- Why I did it
Support PSU voltage high/low thresholds and power max threshold
1. Add thresholds support for voltage and power.
2. As thresholds are not supported on all platforms, we need to check the capability first and fetch thresholds only if it is supported.
- How I did it
- How to verify it
Run regression test and manual test.
Signed-off-by: Stephen Sun <stephens@nvidia.com>
#### Why I did it
Mellanox builds were failing intermittently due to the `issue_version` file and MFT package not building correctly in the Azure pipeline environment (both of these packages were patched to build correctly with bullseye running on the host and buster running on the dockers)
#### How I did it
Fixed two problems:
1. BLDENV is not passed to the Makefiles so the references to this were replaced with correct logic
2. `issue_version` was not defined as a target for bullseye and as such was not cached. Altered the build such that it is defined as a target for bullseye (in the case of buster it builds the file, in the case of bullseye it copies from buster)
The previous PR fixing this was reverted as it is no longer necessary for a passing build and was not a long-term fix. https://github.com/Azure/sonic-buildimage/pull/9235
#### How to verify it
Build on AZP and verify success.
The issu-version file for Mellanox is generated from the Mellanox SDK
libraries. The SDK is installed into a Buster docker container, but the
issu-version file goes onto the base OS, which is Bullseye. To work
around this, the issu-version build rules explicitly copies the
issu-version file to target/files/bullseye/ during the Buster build.
Because of our build infra, if caching is enabled and a cache is being
used, then for issu-version, since it is technically built as part of
Buster, then only target/files/buster/issu-version is saved into the
cache, and target/files/bullseye/issu-version isn't cached. If this
cache gets used, then target/files/bullseye/issu-version is missing, and
the final image build fails.
This is to work around the current build issue where Mellanox builds are
failing. This is so that issu-version is always "built", so that copy is
made into the bullseye directory.
Signed-off-by: Saikrishna Arcot <sarcot@microsoft.com>
* Make neccesary changed to mellanox platform code to build on Debian 11
* Revert use of backported kernel to build mft and elect to only build kernel module under bullseye
Allow mellanox platform to build and successfully switch packets in
Debian 11
Upgraded
* Mellanox SDK
* Mellanox Hardware Management
* Mellanox Firmware
* Mellanox Kernel Patches
Adjusted build system to support host system running bullseye and
dockers running buster.
ISSU will likely be broken. As of right now, the issu-version file is
not being generated during build.
Signed-off-by: Saikrishna Arcot <sarcot@microsoft.com>
- Why I did it
In case an app.ext requires a dependency syncd^1.0.0, the RPC version of syncd will not satisfy this constraint, since 1.0.0-rpc < 1.0.0. This is not correct to put 'rpc' as a prerelease identifier. Instead put 'rpc' as build metadata in the version: 1.0.0+rpc which satisfies the constraint ^1.0.0.
- How I did it
Changed the way how to version in RPC and DBG images are constructed.
- How to verify it
Install app.ext with syncd^1.0.0 dependency on a switch with RPC syncd docker.
Signed-off-by: Stepan Blyshchak <stepanb@nvidia.com>
- Why I did it
* To support systems with dynamic port configuration
* Apply lazy initialization to faster the speed of loading platform API
- How I did it
* Add module.py to implement dynamic port configuration (aka line card model)
* Adjust chassis.py, platform.py, thermal.py, sfp.py to support dynamic port configuration
* Optimize existing code
- How to verify it
Platform regression on MSN4700, MSN3800 and MSN2700, 100% pass
Unit test covers all new changes.
- Why I did it
Bug fix:
bad_param request due to missing parser rest command while running mlxlink
- How I did it
Advance to MFT tool version to 4.17.2-12.
- How to verify it
Manually tested on all mellanox platforms.
- Why I did it
Add NVIDIA Copyright header to "mellanox" files
- How I did it
Add NVIDIA Copyright header as a comment for Mellanox files
- How to verify it
Sanity tests and PR checkers.
Why I did it
Currently the mellanox platform API is validating the file extensions of firmware packages to be installed for basic sanity checking. However, ONIE packages do not have an extension and as such if there is a "." in the name it is taken to be an extension and then fails the sanity check.
How I did it
I removed the check which ensures that ONIE images don't have a file extension.
How to verify it
Name the ONIE updater file 2021.onie and attempt to install it via fwutil install fw 2021.onie --yes
Why I did it
The fwutil update all utility expects the auto_update_firmware method in the Platform API to execute the update_firmware() call and not the install_firmware() call.
How I did it
Changed the method in the mellanox platform API component implementation.
How to verify it
Run fwutil update all with a CPLD update on a Mellanox platform and verify that it properly updates the firmware using the MPFA file.
- Why I did it
Advance to new Nvidia SAI release with the following changes:
New features:
- Align with new SDK/FW version 4.5.1006 and above and in parallel to existing used SDK/FW bundle
- Implement timestamp and egress_queue_index hostif packet attributes.
