' to make it more focus on IPv6 only, and reduce the chance of unsuccessful execution of a command. Add skip for test_ipv4_arp in case of ipv6 only topology. Signed-off-by: Lakshmi Yarramaneni <lakshmi@nexthop.ai>

1) Use "enum_rand_one_per_hwsku_frontend_hostname" for "test_ipv6_nd_incomplete" to select the duthost (as being used by other config fixtures like "config_facts", "ip_and_intf_info" etc). [Test was updated recently in PR20780 to use "enum_rand_one_per_hwsku_frontend_hostname" fixture but it seems mistakenly reverted by #20932] 2) Most of testcases are sending protocol packets (arp, ipv6_echo requests etc) from server towards ToR device. On dualtor, these protocol packets will be processed by both the ToR duts and thus NEIGH_TABLE in APPL_DB will be populated in both the duts (though hardware/ASIC_DB programming will happen only on active ToR). Fix: It's always better to clear arp and fdb tables on both the ToRs before and after the testcase by updating "arp_cache_fdb_cleanup" fixture to iterate over "duthosts" for cleanup. Signed-off-by: Vinod <vkjammala@arista.com> Co-authored-by: vkjammala-arista <152394203+vkjammala-arista@users.noreply.github.com>

…ies . (sonic-net#20932) In IPv6-only topologies, use 'ip -6 neigh flush all' insted of 'ip -stats neigh flush all' to make it more focus on IPv6 only, and reduce the chance of unsuccessful execution of a command. Add skip for test_ipv4_arp in case of ipv6 only topology.

…ies . (sonic-net#20932) In IPv6-only topologies, use 'ip -6 neigh flush all' insted of 'ip -stats neigh flush all' to make it more focus on IPv6 only, and reduce the chance of unsuccessful execution of a command. Add skip for test_ipv4_arp in case of ipv6 only topology. Signed-off-by: Yael Tzur <ytzur@nvidia.com>

1) Use "enum_rand_one_per_hwsku_frontend_hostname" for "test_ipv6_nd_incomplete" to select the duthost (as being used by other config fixtures like "config_facts", "ip_and_intf_info" etc). [Test was updated recently in PR20780 to use "enum_rand_one_per_hwsku_frontend_hostname" fixture but it seems mistakenly reverted by sonic-net#20932] 2) Most of testcases are sending protocol packets (arp, ipv6_echo requests etc) from server towards ToR device. On dualtor, these protocol packets will be processed by both the ToR duts and thus NEIGH_TABLE in APPL_DB will be populated in both the duts (though hardware/ASIC_DB programming will happen only on active ToR). Fix: It's always better to clear arp and fdb tables on both the ToRs before and after the testcase by updating "arp_cache_fdb_cleanup" fixture to iterate over "duthosts" for cleanup. Signed-off-by: Vinod <vkjammala@arista.com> Signed-off-by: Yael Tzur <ytzur@nvidia.com>

…ies . (sonic-net#20932) In IPv6-only topologies, use 'ip -6 neigh flush all' insted of 'ip -stats neigh flush all' to make it more focus on IPv6 only, and reduce the chance of unsuccessful execution of a command. Add skip for test_ipv4_arp in case of ipv6 only topology.

1) Use "enum_rand_one_per_hwsku_frontend_hostname" for "test_ipv6_nd_incomplete" to select the duthost (as being used by other config fixtures like "config_facts", "ip_and_intf_info" etc). [Test was updated recently in PR20780 to use "enum_rand_one_per_hwsku_frontend_hostname" fixture but it seems mistakenly reverted by sonic-net#20932] 2) Most of testcases are sending protocol packets (arp, ipv6_echo requests etc) from server towards ToR device. On dualtor, these protocol packets will be processed by both the ToR duts and thus NEIGH_TABLE in APPL_DB will be populated in both the duts (though hardware/ASIC_DB programming will happen only on active ToR). Fix: It's always better to clear arp and fdb tables on both the ToRs before and after the testcase by updating "arp_cache_fdb_cleanup" fixture to iterate over "duthosts" for cleanup. Signed-off-by: Vinod <vkjammala@arista.com>

