Why I did it
Networking devices need to be responsive. Such responsiveness is harmed when the CPU change state.
There is a latency penalty when a CPU is idle (e.g C2) and need to exit this state to come back to C1 state.
To prevent this from happening the CPU should be forced to remain in C1 state.
How I did it
Generalize the cstate forcing to C1 to all Arista products.
This is done by adding processor.max_cstate=1 to the kernel cmdline for all CPUs.
Additionally Intel CPUs also need intel_idle.max_cstate=0 to fallback to the acpi_idle driver.
How to verify it
Check that processor.max_cstate=1 is present on the cmdline for AMD CPUs
Check that both processor.max_cstate=1 and intel_idle.max_cstate=0 are present on the cmdline for Intel CPUs
Openssh in Debian Bullseye has been updated to 1:8.4p1-5+deb11u2 to fix CVE-2023-38408.
Since we're building openssh with some patches, we need to update our version as well.
Signed-off-by: Saikrishna Arcot <sarcot@microsoft.com>
* [swss] Chassis db clean up optimization and bug fixes
This commit includes the following changes:
- Fix for regression failure due to error in finding CHASSIS_APP_DB in
pizzabox (#PR 16451)
- After attempting to delete the system neighbor entries from
chassis db, before starting clearing the system interface entries,
wait for sometime only if some system neighbors were deleted.
If there are no system neighbors entries deleted for the asic coming up,
no need to wait.
- Similar changes for system lag delete. Before deleting the
system lag, wait for some time only if some system lag memebers were
deleted. If there are no system lag members deleted no need to wait.
- Flush the SYSTEM_NEIGH_TABLE from the local STATE_DB. While asic
is coming up, when system neigh entries are deleted from chassis ap
db (as part of chassis db clean up), there is no orchs/process running to
process the delete messages from chassis redis. Because of this, stale system
neigh are entries present in the local STATE_DB. The stale entries result in
creation of orphan (no corresponding data path/asic db entry) kernel neigh
entries during STATE_DB:SYSTEM_NEIGH_TABLE entries processing by nbrmgr (after
the swss serive came up). This is avoided by flushing the SYSTEM_NEIGH_TABLE from
the local STATE_DB when sevice comes up.
Signed-off-by: vedganes <veda.ganesan@nokia.com>
* [swss] Chassis db clean up bug fixes review comment fix - 1
Debug logs added for deletion of other tables (SYSTEM_INTERFACE and SYSTEM_LAG_TABLE)
Signed-off-by: vedganes <veda.ganesan@nokia.com>
---------
Signed-off-by: vedganes <veda.ganesan@nokia.com>
(cherry picked from commit b13b41fc22)
In #15080, there was a command added to re-add 127.0.0.1/8 to the lo
interface when the networking configuration is being brought down.
However, the trigger for that command is `down`, which, looking at
ifupdown2 configuration files, runs immediately after 127.0.0.1/16 is
removed. This means there may be a period of time where there are no
loopback addresses assigned to the lo interface, and redis commands will
fail.
Fix this by changing this to pre-down, which should run well before
127.0.0.1/16 is removed, and should always leave lo with a loopback
address.
Signed-off-by: Saikrishna Arcot <sarcot@microsoft.com>
Co-authored-by: Saikrishna Arcot <sarcot@microsoft.com>
#### Why I did it
To enable qos config for a certain backend deployment mode, for resource-type "Compute-AI".
This deployment has the following requirement:
- Config below enabled if DEVICE_TYPE as one of backend_device_types
- Config below enabled if ResourceType is 'Compute-AI'
- 2 lossless TCs' (2, 3)
- 2 lossy TCs' (0,1)
- DSCP to TC map uses 4 DSCP code points and maps to the TCs' as follows:
"DSCP_TO_TC_MAP": {
"AZURE": {
"48" : "0",
"46" : "1",
"3" : "3",
"4" : "4"
}
}
- WRED profile has green {min/max/mark%} as {2M/10M/5%}
This required template change <as in the PR> in addition to the vendor qos.json.j2 file (not included here).
### How I did it
#### How to verify it
- with the above change and the vendor config change, generated the qos.json file and verified that the objective stated in "Why I did it" was met
- verified no error
### Description for the changelog
Update qos_config.j2 for Comptue-AI deployment on one of backend device type roles
On S6100 we are seeing almost 100K interrupts per second on intels i801 SMBUS controller which affects systems performance.
We now disable the i801 driver interrupt and instead enable polling
Microsoft ADO (number only): 24910530
How I did it
Disable the interrupt by passing the interrupt disable feature argument to i2c-i801 driver
How to verify it
This fix is NOT applicable for ARM based platforms. Applicable only for intel based platforms:-
- On SN2700 its already disabled in Mellanox hw-mgmt
- Celestica DX010 and E1031
- Dell S6100 verified the interrupts are no longer incrementing.