Bugs fixes:
- Fix compilation issues with gcc10
- Fix return code for buffer overflow for query enum values and query statistics capabilities
- Reduce verbosity of print in case packet ingress on invalid port
- Fix mirror Qos settings
- How I did it
Updated SAI version and submodule pointer
- How to verify it
Run regression tests from sonic-mgmt
Signed-off-by: Volodymyr Samotiy <volodymyrs@nvidia.com>
- Why I did it
Change thermal recover threshold from temp_trip_norm to temp_trip_high, so that thermal algorithm would set fan speed to minimum allowed earlier and save power.
- How I did it
Change thermal recover threshold from temp_trip_norm to temp_trip_high
- How to verify it
Manual test
- Why I did it
Advance to Mellanox SAI ver 1.19.2 to pick up dynamic Policy Based Hashing support.
For this version above the static Policy Based Hashing is no longer supported.
For detailed release notes check https://github.com/Mellanox/SAI-Implementation/blob/sonic2111/release%20notes.txt
- How I did it
Updated SAI-Implementation submodule
- How to verify it
1. make configure PLATFORM=mellanox
2. make target/sonic-mellanox.bin
Run full regression as well as new dynamic PBH tests
Signed-off-by: Nazarii Hnydyn <nazariig@nvidia.com>
#### Why I did it
New PSU could install different type of fan, so fan max/min speed should be read per PSU
#### How I did it
The existing implementation read PSU max/min fan speed from a common file, change it to read from per PSU file
#### How to verify it
Manual test
- Why I did it
Update SDK\FW version to 4.4.3326\2008.3326. This version contains:
New Features:
1. Add support for Fast Boot for SN3800
Bug Fixing:
1. In some cases, when the total number of allocations exceeds the resource limit, an error can occur due to incorrect resource release procedure. This issue is most likely to affect the following resources: flow counters, ACL actions, PBS, WJH filter, Tunnels, ECMP containers, MC (L2 &L3)
2. On Spectrum systems, when using Async Router API with IPV6, an error message in the log regarding failing to remove ECMP container may show up. This error is not functional and can be safely ignored.
3. On Spectrum-2 systems and above, when using warm boot, setting max_bridge_num to a value greater than 1968 will cause an error and potential crash.
4. Some Molex cables do not support speed after reboot
- How I did it
- How to verify it
Was verified by running regression tests that includes complete sonic-mgmt tests supported
- Why I did it
New release of MFT has the following changelog / RN
Fixed an issue that resulted in getting MVPD read errors from the mlxfwmanager during fast reboot.
Fixed mlxuptime sometimes generating a time less than previous due the wrong frequency calculation
- How I did it
Update makefile pointer to new version.
- How to verify it
Manually tested on all Mellanox platforms.
Update SDK\FW version to 4.4.3320\2008.3324. This version contains:
New Features:
* Add support for Fast Boot for SN3800
Bug Fixing:
* In some cases, when the total number of allocations exceeds the resource limit, an error can occur due to incorrect resource release procedure. This issue is most likely to affect the following resources: flow counters, ACL actions, PBS, WJH filter, Tunnels, ECMP containers, MC (L2 &L3)
* On Spectrum systems, when using Async Router API with IPV6, an error message in the log regarding failing to remove ECMP container may show up. This error is not functional and can be safely ignored.
* On Spectrum-2 systems and above, when using warm boot, setting max_bridge_num to a value greater than 1968 will cause an error and potential crash.
* Some Molex cables do not support speed after reboot
Signed-off-by: Dror Prital <drorp@nvidia.com>
The start template script that this value is used in will determine what
network namespace to use, and will add --net=host if it needs to run in
the host namespace.
With Docker 20.10, if --net=host is specified twice in docker run, then
it errors out. Therefore, remove the explicit --net=host in the run
options setting and let the start template script specify it.
Signed-off-by: Saikrishna Arcot <sarcot@microsoft.com>
Why I did it
BIOS upgrade on rare cases cannot guarantee bus value remain the same on every BIOS release. Ignoring this field in order for pcied not to fail but still verify device id in a different way. The solution is future proof and will not require changes in code when new BIOS version is available
How I did it
Since bus is not a fixed value (it is determined by the bios version) we are ignoring this field, and instead checking if there is a device that match on all other fields that and in addition has a matching device id.
How to verify it
Verify no errors or failures in pcied on different BIOS version with the same code base.
- Why I did it
Update SAI version to 1.19.1. The following was changed:
1. Update license
2. Do not remove and re-apply the same SDK mirror session on LAG
3. FEC fix to support all speeds
4. Improve PG counters performance
5. Fix number of switch priorities for port mirroring
Signed-off-by: Dror Prital <drorp@nvidia.com>
Avoid initializing sfp/thermal/components/fan/psu/leds on simx and create vpd_info file on hw_management when we use mellanox simulator platform
- Why I did it
this is a fix for issue in mellanox simulator platforms. the syseepromd failed on the pmon docker. also "decode-syseeprom" failed also
- How I did it
before initializing thermal/components/fan/psu/leds --> check if we are running on simx
creating the vpd_info on the hw_management folder.
