temp PR multi-asic-ready-signal --> multi-asic#71
Closed
noaOrMlnx wants to merge 233 commits intomulti-asicfrom
Closed
temp PR multi-asic-ready-signal --> multi-asic#71noaOrMlnx wants to merge 233 commits intomulti-asicfrom
noaOrMlnx wants to merge 233 commits intomulti-asicfrom
Conversation
- Why I did it This dhclient call was added so that the 202505 image which does not have cid set to mac and 202506 which has explicit set for CID to mac has compatibility (DHCP server will only provide lease to same CIDs, but this WA which was added uses dhclient instead of systemd-networkd - and has no CID in the request) The fix itself is present so that dhclient request does not take too long, since we have direct connectivity between DPU and switch, the request should finish within 5 seconds, unblocking all other systemd services and other daemons - How I did it Add timeout of 5 seconds - How to verify it Verified after upgrade from 202505 to 202506 image to confirm that there are no issues: Signed-off-by: gpunathilell <gpunathilell@nvidia.com>
…t#24634) - Why I did it Fix platform name in eeprom.py to support getting system eeprom on 5640 simx platform - How I did it Use correct platform name "x86_64-nvidia_sn5640_simx-r0" - How to verify it Manual test
…onic-net#24125) - Why I did it Following this HLD for liquid cooling sonic-net/SONiC#2032, I created this pr to add Mellanox platform specific APIs. This pr has dependency on sonic-net/sonic-platform-common#603 - How I did it Add support for liquid cool part of the platform API in case thermal How to verify it Add unit test to cover the new added code once liquid cool is enabled Signed-off-by: Yuanzhe Liu <yualiu@nvidia.com>
…4626) - Why I did it An inconsistency was found between sai_5600.xml in the SN5600 Simx platform (x86_64-nvidia_sn5600_simx-r0) and the non-Simx variant (x86_64-nvidia_sn5600-r0). - How I did it To avoid maintaining duplicate files, I replaced the duplicate file with a symlink pointing to the non-SimX variant. This ensures both platforms use the same correct source of truth.
…acement (sonic-net#24688) - Why I did it CMIS module cached module serial in memory, it causes an issue that sfp.get_temperature_info cannot detect SFP replacement. The PR is a fix for the issue. - How I did it Implement a private sfp._get_serial which reads module serial from EEPROM instead of memory. - How to verify it Manual test passed Unit test passed Signed-off-by: Junchao-Mellanox <junchao@nvidia.com>
…t#24759) Signed-off-by: Ashutosh Agrawal <ashu@cisco.com>
How I did it Update YANG and add related unit tests. How to verify it Unit tests are all passed. Signed-off-by: Jing Kan <jika@microsoft.com>
Why I did it When enough neighbors exit for a switch (>211), LLDP will stop reporting the last few neighbors after swss restarts. This happens even though the tcp packets from those neighbors are still arriving over the ports. This is because the netlink buffer is filled up when the restart occurs and the netlink delete messages are not received by lldp, this leads to the service not including the new neighbors over those ports, which will persist until lldp is restarted: How I did it By increasing both the minimum and maximum netlink receive buffer in LLDP I was able to fix this issue How to verify it Run: show lldp table systemctl restart swss Check without the change to see if the lldp buffer cannot handle all of the neighbors, then install a version of Sonic with the change to check that the larger buffer fixes this issue. Signed-off-by: Connor Roos <croos@nvidia.com>
) gateway is optional field. In smartswitch use case, dhcp server runs in the switch and is used to offer addresses to DPU midplane. But we don't want to install default route over the midplane interface so gateway option is removed from DHCP server config. See sonic-net#24495 (comment). Signed-off-by: Yue Gao <yuega2@cisco.com>
…PE-B-O128 and update comments (sonic-net#24774) Why I did it This PR is addressing a comment in sonic-net#24578 by correctly set sai_stats_support_mask for Arista-7060X6-64PE-B-O128. Note that bit 7 is not needed for Arista-7060X6-16PE-384C-B-O128S2 as srv6 is not supported for this platform. I just updated the comment for clarity. Signed-off-by: kewei <kewei@arista.com> Co-authored-by: Vineet Mittal <46945843+vmittal-msft@users.noreply.github.com>
Using a symlink to point to the buffer and pool values for LT2 role QoS. This will make sure any fixes for LT2 will be carried over. The same bcm config file with applied QoSO settings is already being used. Signed-off-by: Dakota Crozier <dakotac@arista.com>