- Arista 7260CX3
Signed-off-by: Prince George <prgeor@microsoft.com>
* [Mellanox] Update SDK/FW/SAI to 4.6.1020/2012.1020/SAIBuild2305.25.0.3 (#16096)
SONiC changes:
1. Support Spectrum4 ASIC FW binary building.
2. Support new SDK sx-obj-desc lib building since new SAI need it.
3. Remove SX_SCEW debian package from Mellanox SDK build since we are no longer using it (we use libxml2 instead).
4. Update SAI, SDK, FW to version 4.6.1020/2012.1020/SAIBuild2305.25.0.3
SDK/FW bug fixes
1. In SPC-1 platforms: Fastboot mode is not operational for Split port with Force mode in 50G speed
SFP modules are kept in disabled state after set LPM (low power mode) on/off for at least 3 minutes.
2. When preforming fast boot from an old SDK version (currently installed) to a newer one (target version), and the system was initially loaded with a new SDK version (past version), and the system has not been wiped, under specific conditions, the fast boot would use the past version's data and may fail.
SDK/FW Features
1. On SN2700 all ports can support y cable by credo
SAI bug Fixes
1. When creating an ACL rule with SAI_ACL_ENTRY_ATTR_FIELD_SRC_IP/SAI_ACL_ENTRY_ATTR_FIELD_DST_IP enabled, and then disabling the field by setting enable=false, a match on L3_type=IPv4 will remain programmed for the rule Issue resolved after the fix
2. Allow the max scale of virtual routers to be configure for SPC-1, SPC-2, SPC-3 when fastboot enable
3. Remove default hash key of SRC_MAC, DST_MAC and ETH_TYPE
SAI features
1. Port init profile
- How I did it
Update SDK/FW/SAI make files
- How to verify it
Run full sonic-mgmt regression on Mellanox platform
Signed-off-by: Kebo Liu <kebol@nvidia.com>
Conflicts:
platform/mellanox/mlnx-sai.mk
* Fix issue: unprintable character is rendered when handling comments in j2
Use "{#-" and "-#}" to mark comments in jinja template
Signed-off-by: Stephen Sun <stephens@nvidia.com>
---------
Signed-off-by: Stephen Sun <stephens@nvidia.com>
Co-authored-by: Stephen Sun <stephens@nvidia.com>
#### Why I did it
To fix the logic introduced by [[memory_checker] Do not check memory usage of containers which are not created #11129](https://github.com/sonic-net/sonic-buildimage/pull/11129).
There could be a scenario before the reboot, where
1. The `docker service` has stopped
2. In a very short period of time, the monit service performs the `root@sonic:/home/admin# monit status container_memory_telemetry`
In such scenario, the `memory_checker` script will throw an error to the syslog:
```
ERR memory_checker: Failed to retrieve the running container list from docker daemon! Error message is: 'Error while fetching server API version: ('Connection aborted.', FileNotFoundError(2, 'No such file or directory'))'
```
But, actually, this scenario is a correct behavior, because when the docker service is stopped, the Unix socket is destroyed and that is why we could see the `FileNotFoundError(2, 'No such file or directory'` exception in the syslog.
#### How I did it
Change the log severity to the warning and changed the return value.
#### How to verify it
It is really hard to catch the exact moment described in the `Why I did it` section.
In order to check the logic:
1. Change the Unix socket path to non-existing in [/usr/bin/memory_checker](47742dfc2c/files/image_config/monit/memory_checker (L139)) file on the switch.
2. Execute the `root@sonic:/home/admin# monit restart container_memory_telemetry`
3. Check the syslog for such messages:
```
WARNING memory_checker: Failed to retrieve the running container list from docker daemon! Error message is: 'Error while fetching server API version: ('Connection aborte
d.', FileNotFoundError(2, 'No such file or directory'))'
INFO memory_checker: [memory_checker] Exits without checking memory usage since container 'telemetry' is not running!
```
Why I did it
Support FIPS DB configuration
Design Doc: sonic-net/SONiC#1372
Work item tracking
Microsoft ADO (number only): 24411148
How I did it
Add the FIPS Yang model to make FIPS configurable in ConfigDB.
How to verify it
See TestPlan: sonic-net/sonic-mgmt#9092
Build the image and run the tests: sonic-net/sonic-mgmt#9091
Cherypick of #15685
MSFT ADO: 24274591
Why I did it
Two changes:
1 Fix a day1 issue, where check to wait until CONFIG_DB_INITIALIZED is incorrect.
There are multiple places where same incorrect logic is used.