- How to verify it
check if syseepromd process was loaded properly on the pmon docker.
decode-syseeprom is working well without errors/warnings
- Why I did it
to prevent python exception error when executing warm-reboot command on mellanox simulator platform
- How I did it
return None on the watchdog python script on cases that watchdog file is not exist
- How to verify it
warm-reboot is running well without the python error. error message will appear on log on these cases.
in order to avoid this error message we can simulate the watchdog on mellanox simulator platform
Update FW version to 2008.3218, fixing the following issues:
- 50G/100G links that are operationally down before warm-reboot are not coming up after warm-reboot
- 50G/100G links with admin shut / no shut commands are not coming up after warm-reboot
Signed-off-by: Dror Prital <drorp@nvidia.com>
- Why I did it
* For SAI - Advance to adopt the following fixes:
1. Better handle not implement object type for resource availability
2. Fix ext dump when saidump is triggered from 2nd process (saidump utility) other than main adapter host (syncd in SONiC)
* For SDK\FW:
- Changes and new features:
1. Added support in SN4600C systems for new module Finisar ET7402-CWDM4 (100G CWDM4 QSFP28 1310nm SM 2KM).
2. Added support for new module MMS1W50-HM (2km transceiver FR4) for 200GbE
3. Improved performance of "per-port-buffer" counters
4. Added support for Kernel 5.10
- Bugs fixes:
On rare occasions (0.5%), in SN4600C systems, when using 100GbE NRZ mode and Fastboot flow, the link up time may take up to 10 seconds
Signed-off-by: Dror Prital <drorp@nvidia.com>
#### Why I did it
Currently, SONiC use a single value to represent SFP error, however, multiple SFP errors could exist at the same time. This PR is aimed to support it
#### How I did it
Return bitmap instead of single value when a SFP event occurs
Signed-off-by: Stephen Sun <stephens@nvidia.com>
Split and bulk counter bug fixes:
- Init port auto neg to default on static (SAI XML) port split for 2nd+ port
- Fix port stats SAI_PORT_STAT_WRED_DROPPED_PACKETS, SAI_PORT_STAT_ECN_MARKED_PACKETS, SAI_PORT_STAT_ETHER_TX_OVERSIZE_PKTS
- Hide error message when reading not implemented port stat
Signed-off-by: Shlomi Bitton <shlomibi@nvidia.com>
- Why I did it
The methods get_model, get_serial, and get_revision have been implemented by reading relevant information from VPD and then recording the information into relevant fields.
However, there is no VPD data on platforms with fixed PSUs and relevant fields haven't been initialized, which causes the methods to throw exceptions. which in turn prevents psud from inserting fields into PSU table.
Eventually, this causes show platform psustatus doesn't output correct info.
- How I did it
Initialize those fields as N/A on systems with fixed PSUs.
- How to verify it
Manually test.
Signed-off-by: Stephen Sun <stephens@nvidia.com>
- Why I did it
Remove EEPROM cache file and use DB instead
- How I did it
Read EEPROM data from DB if possible
If data is not ready in DB, read from hardware using a visitor pattern
- How to verify it
Manual test and regression
Why I did it
The Mellanox platform is required to support the fwutil auto-update feature defined here
This is to allow switches, when performing SONiC upgrades to choose whether to perform firmware upgrades that may interrupt the data plane through a cold boot.
How I did it
Two methods were added to the component implementations for mellanox.
In the base Component class we add a default function that chooses to skip the installation of any firmware unless the cold boot option is provided. This is because the Mellanox platform, by default, does not support installing firmware on ONIE, the CPLD, or the BIOS "on-the-fly".
In the ComponentSSD class we add a function that behaves similarly but uses the Mellanox specific SSD firmware upgrade tool to check if the current SSD supports being upgraded on the fly in order to decide whether to skip or perform the installation.
How to verify it
Unit tests are included with this PR. These test will run on build of target sonic-mellanox.bin
You may also perform fwutil auto-update ... commands after Azure/sonic-utilities#1242 is merged in.
- Why I did it
Enhance the Python3 support for platform API. Originally, some platform APIs call SDK API which didn't support Python 3. Now the Python 3 APIs have been supported in SDK 4.4.3XXX, Python3 is completely supported by platform API
- How I did it
Start all platform daemons from python3
1. Remove #/usr/bin/env python at the beginning of each platform API file as the platform API won't be started as daemons but be imported from other daemons.
2. Adjust SDK API calls accordingly
- How to verify it
Manually test and run regression platform test
Signed-off-by: Stephen Sun <stephens@nvidia.com>
- Why I did it
Adjust the Makefile for SDK/python-SDK-API to support both python2 and python3
- How to verify it
Build the image and check whether python2 and python3 are both supported by SDK API.
