[snmp][multi-asic]: Fix test_snmp_queue to support multi-asic platform #9115

SuvarnaMeenakshi · 2023-07-25T22:52:56Z

platform

Signed-off-by: Suvarna Meenakshi [email protected]
(cherry picked from commit 38cb1337201d5f2d9f9f1c3fa9f438d907e41be0)

Description of PR

Summary:
Fixes # (issue)

Type of change

Bug fix
Testbed and Framework(new/improvement)
Test case(new/improvement)

Back port request

201911
202012
202205

Approach

What is the motivation for this PR?

test_snmp_queue was modified to get keys "QUEUE|*" from config_db in #6744.
This has to be modified to get keys from all namespace config_db on mulit-asic platform.
Currently, test tries to retrieve from host config_db and skips the test for multi-asic platform.

Test is currently skipped for supervisor node, but should not be skipped for supervisor node of packet chassis.

How did you do it?

Modified to get interfaces with queue configuration from all namespaces.
The test case was not checking the right field in snmp_interfaces SNMP result. Test case was checking for "description" of each interface, from description, interface name was extracted and checked if the interface name is present in QUEUE table.
This check will always be false and pytest fail condition was never hit. The test case was always incorrectly passing.

(Pdb) snmp_facts['snmp_interfaces']['117']['description']
u'ARISTA02T1:Ethernet1'

   for k, v in list(snmp_facts['snmp_interfaces'].items()):
        if "Ethernet" in v['description']:
            intf = v['description'].split(':')
            # 'ARISTA*:Ethernet*'
            if len(intf) == 2:
                if intf[1] in q_interfaces and 'queues' not in v: -- intf[1] in q_interfaces will always be false
                    pytest.fail(
                        "port %s does not have queue counters" % v['name'])  -- will never hit this condition
                intf[1] will always be Ethernet1 which is the interface of the neighbor

Modified test to use 'name' which gives the interface alias instead of interface name present in 'description'

(Pdb) snmp_facts['snmp_interfaces']['117']['name']
u'fortyGigE0/116'

Remove skip of supervisor node.
On voq chassis LC, QUEUE configuration will include hostname and asic namespace as queue configuration is created on system port. Test is modified to get the interface name for voq chassis.

"QUEUE": {
        "<hostname>|asic0|Ethernet0|0": {
            "scheduler": "scheduler.0"
        },

How did you verify/test it?

Verified on single asic VS testbed.

./run_tests.sh -n vms-kvm-t0 -d vlab-01 -c snmp/test_snmp_queue.py  -f vtestbed.yaml -i ../ansible/veos_vtb -e "--skip_sanity --disable_loganalyzer" -u
=== Running tests in groups ===
Running: pytest snmp/test_snmp_queue.py --inventory ../ansible/veos_vtb --host-pattern vlab-01 --testbed vms-kvm-t0 --testbed_file vtestbed.yaml --log-cli-level warning --log-file-level debug --kube_master unset --showlocals --assert plain --show-capture no -rav --allow_recover --ignore=ptftests --ignore=acstests --ignore=saitests --ignore=scripts --ignore=k8s --ignore=sai_qualify --junit-xml=logs/tr.xml --log-file=logs/test.log --skip_sanity --disable_loganalyzer
/usr/local/lib/python2.7/dist-packages/ansible/parsing/vault/__init__.py:44: CryptographyDeprecationWarning: Python 2 is no longer supported by the Python core team. Support for it is now deprecated in cryptography, and will be removed in the next release.
  from cryptography.exceptions import InvalidSignature
===================================================================================================================== test session starts ======================================================================================================================
..                                                                                                                                                                                         

snmp/test_snmp_queue.py::test_snmp_queues[vlab-01] PASSED                                                                                                                                                                                                [100%]

---------------------------------------------------------------------------------------------------- generated xml file: /data/sonic-mgmt/tests/logs/tr.xml ----------------------------------------------------------------------------------------------------
================================================================================================================== 1 passed in 18.02 seconds ===================================================================================================================
INFO:root:Can not get Allure report URL. Please check logs

Verified on packet and voq chassis.

Any platform specific information?

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

Documentation

platform Signed-off-by: Suvarna Meenakshi <[email protected]> (cherry picked from commit 38cb1337201d5f2d9f9f1c3fa9f438d907e41be0)

arlakshm · 2023-08-08T00:04:18Z

tests/snmp/test_snmp_queue.py

-    q_keys = redis_get_keys(duthost, "CONFIG_DB", "QUEUE|*")
+    for asic_id in duthost.get_asic_ids():
+        namespace = duthost.get_namespace_from_asic_id(asic_id)
+        sonic_db_cmd = "sonic-db-cli {}".format("-n " + namespace if namespace else "")


SonicAsic class already function to run sonic-db-cli on a given asic. Use it directly?

sonic-mgmt/tests/common/devices/sonic_asic.py

Line 156 in 21784ae

def run_sonic_db_cli_cmd(self, sonic_db_cmd):

Modified as per comment

Modify test case to check for interface name in snmp_interfaces result. Signed-off-by: Suvarna Meenakshi <[email protected]>

Signed-off-by: Suvarna Meenakshi <[email protected]>

SuvarnaMeenakshi · 2023-08-21T18:39:55Z

/azp run

azure-pipelines · 2023-08-21T18:40:04Z

Azure Pipelines successfully started running 1 pipeline(s).