…omatically (sonic-net#24792) #### Why I did it src/sonic-swss-common ``` * 03c6777 - (HEAD -> master, origin/master, origin/HEAD) Opportunistically use a matching kernel version for building the modules (sonic-net#1138) (2 days ago) [Saikrishna Arcot] * 41acbf8 - Ensure RedisPipeline dtor not to throw (sonic-net#1115) (5 days ago) [Qi Luo] * e8a9592 - Reapply "Fix risky unwrap(), expect(), and casting (sonic-net#1113)" (sonic-net#1118) (sonic-net#1124) (5 days ago) [Qi Luo] ``` #### How I did it #### How to verify it #### Description for the changelog
…tically (sonic-net#24791) #### Why I did it src/sonic-sairedis ``` * 1bf99980 - (HEAD -> master, origin/master, origin/HEAD) changes for vpp release 202510 (sonic-net#1695) (3 days ago) [aronovic] * 0ecbb5b9 - Revert "[meta] do not fail bulk operations if MODE_IGNORE_ERROR (sonic-net#1613)" (sonic-net#1676) (3 days ago) [Nikola Dancejic] * 3ab2f07d - Graceful shutdown vpp to avoid core dump (sonic-net#1714) (4 days ago) [yue-fred-gao] * a5e0e632 - Revert "Temp workaround for sonic-buildimage issue 23387 (sonic-net#1629)" (sonic-net#1700) (4 days ago) [Ze Gan] * b735b473 - Use sonic build pool for building sairedis (sonic-net#1719) (5 days ago) [Saikrishna Arcot] ``` #### How I did it #### How to verify it #### Description for the changelog
Co-authored-by: Hemanth Kumar Tirupati <tirupatihemanthkumar@gmail.com> Signed-off-by: Saikrishna Arcot <sarcot@microsoft.com>
Update the following to the version present in Trixie: * bash to 5.2.37-2 * kdump-tools to 1.10.7 * openssh to 10.0p1-7 * snmpd to 5.9.4-2 Co-authored-by: Hua Liu <58683130+liuh-80@users.noreply.github.com> Signed-off-by: Saikrishna Arcot <sarcot@microsoft.com>
The version of systemd on Trixie no longer allows service generators to write to directories outside of what has been explicitly passed in. This affects DPU and multi-ASIC use cases. Therefore, rework systemd-sonic-generator to meet these requirements. Also, compile systemd-sonic-generator with C++17. The gtest headers no longer support C++11, so it needs to be bumped up to C++14 at minimum. Also, move logs for systemd-sonic-generator into /dev/kmsg (#34) Co-authored-by: Hemanth Kumar Tirupati <tirupatihemanthkumar@gmail.com> Signed-off-by: Saikrishna Arcot <sarcot@microsoft.com>
Co-authored-by: Hemanth Kumar Tirupati <tirupatihemanthkumar@gmail.com> Co-authored-by: Brad House - NextHop <bhouse@nexthop.ai> Signed-off-by: Saikrishna Arcot <sarcot@microsoft.com>
Signed-off-by: Saikrishna Arcot <sarcot@microsoft.com>
Signed-off-by: Saikrishna Arcot <sarcot@microsoft.com>
Loosen the help text check in `test_cfghelp.py` in `sonic-yang-mgmt`. The exact text might change from one Python version to another, and help text itself is more for use by a human rather than a machine. It's better to check that the expected elements of the help text (something about the options that are expected and the descriptions) are there rather than the exact formatting. Signed-off-by: Saikrishna Arcot <sarcot@microsoft.com>
There is one test that is failing for unclear reasons. Signed-off-by: Saikrishna Arcot <sarcot@microsoft.com>
Signed-off-by: Saikrishna Arcot <sarcot@microsoft.com>
Signed-off-by: Saikrishna Arcot <sarcot@microsoft.com>
Signed-off-by: Saikrishna Arcot <sarcot@microsoft.com>
Co-authored-by: Yan Markman <ymarkman@marvell.com> Co-authored-by: Hemanth Kumar Tirupati <tirupatihemanthkumar@gmail.com> Signed-off-by: Saikrishna Arcot <sarcot@microsoft.com>
With the base image upgrade to Trixie, Bookworm-based containers will need to use Boost 1.83. This is because of an incompatibility between rsyslog_plugin that uses Boost 1.83 on Trixie and the eventd container that uses Boost 1.74 on Bookworm; specifically there is an incompatiblity with serialization of objects between the two versions of Boost. Because of this, for Bookworm, use Boost 1.83 instead of the default to the default Boost 1.74. Signed-off-by: Saikrishna Arcot <sarcot@microsoft.com>
With the trixie upgrade, all of the package versions for the base image will have changed, meaning the version control files will not be useful at all for the base image. Take this opportunity to recreate all of the version files (including the ones for the containers). Signed-off-by: Saikrishna Arcot <sarcot@microsoft.com>
Install and use the pam_system module, where systemd creates a user session manager for each user that logs in. This is now required for limiting login sessions, but brings in some advantages of cgroups limiting each user's resources and some resource isolation from the main sshd service. Signed-off-by: Saikrishna Arcot <sarcot@microsoft.com>
* Fix FIPS build issue on trixie * Update sonic-fips.mk --------- Co-authored-by: sonicbld <sonicbld@microsoft.com>
* Add cli command to display macsec fips module * Pull latest. Signed-off-by: Wu Miao<wumiao@nokia.com>
Signed-off-by: Tejaswini Chadaga <tchadaga@microsoft.com> Co-authored-by: Vineet Mittal <46945843+vmittal-msft@users.noreply.github.com>