Signed-off-by: Stephen Sun <stephens@nvidia.com>
Why I did it
* For SAI - Upgrade to Version 1.19.0
- Add support for VxLAN encap TTL uniform model on SPC2/3
- Add ACL entry actions set VRF, set do no learn, add VLAN ID, add VLAN priority
- Add ACL field has VLAN tag
- Bulk counters (improve port statistics performance)
- Create async dump extra as part of debug generate dump
- Create irisc dump on severe health event
- Support 0 port systems (modify get switch mac to work accordingly)
- Set interface vlan up state for ping tool in SONiC
- Support attributes SAI_PORT_ATTR_QOS_SCHEDULER_PROFILE_ID, SAI_PORT_ATTR_QOS_INGRESS_BUFFER_PROFILE_LIST,
SAI_PORT_ATTR_QOS_EGRESS_BUFFER_PROFILE_LIST, SAI_PORT_ATTR_POLICER_ID as part of port create Git stats
* For SDK\FW - Upgrade to Version SDK 4.4.3106, FW 2008_3110
Added Features:
- Increased ACL table
- Enhanced PSAMPLE support
- Added support for Finisar SR4 module in SN3700 systems
- Added support for Python 3.0 in examples.
Fix bugs:
- On LR4 transceivers 00YD278, the firmware incorrectly identified the transceiver
- Reduce memory consumption for virtual LAG
- Fixed PSAMPLE listeners cleanup on SDK drivers unloading.
- On Spectrum-2 and Spectrum-3 systems, slow reaction time to Rx pause packets may lead to buffer overflow on servers.
- BER may be experienced when using 5m DAC cables between SN4700 and SN2700 in 100GbE speed.
- On very rare occasion, when connecting DR4 PAM4 transceiver to 100GbE DR1 NRZ, low BER may be experienced.
- Unexpected packet drops on the port ingress buffer may be experienced when working in 400GbE mode.
Note: When performing ISSU from an older version, this fix won't be applied. For fix to apply, a non-ISSU reset is required.
- Fix SN3800 specific warm boot scenario: Disable interface, Warm Boot, Enable Interface --> link will remain down.
Signed-off-by: Dror Prital <drorp@nvidia.com>
This is due to the fact that we use SONIC_OVERRIDE_BUILD_VARS internally
in our build jobs and this is not accounted in caching framework.
So we add MLNX_SDK_DEB_VERSION to force rebuild if we changed it via
SONIC_OVERRIDE_BUILD_VARS.
Signed-off-by: Stepan Blyschak <stepanb@nvidia.com>
Signed-off-by: Yong Zhao yozhao@microsoft.com
Why I did it
Currently we leveraged the Supervisor to monitor the running status of critical processes in each container and it is more reliable and flexible than doing the monitoring by Monit. So we removed the functionality of monitoring the critical processes by Monit.
How I did it
I removed the script process_checker and corresponding Monit configuration entries of critical processes.
How to verify it
I verified this on the device str-7260cx3-acs-1.
- Why I did it
Pick up fix from new hw-management package:
Fix gearbox thermal zone name, which was lack suffix thermal zone number
- How I did it
Update the hw-management version number in the make file
Update hw-management submodule pointer
- How to verify it
Run platform related test cases on Mellanox platform
#### Why I did it
According to thermalctld hld, each fan must belong to a fan drawer, if the fan drawer does not physically exist, put fan into a virtual fan drawer. This PR is to clear fan from chassis._fan_list
#### How I did it
1. Don't put fan to chassis._fan_list
2. Always query fan from fan_drawer
#### Why I did it
This pull request allows calls to be made through the platform 2.0 API that retrieve the PSU and Chassis hardware revision on Mellanox platforms. Access to these values will aid customers in determining their hardware revisions for debugging and technical support. These values are intended to be eventually exposed through the CLI.
#### How I did it
For the PSU hardware revision I used the existing VPD function calls implemented in https://github.com/Azure/sonic-buildimage/pull/7382
For the Chassis hardware revision I parsed the SMBIOS / DMI type 2 information to retrieve the information.
Originally, SFP modules were always accessed from platform daemons, and arbitrary SFP modules can be accessed in the daemon. So all SFP modules were initialized in one shot once one of the following chassis APIs called
- get_all_sfps
- get_sfp_numbers
- get_sfp
Recently, we noticed that SFP modules can also be accessed from CLI, eg. the latest refactor of `sfputil`.
In this case, only one SFP module is accessed in the chassis object's life cycle.
To initialize all SFP modules in one shot is waste of time and causes the CLI to take much more time to finish.
So we would like to optimize the initialization flow by introducing a two-phase initialization approach:
- Partial initialization, which means the `chassis._sfp_list` has been initialized with proper length and all elements being `None`
- Full initialization, which means all elements in `chassis._sfp_list` are created
If the relevant function is called,
- `get_sfp`, only partial initialization will be done, and then the specific SFP module is initialized.
- `get_all_sfps` or `get_num_sfps`, full initialization will be done, which means all SFP modules are initialized.
Signed-off-by: Stephen Sun <stephens@nvidia.com>
- Why I did it
Updated FW to xx.2008.2526 version.