…) (sonic-net#22066) 1) Use "enum_rand_one_per_hwsku_frontend_hostname" for "test_ipv6_nd_incomplete" to select the duthost (as being used by other config fixtures like "config_facts", "ip_and_intf_info" etc). [Test was updated recently in PR20780 to use "enum_rand_one_per_hwsku_frontend_hostname" fixture but it seems mistakenly reverted by sonic-net#20932] 2) Most of testcases are sending protocol packets (arp, ipv6_echo requests etc) from server towards ToR device. On dualtor, these protocol packets will be processed by both the ToR duts and thus NEIGH_TABLE in APPL_DB will be populated in both the duts (though hardware/ASIC_DB programming will happen only on active ToR). Fix: It's always better to clear arp and fdb tables on both the ToRs before and after the testcase by updating "arp_cache_fdb_cleanup" fixture to iterate over "duthosts" for cleanup. Signed-off-by: Vinod <vkjammala@arista.com> Co-authored-by: vkjammala-arista <152394203+vkjammala-arista@users.noreply.github.com> Signed-off-by: Lakshmi Yarramaneni <lakshmi@nexthop.ai>

1) Use "enum_rand_one_per_hwsku_frontend_hostname" for "test_ipv6_nd_incomplete" to select the duthost (as being used by other config fixtures like "config_facts", "ip_and_intf_info" etc). [Test was updated recently in PR20780 to use "enum_rand_one_per_hwsku_frontend_hostname" fixture but it seems mistakenly reverted by sonic-net#20932] 2) Most of testcases are sending protocol packets (arp, ipv6_echo requests etc) from server towards ToR device. On dualtor, these protocol packets will be processed by both the ToR duts and thus NEIGH_TABLE in APPL_DB will be populated in both the duts (though hardware/ASIC_DB programming will happen only on active ToR). Fix: It's always better to clear arp and fdb tables on both the ToRs before and after the testcase by updating "arp_cache_fdb_cleanup" fixture to iterate over "duthosts" for cleanup. Signed-off-by: Vinod <vkjammala@arista.com> Signed-off-by: nnelluri-cisco <nnelluri@cisco.com>

…ies . (sonic-net#20932) In IPv6-only topologies, use 'ip -6 neigh flush all' insted of 'ip -stats neigh flush all' to make it more focus on IPv6 only, and reduce the chance of unsuccessful execution of a command. Add skip for test_ipv4_arp in case of ipv6 only topology. Signed-off-by: Raghavendran Ramanathan <rraghav@cisco.com>

1) Use "enum_rand_one_per_hwsku_frontend_hostname" for "test_ipv6_nd_incomplete" to select the duthost (as being used by other config fixtures like "config_facts", "ip_and_intf_info" etc). [Test was updated recently in PR20780 to use "enum_rand_one_per_hwsku_frontend_hostname" fixture but it seems mistakenly reverted by sonic-net#20932] 2) Most of testcases are sending protocol packets (arp, ipv6_echo requests etc) from server towards ToR device. On dualtor, these protocol packets will be processed by both the ToR duts and thus NEIGH_TABLE in APPL_DB will be populated in both the duts (though hardware/ASIC_DB programming will happen only on active ToR). Fix: It's always better to clear arp and fdb tables on both the ToRs before and after the testcase by updating "arp_cache_fdb_cleanup" fixture to iterate over "duthosts" for cleanup. Signed-off-by: Vinod <vkjammala@arista.com> Signed-off-by: Raghavendran Ramanathan <rraghav@cisco.com>

…ies . (sonic-net#20932) In IPv6-only topologies, use 'ip -6 neigh flush all' insted of 'ip -stats neigh flush all' to make it more focus on IPv6 only, and reduce the chance of unsuccessful execution of a command. Add skip for test_ipv4_arp in case of ipv6 only topology. Signed-off-by: Raghavendran Ramanathan <rraghav@cisco.com>

1) Use "enum_rand_one_per_hwsku_frontend_hostname" for "test_ipv6_nd_incomplete" to select the duthost (as being used by other config fixtures like "config_facts", "ip_and_intf_info" etc). [Test was updated recently in PR20780 to use "enum_rand_one_per_hwsku_frontend_hostname" fixture but it seems mistakenly reverted by sonic-net#20932] 2) Most of testcases are sending protocol packets (arp, ipv6_echo requests etc) from server towards ToR device. On dualtor, these protocol packets will be processed by both the ToR duts and thus NEIGH_TABLE in APPL_DB will be populated in both the duts (though hardware/ASIC_DB programming will happen only on active ToR). Fix: It's always better to clear arp and fdb tables on both the ToRs before and after the testcase by updating "arp_cache_fdb_cleanup" fixture to iterate over "duthosts" for cleanup. Signed-off-by: Vinod <vkjammala@arista.com> Signed-off-by: Raghavendran Ramanathan <rraghav@cisco.com>

