LEDs, sshd intermittent startup issue

I am having strange issues with reboots. Sometimes sshd does not come up, I get “Connection refused”. Luci shows sshd process is indeed not running, log does not have any messages from sshd. Luci, Startup, sshd is enabled. Click Start or Restart, sshd starts up without issues. Other boots is starts up fine.
LEDs sometimes all come up blue after a reboot. Router is working fine. Running /etc/rc.d/S99start-indicator returns LEDs to the desired config.
I am at loss on what is happening.
While looking at start-indicator noticed it should also log “Router Turris successfully started.” In my log it is right up among the first kernel boot messages when it should be the last rc scripts run. Is this just a log buffer issue?

Sep 28 18:06:49 turris syslog-ng[2769]: syslog-ng starting up; version='4.2.0'
Sep 28 18:06:49 turris kernel: [    0.000000] Booting Linux on physical CPU 0x0
Sep 28 18:06:49 turris kernel: [    0.000000] Linux version 5.15.127 (packaging@turris.cz) (arm-openwrt-linux-muslgnueabi-gcc (OpenWrt GCC 8.4.0 r16862+127-170d9e447d) 8.4.0, GNU ld (GNU Binutils) 2.34) #0 SMP Thu Aug 24 08:16:54 2023
Sep 28 18:06:49 turris kernel: [    0.000000] CPU: ARMv7 Processor [414fc091] revision 1 (ARMv7), cr=10c5387d
Sep 28 18:06:49 turris kernel: [    0.000000] CPU: PIPT / VIPT nonaliasing data cache, VIPT aliasing instruction cache
Sep 28 18:06:49 turris kernel: [    0.000000] OF: fdt: Machine model: Turris Omnia
Sep 28 18:06:49 turris kernel: [    0.000000] Memory policy: Data cache writealloc
Sep 28 18:06:49 turris kernel: [    0.000000] Zone ranges:
Sep 28 18:06:49 turris kernel: [    0.000000]   Normal   [mem 0x0000000000000000-0x000000002fffffff]
Sep 28 18:06:49 turris kernel: [    0.000000]   HighMem  [mem 0x0000000030000000-0x000000003fffffff]
Sep 28 18:06:49 turris kernel: [    0.000000] Movable zone start for each node
Sep 28 15:06:49 turris turris: Router Turris successfully started.
Sep 28 18:06:49 turris kernel: [    0.000000] Early memory node ranges

it’s the syslog daemon that fetches logs from kernel and other sources, so ir may fetch them in different order than they were created.

I had the same problem with random reboots, usually after ppp link failed. So far no reason found.

I have msata drive where I log everything, nothing in logs indicates a problem.

Also can confirm after mcu (with omnia-mcutool) upgrade to newest version all LEDs becomes blue (every second reboot). With old mcu firmware was no problem. TOS 6.4.4. Update details below:

root@turris:/usr/share/omnia-mcu-firmware# ./omnia-mcutool.static -v
Bootloader version: b5a8a24e007eb72be16aeb3fff6f03ec647023e4
Application version: b5a8a24e007eb72be16aeb3fff6f03ec647023e4
MCU type: STM32
Board firmware type: stm32-rev23-user-regulator

Features: 0x0000 (none)

after upgrade

root@turris:/usr/share/omnia-mcu-firmware# ./omnia-mcutool.static -v
Bootloader version: b5a8a24e007eb72be16aeb3fff6f03ec647023e4
Application version: 67eddc9540526d0a9d9660f7a7867af9a28a68d6
MCU type: STM32
Board firmware type: stm32-rev23-user-regulator
Features: 0x1f6e
EXT_CMDS
WDT_PING
LED_STATE_EXT
LED_GAMMA_CORRECTION
NEW_INT_API
FLASHING
NEW_MESSAGE_API
BRIGHTNESS_INT
POWEROFF_WAKEUP
CAN_OLD_MESSAGE_API
Application firmware length: 21400 Bytes
Application firmware checksum: 0xfeeb0f31