Fixed issues:
1. Spectrum-2, Spectrum-3 | sFlow | High CPU load and high on fully loaded switch.
2. Spectrum-2, Spectrum-3 | Fine grain LAG | in rare cases doesn’t update the right entry
- How I did it
Updated submodule pointer and version in a Makefile.
- How to verify it
Full regression and bugs validation
Signed-off-by: Shlomi Bitton <shlomibi@nvidia.com>
#### Why I did it
We want to add the ability for the command `show platform psustatus` to show the serial number and part number of the PSU devices on Mellanox platforms. This will be useful for data-center management of field replaceable units (FRUs) on switches.
#### How I did it
I implemented the platform 2.0 functions `get_model()` and `get_serial()` for the PSU in the mellanox platform API by referencing the sysfs nodes provided by the [hw-management](https://github.com/Azure/sonic-buildimage/tree/master/platform/mellanox/hw-management) module.
#### Why I did it
Adopt a single way to get fan direction for all ASIC types.
It depends on hw-mgmt V.7.0010.2000.2303. Depends on https://github.com/Azure/sonic-buildimage/pull/7419
#### How I did it
Originally, the get_direction was implemented by fetching and parsing `/var/run/hw-management/system/fan_dir` on the Spectrum-2 and the Spectrum-3 systems. It isn't supported on the Spectrum system.
Now, it is implemented by fetching `/var/run/hw-management/thermal/fanX_dir` for all the platforms.
Signed-off-by: Stephen Sun <stephens@nvidia.com>
The following error message is observed during chassis object being destroyed
"Exception ignored in: <function Chassis.__del__ at 0x7fd22165cd08>
Traceback (most recent call last):
File "/usr/local/lib/python3.7/dist-packages/sonic_platform/chassis.py", line 83, in __del__
ImportError: sys.meta_path is None, Python is likely shutting down
The chassis tries to import deinitialize_sdk_handle during being destroyed for the purpose of releasing the sdk_handle.
However, importing another module during shutting down can cause the error because some of the fundamental infrastructures are no longer available."
This error occurs when a chassis object is created and then destroyed in the Python shell.
- How I did it
To fix it, record the deinitialize_sdk_handle in the chassis object when sdk_handle is being initialized and call the deinitialize handler when the chassis object is being destroyed
- How to verify it
Manually test.
- Why I did it
Upgrade hw-mgmt to 7.0100.2303
Bug fixes
1. Fan direction feature fix for fixed FAN system (using shell instead of binutils/strings)
2. Remove cpld 4th link on systems with only 3 CPLD's
3. hw-mgmt: thermal: Add hardcoded critical trip point. Follow-up after patch "Removing critical thermal zones to prevent unexpected software system shutdown".
4. Fix sensor attribute mapping to be label based instead of index based to allow common handling of voltage regulator names independently of hardware changes.
5. Update 'lm-sensors' custom configuration file. Relevant only for users utilizing sensors.conf files coming along with hw-management package.
6. For full feature list please follow https://github.com/Mellanox/hw-mgmt/blob/V.7.0010.2300_BR/debian/Release.txt
- How I did it
Update hw-mgmt pointer
Remove unused patches
Fix existing patch to make sure it apply successfully
- How to verify it
Full platform regression on all mellanox platforms
Changes in the new release:
Fix 10G and 50G speeds in SAI XML to support all interface types
Enable SMAC=DMAC and SMAC MC in tunnel debug counter
Add tunnel statistics
Add isolation group API implementation
Fix ACL ANY debug counter to correctly track ACL drops
Add VXLAN source port hard coded range, controlled by K/V
FW dump me now feature
Add mlxtrace to saidump
Speed lane setting and AN control
Implement query stats API
VNI miss part of tunnel decal drop reason
Align with SAI API v1.8.1
Signed-off-by: Dror Prital <drorp@nvidia.com>
Signed-off-by: Stepan Blyschak stepanb@nvidia.com
This PR is part of SONiC Application Extension
Depends on #5938
- Why I did it
To provide an infrastructure change in order to support SONiC Application Extension feature.
- How I did it
Label every installable SONiC Docker with a minimal required manifest and auto-generate packages.json file based on
installed SONiC images.
- How to verify it
Build an image, execute the following command:
admin@sonic:~$ docker inspect docker-snmp:1.0.0 | jq '.[0].Config.Labels["com.azure.sonic.manifest"]' -r | jq
Cat /var/lib/sonic-package-manager/packages.json file to verify all dockers are listed there.
New features and fixes in the new SDK/FW:
SN4600C | AN/LT support
SN2700 | AN/LT bugs fixes
WJH | FID_MISS support
Signed-off-by: Kebo Liu <kebol@nvidia.com>
- Why I did it
Add missed files for dynamic buffer calculation for ACS-MSN3420 and ACS-MSN4410
- How I did it
asic_table.j2: Add mapping from platform to ASIC
Add buffer_dynamic.json.j2 for ACS-MSN4410.
- How to verify it
Check whether the dynamic buffer calculation daemon starts successfully.
Signed-off-by: Stephen Sun <stephens@nvidia.com>
Integrate hw-management package V.7.0010.2002
Bug fixes:
Removing critical thermal zones to prevent unexpected software system shutdown:
*Kernel 4.9 -0071-mlxsw-core-Remove-critical-trip-point-from-thermal-z.patch
*Kernel 4.19 -076-mlxsw-core-Remove-critical-trip-point-from-thermal-z.patch
Removing redundant link for cpld3 for fixed systems (SN2100, SN2010).
Fix an issue with missed attribute for cpld3 (port CPLD) for SN2700, SN2410.
Signed-off-by: Stephen Sun <stephens@nvidia.com>
Fix the following issues:
Spectrum-2, Spectrum-3 | Port | Fix link issue when using 25 GbE rate between two ports while one is on Spectrum-2-based system and the other is on Spectrum-3-based system
All | warmboot | fail to upgrade from earlier SONiC versions with official SDK/FW 4.4.2306 (was on SONiC 201911)
All | What-Just-Happened | When enabling or disabling WJH under high traffic load to the host CPU, in very specific and low probability conditions, an error could occur, that may result in loss of data, channel failure or in extreme cases SW failure
Signed-off-by: Volodymyr Samotiy <volodymyrs@nvidia.com>
- Why I did it
The existing Fan led and Psu led object initialize itself to green color in init method. However, there are multiple daemons calls sonic platform API and there could be a case that:
A PSU is removed from system
Reboot switch
psud detects that 1 PSU is missing and set PSU led to red
Other daemon just start up and call sonic platform API, the API set PSU led to green by call PsuLed.init
This PR is a partial fix for the issue. As we also need guarantee that the led is initialized with a correct value. I checked existing psud and thermalctld code. psud always initialize the PSU led color on boot up, thermalcltd need some changes to initialize led color on the first run
- How I did it
Remove the led color initialization code from FanLed.init and PsuLed.init
- How to verify it
Manual test
There was a change to replace platform utils with sonic platform API in psuutil. However, psu API is not initialized on host side. The PR is to fix it.
- Why I did it
To pick up new features and fix from SDK/FW and SAI
SDK/FW new Feature:
All | Added support for multiple modules and cable types. For full list contact Nvidia networking support
Spectrum-3 | SN46000C | Added support for up to 5W on ports 49 to 64 .
SDK/FW bugs' fix:
All | fast reboot | fast boot failure from latest 201811 to 201911 and above
Spectrum | 10GbE/1GbE Transceiver (FTLX8574D3BCV) stopped working after firmware upgrade
Spectrum-2 | When device is rebooted with locked Optical Transceivers in split mode, the firmware may get stuck
Spectrum-2 | SN3700 | When connecting at 200GbE to Ixia K400, Ixia receives CRC errors
Spectrum-2 | SN3800 | On rare occasions packets loss may be experienced due to signal integrity issues
Spectrum-2 | When the port is a member of a LAG, after a warmboot and port toggle on the peer-side, the port remains down
Spectrum-3 | SN4700 | While using Optic cable in Split 4x1 mode in PAM4, when two first ports are toggled, the other 2 ports go down
Spectrum-3 | SN4700 | When working in 400GbE, deleting the headroom configuration (changing buffer size to zero) on the fly may cause continual packet drops
SAI
All | sFlow | Use hardcoded value 1 as netlink group number ax expected by hsflowd
- How I did it
Update the related version number in the make files and update the submodule pointer accordingly.
- How to verify it
Run regression test and everything works good.
#### Why I did it
Recently, CLI sfputil replace the old sonic platform utils with sonic platform API. However, sonic platform API does not support SFP low power mode and reset related operation. The PR is to fix it.
The change to replace platform utils with sonic platform API was reverted on 202012, once this PR is merged, we can cherry-pick these two PRs to 202012 together.
#### How I did it
In low power mode and reset related operation, use "docker exec" if the command is running on host side.
- Why I did it
Update MFT tool version to 4.16.0
Bugs fixes:
mlxlink: Fixed an issue that caused the margin scan to fail with the following message: Eye scan not completed.
mlxcable: Cable firmware burning capability is not supported.
New features:
mlxlink: Enabled margin scan on Network links.
mlxlink: Added PRBS TX/RX polarity inversion using the following flags: --invert_tx_polarity / --invert_rx_polarity
- How I did it
Update MFT make file with new version number.
- How to verify it
Build image and test related functions on Mellanox platform
- Why I did it
Bug fixes
- In rare cases when thermal algorithm is reactivated after FAN/PSU insertion, FAN remains at high rpm
- When stop hw-management code received error in the log instead of exit code '0'.
- In SPC1 i2c sometimes collide with chip reset coming from SDK
- Remove raw eeprom data link, when working with PSU which don't have eeprom for "msn274x", "msn24xx" and "msn27xx" systems
- Fix memory leak on mlxsw_core_bus_device module removal
- How I did it
Update the hw-mgmt version number in the make file
Update the hw-mgmt repo pointer
- How to verify it
run platform related test cases on all Mellanox platform
Signed-off-by: Kebo Liu <kebol@nvidia.com>
In preparation for the merging of Azure/sonic-platform-common#173, which properly defines class and instance members in the Platform API base classes.