SuvarnaMeenakshi · 2023-08-21T18:40:24Z

@qiluo-msft can re-review, after the previous approval, made a fix in the log message.

mssonicbld · 2023-08-28T07:54:10Z

@SuvarnaMeenakshi PR conflicts with 202205 branch

sonic-net#9115) What is the motivation for this PR? test_snmp_queue was modified to get keys "QUEUE|*" from config_db in sonic-net#6744. This has to be modified to get keys from all namespace config_db on mulit-asic platform. Currently, test tries to retrieve from host config_db and skips the test for multi-asic platform. Test is currently skipped for supervisor node, but should not be skipped for supervisor node of packet chassis. How did you do it? Modified to get interfaces with queue configuration from all namespaces. The test case was not checking the right field in snmp_interfaces SNMP result. Test case was checking for "description" of each interface, from description, interface name was extracted and checked if the interface name is present in QUEUE table. This check will always be false and pytest fail condition was never hit. The test case was always incorrectly passing. (Pdb) snmp_facts['snmp_interfaces']['117']['description'] u'ARISTA02T1:Ethernet1' for k, v in list(snmp_facts['snmp_interfaces'].items()): if "Ethernet" in v['description']: intf = v['description'].split(':') # 'ARISTA*:Ethernet*' if len(intf) == 2: if intf[1] in q_interfaces and 'queues' not in v: -- intf[1] in q_interfaces will always be false pytest.fail( "port %s does not have queue counters" % v['name']) -- will never hit this condition intf[1] will always be Ethernet1 which is the interface of the neighbor Modified test to use 'name' which gives the interface alias instead of interface name present in 'description' (Pdb) snmp_facts['snmp_interfaces']['117']['name'] u'fortyGigE0/116' Remove skip of supervisor node. On voq chassis LC, QUEUE configuration will include hostname and asic namespace as queue configuration is created on system port. Test is modified to get the interface name for voq chassis. "QUEUE": { "<hostname>|asic0|Ethernet0|0": { "scheduler": "scheduler.0" }, How did you verify/test it? Verified on single asic VS testbed.

mssonicbld · 2023-08-28T07:54:15Z

Cherry-pick PR to 202305: #9663

#9115) What is the motivation for this PR? test_snmp_queue was modified to get keys "QUEUE|*" from config_db in #6744. This has to be modified to get keys from all namespace config_db on mulit-asic platform. Currently, test tries to retrieve from host config_db and skips the test for multi-asic platform. Test is currently skipped for supervisor node, but should not be skipped for supervisor node of packet chassis. How did you do it? Modified to get interfaces with queue configuration from all namespaces. The test case was not checking the right field in snmp_interfaces SNMP result. Test case was checking for "description" of each interface, from description, interface name was extracted and checked if the interface name is present in QUEUE table. This check will always be false and pytest fail condition was never hit. The test case was always incorrectly passing. (Pdb) snmp_facts['snmp_interfaces']['117']['description'] u'ARISTA02T1:Ethernet1' for k, v in list(snmp_facts['snmp_interfaces'].items()): if "Ethernet" in v['description']: intf = v['description'].split(':') # 'ARISTA*:Ethernet*' if len(intf) == 2: if intf[1] in q_interfaces and 'queues' not in v: -- intf[1] in q_interfaces will always be false pytest.fail( "port %s does not have queue counters" % v['name']) -- will never hit this condition intf[1] will always be Ethernet1 which is the interface of the neighbor Modified test to use 'name' which gives the interface alias instead of interface name present in 'description' (Pdb) snmp_facts['snmp_interfaces']['117']['name'] u'fortyGigE0/116' Remove skip of supervisor node. On voq chassis LC, QUEUE configuration will include hostname and asic namespace as queue configuration is created on system port. Test is modified to get the interface name for voq chassis. "QUEUE": { "<hostname>|asic0|Ethernet0|0": { "scheduler": "scheduler.0" }, How did you verify/test it? Verified on single asic VS testbed.

sonic-net#9115) What is the motivation for this PR? test_snmp_queue was modified to get keys "QUEUE|*" from config_db in sonic-net#6744. This has to be modified to get keys from all namespace config_db on mulit-asic platform. Currently, test tries to retrieve from host config_db and skips the test for multi-asic platform. Test is currently skipped for supervisor node, but should not be skipped for supervisor node of packet chassis. How did you do it? Modified to get interfaces with queue configuration from all namespaces. The test case was not checking the right field in snmp_interfaces SNMP result. Test case was checking for "description" of each interface, from description, interface name was extracted and checked if the interface name is present in QUEUE table. This check will always be false and pytest fail condition was never hit. The test case was always incorrectly passing. (Pdb) snmp_facts['snmp_interfaces']['117']['description'] u'ARISTA02T1:Ethernet1' for k, v in list(snmp_facts['snmp_interfaces'].items()): if "Ethernet" in v['description']: intf = v['description'].split(':') # 'ARISTA*:Ethernet*' if len(intf) == 2: if intf[1] in q_interfaces and 'queues' not in v: -- intf[1] in q_interfaces will always be false pytest.fail( "port %s does not have queue counters" % v['name']) -- will never hit this condition intf[1] will always be Ethernet1 which is the interface of the neighbor Modified test to use 'name' which gives the interface alias instead of interface name present in 'description' (Pdb) snmp_facts['snmp_interfaces']['117']['name'] u'fortyGigE0/116' Remove skip of supervisor node. On voq chassis LC, QUEUE configuration will include hostname and asic namespace as queue configuration is created on system port. Test is modified to get the interface name for voq chassis. "QUEUE": { "<hostname>|asic0|Ethernet0|0": { "scheduler": "scheduler.0" }, How did you verify/test it? Verified on single asic VS testbed. (cherry picked from commit 8123ab7)

#9115) (#9973) Cherry-pick of #9115 What is the motivation for this PR? test_snmp_queue was modified to get keys "QUEUE|*" from config_db in #6744. This has to be modified to get keys from all namespace config_db on mulit-asic platform. Currently, test tries to retrieve from host config_db and skips the test for multi-asic platform. Test is currently skipped for supervisor node, but should not be skipped for supervisor node of packet chassis. How did you do it? Modified to get interfaces with queue configuration from all namespaces. The test case was not checking the right field in snmp_interfaces SNMP result. Test case was checking for "description" of each interface, from description, interface name was extracted and checked if the interface name is present in QUEUE table. This check will always be false and pytest fail condition was never hit. The test case was always incorrectly passing. (Pdb) snmp_facts['snmp_interfaces']['117']['description'] u'ARISTA02T1:Ethernet1' for k, v in list(snmp_facts['snmp_interfaces'].items()): if "Ethernet" in v['description']: intf = v['description'].split(':') # 'ARISTA*:Ethernet*' if len(intf) == 2: if intf[1] in q_interfaces and 'queues' not in v: -- intf[1] in q_interfaces will always be false pytest.fail( "port %s does not have queue counters" % v['name']) -- will never hit this condition intf[1] will always be Ethernet1 which is the interface of the neighbor Modified test to use 'name' which gives the interface alias instead of interface name present in 'description' (Pdb) snmp_facts['snmp_interfaces']['117']['name'] u'fortyGigE0/116' Remove skip of supervisor node. On voq chassis LC, QUEUE configuration will include hostname and asic namespace as queue configuration is created on system port. Test is modified to get the interface name for voq chassis. "QUEUE": { "<hostname>|asic0|Ethernet0|0": { "scheduler": "scheduler.0" }, How did you verify/test it? Verified on single asic VS testbed. (cherry picked from commit 8123ab7)

sonic-net#9115) What is the motivation for this PR? test_snmp_queue was modified to get keys "QUEUE|*" from config_db in sonic-net#6744. This has to be modified to get keys from all namespace config_db on mulit-asic platform. Currently, test tries to retrieve from host config_db and skips the test for multi-asic platform. Test is currently skipped for supervisor node, but should not be skipped for supervisor node of packet chassis. How did you do it? Modified to get interfaces with queue configuration from all namespaces. The test case was not checking the right field in snmp_interfaces SNMP result. Test case was checking for "description" of each interface, from description, interface name was extracted and checked if the interface name is present in QUEUE table. This check will always be false and pytest fail condition was never hit. The test case was always incorrectly passing. (Pdb) snmp_facts['snmp_interfaces']['117']['description'] u'ARISTA02T1:Ethernet1' for k, v in list(snmp_facts['snmp_interfaces'].items()): if "Ethernet" in v['description']: intf = v['description'].split(':') # 'ARISTA*:Ethernet*' if len(intf) == 2: if intf[1] in q_interfaces and 'queues' not in v: -- intf[1] in q_interfaces will always be false pytest.fail( "port %s does not have queue counters" % v['name']) -- will never hit this condition intf[1] will always be Ethernet1 which is the interface of the neighbor Modified test to use 'name' which gives the interface alias instead of interface name present in 'description' (Pdb) snmp_facts['snmp_interfaces']['117']['name'] u'fortyGigE0/116' Remove skip of supervisor node. On voq chassis LC, QUEUE configuration will include hostname and asic namespace as queue configuration is created on system port. Test is modified to get the interface name for voq chassis. "QUEUE": { "<hostname>|asic0|Ethernet0|0": { "scheduler": "scheduler.0" }, How did you verify/test it? Verified on single asic VS testbed.