Current logic (until [[ $($SONIC_DB_CLI CONFIG_DB GET "CONFIG_DB_INITIALIZED") ]];) will always result in pass, irrespective of the result of GET operation.
root@str2-7060cx-32s-29:~# sonic-db-cli CONFIG_DB GET "CONFIG_DB_INITIALIZED"
1
root@str2-7060cx-32s-29:~# until [[ $(sonic-db-cli CONFIG_DB GET "CONFIG_DB_INITIALIZED") ]]; do echo "entered here"; done
root@str2-7060cx-32s-29:~#
root@str2-7060cx-32s-29:~#
root@str2-7060cx-32s-29:~# sonic-db-cli CONFIG_DB GET "CONFIG_DB_INITIALIZED"
0
root@str2-7060cx-32s-29:~# until [[ $(sonic-db-cli CONFIG_DB GET "CONFIG_DB_INITIALIZED") ]]; do echo "entered here"; done
root@str2-7060cx-32s-29:~#
Fix this logic by checking for value of flag to be "1".
root@str2-7060cx-32s-29:~# until [[ $(sonic-db-cli CONFIG_DB GET "CONFIG_DB_INITIALIZED") -eq 1 ]]; do echo "entered here"; done
entered here
entered here
entered here
This gap in logic was highlighted when another fix was merged: #14933
The issue being fixed here caused warmboot-finalizer to not wait until config-db is initialized.
2 Set and unset CONFIG_DB_INITIALIZED for warm-reboot case
Currently, during warm shutdown CONFIG_DB_INITIALIZED's value is stored in redis db backup. This is restored back when the dump is loaded during warm-recovery.
So the value of CONFIG_DB_INITIALIZED does not depend on config db's state, however it remain what it was before reboot.
Fix this by setting CONFIG_DB_INITIALIZED to 0 as when the DB is loaded, and set it to 1 after db_migrator is done.
Work item tracking
Microsoft ADO (number only):
How I did it
How to verify it
cherry-pick: #14513
depends: https://github.com/sonic-net/sonic-utilities/pull/2939
* Add an ability to configure remote syslog servers
* Add an initial configuration for remote syslog
* Extend YANG module and add unit tests
#### Why I did it
Adding the following functionality to rsyslog feature:
* Configure remote syslog servers: protocol, filter, severity level
* Update global syslog configuration: severity level, message format
#### How I did it
added parameters to syslog server and global configuration.
#### How to verify it
create syslog server using CLI/adding to Redis-DB
verify server is added to file /etc/rsyslog.conf and server is functional.
#### Description for the changelog
extend rsyslog capabilities, added server and global configuration parameters.
#### Link to config_db schema for YANG module changes
[sonic-syslog.yang](https://github.com/sonic-net/sonic-buildimage/blob/master/src/sonic-yang-models/yang-models/sonic-syslog.yang)
Fixes#15667 and #13293
Work item tracking
Microsoft ADO 24472854:
How I did it
On chassis supervisor bgp feature is disabled in hostcfgd. The dependency between swss and bgp causes the bgp containers to start even though the feature is disabled.
How to verify it
Tests on chassis supervisor and LC
Co-authored-by: Arvindsrinivasan Lakshmi Narasimhan <55814491+arlakshm@users.noreply.github.com>
Why I did it
Sharing the storage of syncd with other proprietary application extensions allows them to communicate with syncd in differnt ways.
If one container wants to pass some information to syncd then shared storage can be used. However, today the shared storage isn't cleaned on restarts making it possible for syncd to read out-of-date information generated in the past.
NOTE: No plans to use it for standard SONIC dockers and we are working on removing the SDK dependency from PMON docker
How I did it
Implemented new service to clean the shared storage.
How to verify it
Do reboot/fast-reboot/warm-reboot/config-reload/systemctl restart swss and verify /tmp/ is cleaned after each restart in syncd container.
Signed-off-by: Stepan Blyschak <stepanb@nvidia.com>
Co-authored-by: Stepan Blyshchak <38952541+stepanblyschak@users.noreply.github.com>
There is a redundant line in init_cfg.json.j2. It would cause pmon service always has "delayed=False". However, we know that PMON has a timer now. So, I try to fix it here.
Why I did it
To reduce the container's dependency from host system
Work item tracking
Microsoft ADO (number only):
17713469
How I did it
Move the k8s container startup script to config engine container, other than mount it from host.
How to verify it
Check file path(/usr/share/sonic/scripts/container_startup.py) inside config engine container.