1) Use "enum_rand_one_per_hwsku_frontend_hostname" for "test_ipv6_nd_incomplete" to select the duthost (as being used by other config fixtures like "config_facts", "ip_and_intf_info" etc). [Test was updated recently in PR20780 to use "enum_rand_one_per_hwsku_frontend_hostname" fixture but it seems mistakenly reverted by sonic-net#20932] 2) Most of testcases are sending protocol packets (arp, ipv6_echo requests etc) from server towards ToR device. On dualtor, these protocol packets will be processed by both the ToR duts and thus NEIGH_TABLE in APPL_DB will be populated in both the duts (though hardware/ASIC_DB programming will happen only on active ToR). Fix: It's always better to clear arp and fdb tables on both the ToRs before and after the testcase by updating "arp_cache_fdb_cleanup" fixture to iterate over "duthosts" for cleanup. Signed-off-by: Vinod <vkjammala@arista.com> Signed-off-by: Zhuohui Tan <zhuohui.tan@amd.com>

gshemesh2 added 30 commits May 28, 2025 16:27

example for PR

dd152cb

example

Merge branch 'sonic-net:master' into master

9a35c5f

Adding WA to for Port Error Test

8df4b25

1. adding reeboot before running the test to clean up the db for reset the faults counters 2. adding command that bringup all the ports od the device 3. addding condition for the picked ports, that they are up and connected to optical module.

Merge branch 'sonic-net:master' into master

8045917

Update test_port_error.py

72572a2

Update test_port_error.py

dd22460

Update test_port_error.py

e58fe89

Update test_port_error.py

c263165

Update test_port_error.py

50fa7e4

Update test_port_error.py

e6e28ec

Update test_port_error.py

b40aa2c

Update test_port_error.py

51b1052

Update test_port_error.py

81d7e2a

Update code-review-guidlines.md

a5fa257

Update test_port_error.py

7bf6710

Update test_port_error.py

5d4a01f

Update test_port_error.py

8cfa578

Update test_port_error.py

1a67218

Update test_port_error.py

4e115ca

Update test_port_error.py

32663a4

Update test_port_error.py

5710fca

Update test_port_error.py

4916a68

Update test_port_error.py

9c21be4

Update test_port_error.py

8a7f377

Update test_port_error.py

2a9fd16

Update test_port_error.py

9cf53ff

Update test_port_error.py

75d5d4a

remove Nvidia platform checks

Update test_port_error.py

45087e2

Merge branch 'sonic-net:master' into master

ca6f72b

Update transceiver_utils.py

29079a7

mssonicbld mentioned this pull request Jan 20, 2026

[action] [PR:21616] [sonic-mgmt] Fix flakiness of "arp/test_stress_arp.py" #22066

Merged

12 tasks

r12f added Included in msft-202412 Branch Request for 202511 branch Request to backport a change to 202511 branch Approved for 202511 branch Created PR to 202511 branch and removed Created PR to msft-202412 Branch Cherry Pick Conflict_msft-202412 labels Jan 23, 2026

mssonicbld mentioned this pull request Jan 23, 2026

[action] [PR:20932] Adjusting test_stress_arp is it can handle cases of ipv6-only topologies . #22112

Open

11 tasks

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Comments

Adjusting test_stress_arp is it can handle cases of ipv6-only topologies .#20932

Adjusting test_stress_arp is it can handle cases of ipv6-only topologies .#20932
roy-sror merged 95 commits intosonic-net:masterfrom
gshemesh2:stress_arp_ipv6

gshemesh2 commented Oct 6, 2025

Uh oh!

mssonicbld commented Jan 23, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

Comments

Conversation

gshemesh2 commented Oct 6, 2025

Description of PR

Type of change

Back port request

Approach

What is the motivation for this PR?

How did you do it?

How did you verify/test it?

Any platform specific information?

Supported testbed topology if it's a new test case?

Documentation

Uh oh!

mssonicbld commented Jan 23, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants