Fix qos test issues in watermark and pfc on Mellanox platform#2174
Merged
neethajohn merged 2 commits intosonic-net:masterfrom Sep 5, 2020
stephenxs:fix-qos-watermark-pfc
Merged
Fix qos test issues in watermark and pfc on Mellanox platform#2174neethajohn merged 2 commits intosonic-net:masterfrom stephenxs:fix-qos-watermark-pfc
neethajohn merged 2 commits intosonic-net:masterfrom
stephenxs:fix-qos-watermark-pfc
Conversation
In the headroom watermark test, it is expected that the watermark is multiple of cell size. However, there is some deviation caused by the below reasons: - The headroom size in SONiC is in bytes and isn't the multiple of cell size. - The SAI calculate the headroom watermark as "sdk max occupancy" - "sonic configuration", if the value configured in SONiC doesn't aligned with cell size boundary, the watermark isn't cell size aligned. Especially, when the packets start to trek to headroom (xoff zone), the headroom size isn't 0 (but it should be), which fails the headroom watermark test. To tolerance the deviation, the number of packets to occupy all the shared buffer is a bit less than it should be. Signed-off-by: Stephen Sun <[email protected]>
1. the deviation in pkts_num_trig_pfc 2. the deviation in xon Signed-off-by: Stephen Sun <[email protected]>
stephenxs
changed the title
neethajohn
approved these changes
Sep 2, 2020
Contributor
|
retest this please |
kazinator-arista
pushed a commit
to kazinator-arista/sonic-mgmt
that referenced
this pull request
Mar 4, 2026
13ec600 [generic-config-updater] Add NTP validator (sonic-net#2212) 4fc09b1 [GCU] Handling non-compliant leaf-list with string values (sonic-net#2174) ac89489 Modify override testcase to cover PORT admin_status (sonic-net#2165) d7953d2 [GCU] Validate peer_group_range ip_range are correct (sonic-net#2145)
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Description of PR
Summary:
Fix two QoS test issues on Mellanox platform.
Signed-off-by: Stephen Sun [email protected]
Type of change
Approach
What is the motivation for this PR?
Fix two QoS test issues on Mellanox platform.
How did you do it?
In the headroom watermark test, it is expected that the watermark is multiple of cell size.
However, there is some deviation caused by the below reasons:
if the value configured in SONiC doesn't aligned with cell size boundary, the watermark
isn't cell size aligned.
Especially, when the packets start to trek to headroom (xoff zone), the headroom size isn't 0
(but it should be), which fails the headroom watermark test.
To tolerance the deviation, the number of packets to occupy all the shared buffer
is a bit less than it should be.
In the PFCXon test on 400G port, need to increase the hysteresis and margin size.
How did you verify/test it?
Any platform specific information?
Supported testbed topology if it's a new test case?
Documentation