It is proper object-oriented methodology to call the base class initializer, even if it is only the default initializer. This also future-proofs the potential addition of custom initializers in the base classes down the road.
- Why I did it
System is stuck on 'starting' state on SimX platform because of infinite loop on 'hw-management-ready.sh' script .
The loop is polling to check if the hw-mgmt sysfs created before proceeding with the flow, for SimX platform the sysfs will never create so the system is not starting properly.
- How I did it
Add a condition to poll on hw-mgmt sysfs only if the switch is real HW and not SimX platform.
- How to verify it
Check "systemctl status hw-management.service" output on a SimX switch with this patch, the state will be "active".
Signed-off-by: Shlomi Bitton <shlomibi@nvidia.com>
- Why I did it
Add support for new 64x200G SN4600 systems
- How I did it
Add all relevant files (w/o platform.json and hwsku.json as they will come later) with default SKU.
- How to verify it
Install image on switch, verify all ports are up and configured properly, run full platform SONiC tests.
- Why I did it
Mellanox SDK APIs support python 2 at the moment.
- How I did it
Mellanox SDK APIs support python 2 at the moment.
- How to verify it
Add python 2 to Mellanox syncd only.
- Which release branch to backport (provide reason below if selected)
docker exec -t syncd /bin/bash -c "sx_api_dbg_generate_dump.py /home/sx_api_dbg_dump"
You can see that it will work and generate /home/sx_api_dbg_dump
Signed-off-by: allas <allas@nvidia.com>
Open ACL Outer VLAN ID for egress for ports part of VLAN RIF
- Why I did it
Open ACL Outer VLAN ID for egress for ports part of VLAN RIF
- How I did it
Updated SAI submodule pointer
- How to verify it
Build an image, deploy and check all is up and running.
Verify ACL sonic-mgmt test is passing
Signed-off-by: Volodymyr Samotiy <volodymyrs@nvidia.com>
To have the following fixes:
* All | Port status remains down after warm boot and flapping the port on peer side
* All | LAG HASH | IPv6 SRC_IP is not accounted in LAG hashing [
* All | ASIC driver | Kernel crash observed when driver reload is initiated before it fully loaded
* Spectrum-3 | Buffer | In lossless configuration, headroom is been evicted only when the shared buffers is free
* All | prevent FW access during ISSU
Signed-off-by: Volodymyr Samotiy <volodymyrs@nvidia.com>
Include SAI bug fixes:
- Apply device MAC on port host interface when port is removed from LAG.
- [Shared Headroom]: fixed watermark handling for SHP flow
- Decrease verbosity of policer unbind message when no policer is attached
Signed-off-by: Stepan Blyschak <stepanb@nvidia.com>
Submodule commits included:
* src/sonic-platform-common 6ad0004...bd4dc03 (1):
> [sonic_sfp/qsfp_dd.py] Update DOM capability method name to align with other drivers (#163)
Also align all calling function names to match.
When Building syncd-rpc, libthrift has dependency on libboost-atomic1.71.0,
however the debian packager install version 1.67 instead. This PR
preinstalls libboost-atomic v 1.71 to avoid falling back to v 1.67.
signed-off-by: Tamer Ahmed <tamer.ahmed@microsoft.com>
**- Why I did it**
After migrating to python3, the operator '/' always get a float result, but it gets integer result in python2. Need fix this in thermal_conditions.
**- How I did it**
1. cast float value to int
2. change the unit test case to cover this situation
**- How to verify it**
Manually test and regression test
**- Why I did it**
PR https://github.com/Azure/sonic-platform-common/pull/102 modified the name of the SFF-8436 (QSFP) method to align the method name between all drivers, renaming it from `parse_qsfp_dom_capability` to `parse_dom_capability`. Once the submodule was updated, the callers using the old nomenclature broke. This PR updates all callers to use the new naming convention.
**- How I did it**
Update the name of the function globally for all calls into the SFF-8436 driver.
Note that the QSFP-DD driver still uses the old nomenclature and should be modified similarly. I will open a PR to handle this separately.
**- Why I did it**
Reduce the time it takes for the ASIC FW burn as part of the automatic FW upgrade procedure.
**- How I did it**
Add -d option to mlxfwmanager tool to use the faster MST device and not the default one which is not the fastest one.
**- How to verify it**
I manually changed ASIC FW followed by reboot command in order for FW upgrade to take place on deinit.
I manually changed ASIC FW followed by hard reset in order for FW upgrade to take place on init.
Signed-off-by: liora <liora@nvidia.com>
**Why I did it**
Disable SDK extended dump due to issue found
**How I did it**
Update SAI submodule
**How to verify it**
Verify the SDK extended dump is not called.
Signed-off-by: Eran Dahan <erand@nvidia.com>
- Why I did it
Fix issue: ptf_nn_agent isn't able to start in syncd-rpc docker on buster.