…onic-net#25168) - Why I did it To include latest fixes and new functionality - How I did it SDK_VERSION_DPU 25.10-RC5 -> 26.1-RC1 FW_VERSION 47.1080 -> 48.0318 SAI_VERSION SAIBuild0.0.47.0 -> SAIBuild0.0.48.0 - How to verify it Build an image and run tests from "sonic-mgmt". Signed-off-by: Oleksandr Ivantsiv <oivantsiv@nvidia.com>
…25031) - Why I did it The thermal updater's load_tc_config method was unable to correctly parse tc_config files that use regex-based parameter. - How I did it Added _find_matching_key method: Implements regex pattern matching to find parameter keys in dictionary. Updated load_tc_config method: Changed ASIC parameter lookup to use regex pattern r'asic\\d*' Changed Module parameter lookup to use regex pattern r'module\\d+' - How to verify it Test in switch, check if corrected parsed. Signed-off-by: Jianyue Wu <jianyuew@nvidia.com>
This list is designed to record the previously found leak sensor. It can not be clear otherwise we will loose track. - Why I did it resolve the leaking sensor recovery message can't be sent - How I did it remove the leaking sensor list from reset function so we always have a track for the previous found leaking sensors, this attribute life cycle should be same as the hardware checker object. Signed-off-by: Yuanzhe Liu <yualiu@nvidia.com>
sonic-net#25079) This reverts commit ee76ce5. - Why I did it As the original PR was created as a workaround for a driver issue, This commit is reverted, since now it is not requried anymore. This revert will change the behaviour during system reboot for DPUs dpus will start the startup process and then we proceed with switch reboot - How I did it Revert commit - How to verify it Execute reboot command Signed-off-by: gpunathilell <gpunathilell@nvidia.com>
Signed-off-by: Boyang Yu <byu@arista.com>
…apable (sonic-net#25153) Why I did it To enable macsec on UpperSpineRouters when the device is capable at init time How I did it Modify init_cfg.json to enable macsec on UpperSpineRouter when platform is macsec capable How to verify it Macsec feature is enabled on supported device if it is a SpineRouter or UpperSpineRouter Signed-off-by: Tejaswini Chadaga <tchadaga@microsoft.com> Co-authored-by: Vineet Mittal <46945843+vmittal-msft@users.noreply.github.com>
Why I did it Updated Agera2 gearbox SW version to 3.14.0-2 to include certain fixes.
…D automatically (sonic-net#25247) #### Why I did it src/sonic-platform-daemons ``` * cb899f1 - (HEAD -> master, origin/master, origin/HEAD) Stop the `config_manager` child process on exception in `chassisd` (sonic-net#727) (9 hours ago) [arista-nwolfe] * f0d0f27 - Create handle_cmis_inserted_state function for CMIS_STATE_INSERTED (sonic-net#738) (2 days ago) [Bobby McGonigle] * 282525b - xcvrd: Support regex matching against the medium lane speed key (sonic-net#731) (2 days ago) [Brian Gallagher] * 8b81505 - [stormond] Refactor daemon to reuse CONFIG_DB connection (sonic-net#730) (5 days ago) [gechiang] * dabe1f3 - Update the host lane mask for decommissioning based on the module capability (sonic-net#733) (6 days ago) [arpit-nexthop] ``` #### How I did it #### How to verify it #### Description for the changelog
… automatically (sonic-net#25245) #### Why I did it src/sonic-platform-common ``` * 367d000 - (HEAD -> master, origin/master, origin/HEAD) Fix log prefix hijack in c_cmis.py (sonic-net#609) (4 days ago) [nate-nexthop] * a5149fb - [BMC] make bmc-fw-update.py common (sonic-net#617) (7 days ago) [Ben Levi] ``` #### How I did it #### How to verify it #### Description for the changelog
Include the Mellanox firmware manager Python wheel in the KVM image build to enable firmware management capabilities in KVM deployments.
Signed-off-by: Stepan Blyschak <stepanb@nvidia.com>
Signed-off-by: Stepan Blyschak <stepanb@nvidia.com>
Signed-off-by: Yakiv Huryk <yhuryk@nvidia.com>
- sonic-sairedis - sonic-utilities
- sonic-utilities
noaOrMlnx
force-pushed
the
multi-asic-ready-signal-noaor
branch
from
629a781 to
2c95c20
Compare
noaOrMlnx
pushed a commit
that referenced
this pull request
Feb 18, 2026
…ly (sonic-net#25536) #### Why I did it src/sonic-ztp ``` * 170acb0 - (HEAD -> master, origin/master, origin/HEAD) Skip ZTP service during warm boot (#71) (5 hours ago) [Ying Xie] ``` #### How I did it #### How to verify it #### Description for the changelog
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
20 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.
Why I did it
Work item tracking
How I did it
How to verify it
Which release branch to backport (provide reason below if selected)
Tested branch (Please provide the tested image version)
Description for the changelog
Link to config_db schema for YANG module changes
A picture of a cute animal (not mandatory but encouraged)