Downloading a large file using the hAPac as a (not transparent) proxy on my home network. I use the proxy for some machines on my network to only access whitelisted sites and deny everything else. the hAPac also runs capsman and does routing between a few vlans - average approx 20Mbs traffic.

Downloading at around 70 Mbs with the CPU pegged at 100%. After 10 minutes or so the router rebooted due to watchdog timer. The download started again, and the router rebooted a few minutes later again. Profiler was around 30% proxy, 20% ethernet 15% firewall from memory

I downgraded this router to 6.42.9 and the download completed successfully with no further reboots.

Anyone come across this before? If no immediate explanation, I can upgrade again and capture logs and config.

No one else seen this?

Downloading a large file using the hAPac as a (not transparent) proxy on my home network. I use the proxy for some machines on my network to only access whitelisted sites and deny everything else. the hAPac also runs capsman and does routing between a few vlans - average approx 20Mbs traffic.

Downloading at around 70 Mbs with the CPU pegged at 100%. After 10 minutes or so the router rebooted due to watchdog timer. The download started again, and the router rebooted a few minutes later again. Profiler was around 30% proxy, 20% ethernet 15% firewall from memory

I downgraded this router to 6.42.9 and the download completed successfully with no further reboots.

Anyone come across this before? If no immediate explanation, I can upgrade again and capture logs and config.

We have more than 100 MikroTik devices in our network, and since upgrading to 6.43 (and especially 6.43.2) we have been seeing multiple devices being rebooted by the watchdog timer: mostly devices with slower CPUs like 1st gen OmniTiks, but a few newer devices that have high CPU load. Including our edge router, yesterday morning. This morning, both CPU cores (a 3Ghz x86) are pegged at 100%, with the "ssl" task taking up about 50% of each core.

I figure it's only a matter of time before the watchdog timer reboots it again, but we're scheduling a reboot for midnight in case it makes it that far. In the meantime, I have no idea how this is impacting performance, like speed or SIP quality, but taking all our subscribers offline during business hours would arguably be worse.

I'm less than thrilled. Seriously considering reverting everything to 6.42 until MT gets this sorted out.

+1
we faced the same issue. After upgrade to 6.43.2 our hAP AC became to hang up from time to time.
So we had to downgrade it. 6.42.9 works more stable.

+1
Recently have noticed this problem both on hAP ac and hEX PoE. Never had any instabilities likes this before 6.43.2+ Router OS whatsoever.

Time Oct/22/2018 19:15:05
Buffer memory
Topics
system
error
critical
Message router was rebooted without proper shutdown by watchdog timer