[reboot-cause] failed to get reboot cause

[Xichen96]

Description

Reboot tests failed on getting reboot cause, if they are not stopped by other checks first

Steps to reproduce the issue

1. reboot
2. show reboot-cause

Describe the results you received

Unknown

Describe the results you expected

User issued 'reboot' command [User: admin, Time: xxx]

Additional information you deem important (e.g. issue happens only occasionally)

Output of show version

(paste your output here)

[Xichen96]

Output of show version

SONiC Software Version: SONiC.20201231.23
Distribution: Debian 10.10
Kernel: 4.19.0-12-2-amd64
Build commit: 7c791db
Build date: Fri Sep 3 12:17:54 UTC 2021
Built by: AzDevOps@sonic-int-build-workers-0003EK

Platform: x86_64-cel_e1031-r0
HwSKU: Celestica-E1031-T48S4
ASIC: broadcom
ASIC Count: 1
Serial Number: R0882F2B039723BY000014
Uptime: 05:28:56 up 54 min, 1 user, load average: 1.83, 2.08, 2.18

Docker images:
REPOSITORY TAG IMAGE ID SIZE
docker-syncd-brcm 20201231.23 6dea96746b36 694MB
docker-syncd-brcm latest 6dea96746b36 694MB
docker-snmp 20201231.23 49c04cec0029 443MB
docker-snmp latest 49c04cec0029 443MB
docker-teamd 20201231.23 36f555b4fef8 412MB
docker-teamd latest 36f555b4fef8 412MB
docker-router-advertiser 20201231.23 a90afecca506 402MB
docker-router-advertiser latest a90afecca506 402MB
docker-platform-monitor 20201231.23 c920d7e90943 612MB
docker-platform-monitor latest c920d7e90943 612MB
docker-lldp 20201231.23 10756e3f3d1a 442MB
docker-lldp latest 10756e3f3d1a 442MB
docker-dhcp-relay 20201231.23 c4a2e143c2de 409MB
docker-dhcp-relay latest c4a2e143c2de 409MB
docker-database 20201231.23 17492b436856 402MB
docker-database latest 17492b436856 402MB
docker-orchagent 20201231.23 78137aebb544 431MB
docker-orchagent latest 78137aebb544 431MB
docker-sonic-telemetry 20201231.23 df8929ddd2ec 491MB
docker-sonic-telemetry latest df8929ddd2ec 491MB
docker-fpm-frr 20201231.23 cacf6e6cd65c 431MB
docker-fpm-frr latest cacf6e6cd65c 431MB

[Xichen96]

@sujinmkang This is the issue I opened

[sujinmkang]

@Xichen96 Is this specific platform issue? If so, what is the platform?

[sujinmkang]

@Xichen96 Please specify what the commend of the reboot you performed on which platform.,

[Xichen96]

It turns out that if I let determine-reboot-cause to sleep for 30s before running, it produces the expected result. Otherwise, determine-reboot-cause fail mid execution.

[Xichen96]

determine-reboot-cause script hangs on command "docker", which is used to determine if the script is running inside of a container. "docker" hangs because it is called too early, before dockerd is fully initialized. "docker" can be replaced by a read on /proc/1/comm and checking if it is "systemd", but it will fail on other problems. Making the script run later in booting up solved all problems.