Backporting #13969
Why I did it
Implementing code changes for sonic-net/SONiC#1203
Work item tracking
Microsoft ADO (number only):
How I did it
Removed the timers and delayed target since the delayed services would start based on event driven approach.
Cleared port table during config reload and cold reboot scenario.
Modified yang model, init_cfg.json to change has_timer to delayed
How to verify it
Added UT to verify
Before this change, a process running inside every SONiC container dealt with FEATURE table 'auto_restart' field and depending on the value decided whether a container has to be killed or not.
If killed service auto restart mechanism restarts the container.
This change moves the logic from container to the host daemon - hostcfgd.
The 'auto_restart' handling is kept in supervisor-proc-exit-listener but now it is not required for container that wants to support auto restart feature.
hostcfgd refactoring - move feature handling in another class.
override systemd service Restart= setting from hostcfgd.
remove default systemd Restart=always.
Signed-off-by: Stepan Blyshchak stepanb@nvidia.com
- Why I did it
Remove the need to deal with container orchestration logic from the container itself. Leave this logic to the orchestrator - host OS.
- How I did it
hostcfgd configures 'Restart=' value for systemd service.
- How to verify it
root@r-tigon-11:/home/admin# sudo config feature autorestart lldp enabled
root@r-tigon-11:/home/admin# show feature status | grep lldp
lldp enabled enabled
root@r-tigon-11:/home/admin# docker exec -it lldp pkill -9 lldpd
root@r-tigon-11:/home/admin# docker ps -a | grep lldp
65058396277c docker-lldp:latest "/usr/bin/docker-lld…" 2 days ago Exited (0) 20 seconds ago lldp
root@r-tigon-11:/home/admin# docker ps -a | grep lldp
65058396277c docker-lldp:latest "/usr/bin/docker-lld…" 2 days ago Up 5 seconds lldp
root@r-tigon-11:/home/admin# sudo config feature autorestart lldp disabled
root@r-tigon-11:/home/admin# docker exec -it lldp pkill -9 lldpd
root@r-tigon-11:/home/admin# docker ps -a | grep lldp
65058396277c docker-lldp:latest "/usr/bin/docker-lld…" 2 days ago Up 35 seconds lldp
root@r-tigon-11:/home/admin# docker ps -a | grep lldp
65058396277c docker-lldp:latest "/usr/bin/docker-lld…" 2 days ago Exited (0) 3 seconds ago lldp
root@r-tigon-11:/home/admin# docker ps -a | grep lldp
65058396277c docker-lldp:latest "/usr/bin/docker-lld…" 2 days ago Exited (0) 39 seconds ago lldp
root@r-tigon-11:/home/admin#
- Why I did it
Group all SONiC services together and able to manage them together. Will be used in config reload command as much simpler and generic way to restart services.
- How I did it
Add services to sonic.target
- How to verify it
Together with Azure/sonic-utilities#1199
config reload -y
Signed-off-by: Stepan Blyshchak <stepanb@nvidia.com>
Delay CPU intensive services at boot
- How I did it
Made snmp.timer work and add telemetry.timer.
But this is not enough because it breaks the existing snmp dependency on swss.
So, in this solution snmp timer is a wanted by swss service, but since OnBootSec timer expires only once it will not trigger snmp service, so I added line "OnUnitActiveSec=0 sec" which will start snmp service based on the last time it was active. On boot only OnBootSec will expire, on swss start/restarts only second timer will expire immediately and trigger snmp service.
However, snmp service will not stop after "systemctl stop snmp" because of the second timer which will always expire when snmp service because unavailable.
So there is a conflict which will be handled by systemd if we add "Conflicts=" line to both snmp.service and snmp.timer.
So during boot:
snmp does not start by default
swss starts and starts snmp timer
OnUnitActiveSec=0 does not expire since there is no snmp active
OnBootSec expires and starts snmp service and snmp timer gets stopped
During "systemctl restart swss"
snmp stops because of Requisite on swss
snmp unblocks snmp timer from running
swss starts and starts snmp timer
OnUnitActiveSec=0 expires imidiately and start snmp which stops snmp timer
During "systemctl stop snmp"
stop of snmp service unblocks snmp timer but no one starts the timer so it is not started by "OnUnitActiveSec=0"
* [service] Restart SwSS Docker container if orchagent exits unexpectedly
* Configure systemd to stop restarting swss if it attempts to restart more than 3 times in 20 minutes
* Move supervisor-proc-exit-listener script
* [docker-dhcp-relay] Enhance wait_for_intf.sh.j2 to utilize STATEDB
* Ensure dependent services stop/start/restart with SwSS
* Change 'StartLimitInterval' to 'StartLimitIntervalSec', as Stretch installs systemd 232 (>= v230)
* Also update journald.conf options
* Remove 'PartOf' option from unit files
* Add '$(SUPERVISOR_PROC_EXIT_LISTENER_SCRIPT)' to new shared docker-orchagent makefile
* Make supervisor-proc-exit-listener script read from 'critical_processes' file inside container
* Update critical_processes file for swss container
- Move all minigraph-related action from rc.local to updategraph
- updategraph service is now after database. All feature services are now after and depending on updategraph
snmp.service needs to read chassis serial number for one of its mibs.
We save this value in state DB so that it is accessible from container.
Signed-off-by: marian-pritsak <marianp@mellanox.com>
When Type=simple, systemd will consider the service activated immediately
after specified in ExecStart process is started. If there is downstream
service depending on the state prepared in ExecStart, there will be race condition.
For example, issue #390. In this case, database.service calls database.sh, which
calls docker run or docker start -a to start database container. However, systemd
considers database.service successfully started at the time database.sh begins,
not after docker run finishes. As database.service is consider started, bgp.service
can be started. The redis database, which bgp service depends on, might or might not
have been started at this time point.
To fix this issue (and still keeping the functionality to monitor docker status with
systemd), we split the ExecStart process into an ExecStartPre part and an ExecStart
part. docker run is splitted into docker run -d then docker attach , while docker start
-a is splitted into docker start and then docker attach. In this way, we make sure
the downstream services are blocked until container is successfully started.
* Single image
* Fix review comments
* Update syncd service. Add HW mgmt to Mellanox single image.
* Add single image template for Broadcom platform.
SKU should be provided during configure:
make configure PLATFORM=broadcom SKU=Force10-S6000
* Add single image template for Cavium platform.
SKU should be provided during configure:
make configure PLATFORM=cavium SKU=AS7512
* Add description to sonic_debian_extension.j2 file.