[snmp][multi-asic]: Fix test_snmp_queue to support multi-asic

63d2f5a

platform Signed-off-by: Suvarna Meenakshi <[email protected]> (cherry picked from commit 38cb1337201d5f2d9f9f1c3fa9f438d907e41be0)

SuvarnaMeenakshi requested a review from arlakshm July 25, 2023 22:53

SuvarnaMeenakshi added the Request for 202205 branch label Jul 25, 2023

SuvarnaMeenakshi requested a review from abdosi July 28, 2023 15:34

abdosi previously approved these changes Aug 2, 2023

View reviewed changes

arlakshm reviewed Aug 8, 2023

View reviewed changes

Fix as per review comment.

627c553

Modify test case to check for interface name in snmp_interfaces result. Signed-off-by: Suvarna Meenakshi <[email protected]>

SuvarnaMeenakshi dismissed abdosi’s stale review via 627c553 August 8, 2023 21:06

SuvarnaMeenakshi requested a review from qiluo-msft August 8, 2023 21:18

qiluo-msft previously approved these changes Aug 9, 2023

View reviewed changes

Fix assert comment

99f8257

Signed-off-by: Suvarna Meenakshi <[email protected]>

SuvarnaMeenakshi dismissed qiluo-msft’s stale review via 99f8257 August 9, 2023 00:05

SuvarnaMeenakshi requested a review from arlakshm August 11, 2023 19:29

qiluo-msft approved these changes Aug 21, 2023

View reviewed changes

SuvarnaMeenakshi added the Request for 202305 branch label Aug 22, 2023

SuvarnaMeenakshi merged commit 8123ab7 into sonic-net:master Aug 22, 2023
12 checks passed

wangxin added Approved for 202205 branch Approved for 202305 branch labels Aug 28, 2023

mssonicbld added the Cherry Pick Conflict_202205 label Aug 28, 2023

mssonicbld added the Created PR to 202305 branch label Aug 28, 2023

mssonicbld mentioned this pull request Aug 28, 2023

[action] [PR:9115] [snmp][multi-asic]: Fix test_snmp_queue to support multi-asic platform #9663

Merged

6 tasks

mssonicbld added Included in 202305 branch and removed Created PR to 202305 branch labels Aug 28, 2023

SuvarnaMeenakshi mentioned this pull request Sep 6, 2023

test_snmp_queue fails due to unavailable queue counters #8934

Closed

SuvarnaMeenakshi mentioned this pull request Sep 14, 2023

[snmp][multi-asic][202205]: Fix test_snmp_queue to support multi-asic platform #9973

Merged

6 tasks

wangxin added the Included in 202205 branch label Sep 15, 2023

veronica-arista mentioned this pull request Sep 22, 2023

[sonic-mgmt snmp] Multi-asic support added test_snmp_queue.py does not work for multiple linecards #10099

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[snmp][multi-asic]: Fix test_snmp_queue to support multi-asic platform #9115

[snmp][multi-asic]: Fix test_snmp_queue to support multi-asic platform #9115

SuvarnaMeenakshi commented Jul 25, 2023 •

edited

Loading

arlakshm Aug 8, 2023

SuvarnaMeenakshi Aug 8, 2023

SuvarnaMeenakshi commented Aug 21, 2023

azure-pipelines bot commented Aug 21, 2023

SuvarnaMeenakshi commented Aug 21, 2023

mssonicbld commented Aug 28, 2023

mssonicbld commented Aug 28, 2023

[snmp][multi-asic]: Fix test_snmp_queue to support multi-asic platform #9115

[snmp][multi-asic]: Fix test_snmp_queue to support multi-asic platform #9115

Conversation

SuvarnaMeenakshi commented Jul 25, 2023 • edited Loading

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

arlakshm Aug 8, 2023

Choose a reason for hiding this comment

SuvarnaMeenakshi Aug 8, 2023

Choose a reason for hiding this comment

SuvarnaMeenakshi commented Aug 21, 2023

azure-pipelines bot commented Aug 21, 2023

SuvarnaMeenakshi commented Aug 21, 2023

mssonicbld commented Aug 28, 2023

mssonicbld commented Aug 28, 2023

SuvarnaMeenakshi commented Jul 25, 2023 •

edited

Loading