- How I did it
The issue is fixed by installing python-dev, cffi and nnpy for python 2 explicitly.
- How to verify it
Run copp test on RPC image.
Changes in the new release:
1. Policy based hashing optimization
2. New attribute support for Max port headroom
3. Tunnel ECN map fixes
4. Tunnel EVPN skeleton extensions (peer attrib, maps)
5. Bridge port admin not affecting port admin (optimize port down time)
6. CRM new API for neighbors and tunnel termination entries
7. Improve FDB event for flush by bridge port (before, null bridge was reported to SONiC, now the bridge will be extracted from bridge port)
8. DHCP L2 v4+v6 traps (for ZTP use case)
9. Generic counter implementation
Signed-off-by: Kebo Liu <kebol@nvidia.com>
- combine docker-ptf-saithrift into docker-ptf docker
- build docker-ptf under platform vs
- remove docker-ptf for other platforms
Signed-off-by: Guohan Lu <lguohan@gmail.com>
During ISSU, "mlxsw_minimal" driver still trying to access firmware, in some cases FW could return some wrong critical threshold value which will cause switch shutdown.
**- How I did it**
In order to prevent "mlxsw_minimal" driver from accessing ASIC during ISSU, SDK will raise "OFFLINE" 'udev' event
at the early beginning of such flow. When this event is received, hw-management will remove "mlxsw_minimal" driver.
There is no need to implement the opposite "ONLINE" event since this flow is ended up with "kexec".
**- How to verify it**
repeatedly perform warm reboot, make sure there is no switch shutdown occurred.
Bugs fixes:
All | Kernel | During system reload when CPU is loaded with heavy traffic, a Kernel Panic may occur.
All | Modules, Port split | FW stuck when device rebooted with locked Optical Transceivers in split mode
Spectrum-3 | PFC | On Spectrum-3 systems, slow reaction time to Rx pause packets on 40GbE ports may lead to buffer overflow on servers.
Spectrum-3 | SN4700, Port Split | On rare occasion SN4700, conducting 100G split (4x25G) in NRZ when splitter port 1 or 2 are down, ports 3 and 4 will also go down.
Enahncments:
All | Kernel | new notification on ISSU start, so other kernel drivers can disable any interface to ASIC
Signed-off-by: Kebo Liu <kebol@nvidia.com>
- Why I did it
Initially, we used Monit to monitor critical processes in each container. If one of critical processes was not running
or crashed due to some reasons, then Monit will write an alerting message into syslog periodically. If we add a new process
in a container, the corresponding Monti configuration file will also need to update. It is a little hard for maintenance.
Currently we employed event listener of Supervisod to do this monitoring. Since processes in each container are managed by
Supervisord, we can only focus on the logic of monitoring.
- How I did it
We borrowed the event listener of Supervisord to monitor critical processes in containers. The event listener will take
following steps if it was notified one of critical processes exited unexpectedly:
The event listener will first check whether the auto-restart mechanism was enabled for this container or not. If auto-restart mechanism was enabled, event listener will kill the Supervisord process, which should cause the container to exit and subsequently get restarted.
If auto-restart mechanism was not enabled for this contianer, the event listener will enter a loop which will first sleep 1 minute and then check whether the process is running. If yes, the event listener exits. If no, an alerting message will be written into syslog.
- How to verify it
First, we need checked whether the auto-restart mechanism of a container was enabled or not by running the command show feature status. If enabled, one critical process should be selected and killed manually, then we need check whether the container will be restarted or not.
Second, we can disable the auto-restart mechanism if it was enabled at step 1 by running the commnad sudo config feature autorestart <container_name> disabled. Then one critical process should be selected and killed. After that, we will see the alerting message which will appear in the syslog every 1 minute.
- Which release branch to backport (provide reason below if selected)
201811
201911
[x ] 202006
**- Why I did it**
On the Mellanox platform, reboot cause is fetched from some certain sysfs which is created by the hw-management service. So determine-reboot-cause service shall start after hw-management, otherwise it could fail due to the related sysfs is not available yet.
**- How I did it**
Add a patch to the hw-management service to make sure determine-reboot-cause service should start after it.
Signed-off-by: Kebo Liu <kebol@nvidia.com>
In rare case can see that xcvrd failed due to "UnboundLocalError: local variable 'label_port' referenced before assignment"
Init "label_port" as None at the beginning of the function, to avoid the case that "label_port" not assigned.
Features:
Spectrum-3 | Systems | Added GA-level support for SN4700 A0 system
All | Shared headroom | Added GA-level support for Shared headroom between PGs
Bugs fixes:
All | Counters | Sent traffic in certain size is wrongly increase to a smaller size counter, because port extended counter has a counter for sent traffic per packet-size range
All | Shared buffer | Configuring shared buffer on the fly may, on occasion, cause the chip to get stuck
Spectrum-2 | Modules | On occasion, link down is experienced with INPHI COLORZ PAM4 100G optic cables on SN3700 systems