Signed-off-by: Yun Li <yunli1@microsoft.com>
Co-authored-by: Qi Luo <qiluo-msft@users.noreply.github.com>
* Re-add 127.0.0.1/8 when bringing down the interfaces
With #5353, 127.0.0.1/16 was added to the lo interface, and then
127.0.0.1/8 was removed. However, when bringing down the lo interface,
like during a config reload, 127.0.0.1/16 gets removed, but 127.0.0.1/8
isn't added back to the interface. This means that there's a period of
time where 127.0.0.1 is not available at all, and services that need to
connect to 127.0.01 (such as for redis DB) will fail.
To fix this, when going down, add 127.0.0.1/8. Add this address before
the existing configuration gets removed, so that 127.0.0.1 is available
at all times.
Note that running `ifdown lo` doesn't actually bring down the loopback
interface; the interface always stays "physically" up.
Signed-off-by: Saikrishna Arcot <sarcot@microsoft.com>
Add watchdog mechanism to swss service and generate alert when swss have issue.
**Work item tracking**
Microsoft ADO (number only): 16578912
**What I did**
Add orchagent watchdog to monitor and alert orchagent stuck issue.
**Why I did it**
Currently SONiC monit system only monit orchagent process exist or not. If orchagent process stuck and stop processing, current monit can't find and report it.
**How I verified it**
Pass all UT.
Manually test process_monitoring/test_critical_process_monitoring.py can pass.
Add new UT https://github.com/sonic-net/sonic-mgmt/pull/8306 to check watchdog works correctly.
Manually test, after pause orchagent with 'kill -STOP <pid>', check there are warning message exist in log:
Apr 28 23:36:41.504923 vlab-01 ERR swss#supervisor-proc-watchdog-listener: Process 'orchagent' is stuck in namespace 'host' (1.0 minutes).
**Details if related**
Heartbeat message PR: https://github.com/sonic-net/sonic-swss/pull/2737
UT PR: https://github.com/sonic-net/sonic-mgmt/pull/8306
For T2 systems using packet mode, the backplane interfaces (Ethernet-BP#) and the fabric card ethernet interfaces are not visible as neighbor interfaces.
In packet mode, these interfaces needs qos and buffer config as well.
This fix addresses that issue and adds the backplane interfaces to the PORTS_ACTIVE list
Why I did it
After docker_inram is enabled, the docker folder's default max size is 1.5G.
It's not big enough for some tests which need to install additional docker images or install extra packages.
Work item tracking
Microsoft ADO 24199761:
How I did it
add docker_inram into cmdline_allowlist
How to verify it
sudo sh -c 'echo "docker_inram_size=3000M" >> kernel-cmdline-append'
sudo reboot and check the docker folder size
This reverts commit 44427a2f6b.
Docker image not updated during PR validation and caused PR check failures.
Force merge this revert. After cache is updated after this PR is merged, issue should be fixed.
This PR depends on https://github.com/sonic-net/sonic-swss/pull/2737 merge first.
**What I did**
Add orchagent watchdog to monitor and alert orchagent stuck issue.
**Why I did it**
Currently SONiC monit system only monit orchagent process exist or not. If orchagent process stuck and stop processing, current monit can't find and report it.
**How I verified it**
Pass all UT.
Add new UT https://github.com/sonic-net/sonic-mgmt/pull/8306 to check watchdog works correctly.
Manually test, after pause orchagent with 'kill -STOP <pid>', check there are warning message exist in log:
Apr 28 23:36:41.504923 vlab-01 ERR swss#supervisor-proc-watchdog-listener: Process 'orchagent' is stuck in namespace 'host' (1.0 minutes).
**Details if related**
Heartbeat message PR: https://github.com/sonic-net/sonic-swss/pull/2737
UT PR: https://github.com/sonic-net/sonic-mgmt/pull/8306
In the PR sonic-net/sonic-utilities#2850 , for support remote access of linecards paramiko package is installed in sonic-utilities. libffi-dev needs to installed to be able to compile for armhf image
Signed-off-by: Arvindsrinivasan Lakshmi Narasimhan <arlakshm@microsoft.com>
Why I did it
Fix the issue where db_migrator is called before DB is loaded w/ config. This leads to db_migrator:
Not finding anything, and resumes to incorrectly migrate every missing config
This is not expected. migration should happen after the old config is loaded and only new schema changes need migration.
Since DB does not have anything when migrator is called, db_migrator fails when some APIs return None.
The reason for incorrect call is that:
database service starts db_migrator as part of startup sequence.
config-setup service loads data from old-config/minigraph. However, since it has Requires=database.service.
Hence, config-setup starts only when database service is started. And database service is started when db_migrator is completed.
Fixed by:
Check if this is first time boot by checking pending_config_migration flag.
If pending_config_migration is enabled, then do not call db_migrator as part of database service startup.
Let database service start which triggers config-setup service to start.
Now call db_migrator after when config-setup service loads old-config/minigraph
