SG-2100 unresponsive after powering down nearby Synology NAS

Cabledude

Hi all,
This could hardly be coincidence. When I powered down my Synology NAS DS720+, which is in the same cabinet as the SG-2100 Max and just a couple of inches away, the Netgate lost its way. No routing, no DHCP, nothing. Had to perform hard power down (unplug) and reboot.
The power plugs are plugged into the same power strip. Nothing in the logs either. What gives?

Worth mentioning is this separate topic about the WARNING in the logs below, though these may be totally unrelated.

Here is the log just before and after the reboot:

May 18 19:32:53	kernel		Copyright (c) 1979, 1980, 1983, 1986, 1988, 1989, 1991, 1992, 1993, 1994
May 18 19:32:53	kernel		Copyright (c) 1992-2024 The FreeBSD Project.
May 18 19:32:53	kernel		WARNING: DTB version is 6.4 while kernel expects 6.8, please update the DTB in the ESP
May 18 19:32:53	kernel		KDB: current backend: ddb
May 18 19:32:53	kernel		KDB: debugger backends: ddb gdb
May 18 19:32:53	kernel		GDB: current port: uart
May 18 19:32:53	kernel		GDB: debug ports: uart
May 18 19:32:53	kernel		---<<BOOT>>---
May 18 19:32:53	syslogd		kernel boot file is /boot/kernel/kernel
May 18 19:00:35	php	96168	[pfBlockerNG] No changes to Firewall rules, skipping Filter Reload
May 18 19:00:01	php	96168	[pfBlockerNG] Starting cron process.
May 18 18:00:35	php	74739	[pfBlockerNG] No changes to Firewall rules, skipping Filter Reload
May 18 18:00:01	php	74739	[pfBlockerNG] Starting cron process.
May 18 17:00:34	php	52745	[pfBlockerNG] No changes to Firewall rules, skipping Filter Reload
May 18 17:00:01	php	52745	[pfBlockerNG] Starting cron process.
May 18 16:00:34	php	39197	[pfBlockerNG] No changes to Firewall rules, skipping Filter Reload
May 18 16:00:01	php	39197	[pfBlockerNG] Starting cron process.

At present there I have a normally operating unit but I still post this to ask if anyone thinks there is a connection.

After the reboot I powered down the Synology NAS again, but this time the Netgate unit didn't flinch.

Thanks,
Pete

stephenw10

Where were you testing from when it failed?

One possibility is that the switch lost it's config. I have seen that given a big enough power surge. But that would only affect LAN side devices since the WAN in the 2100 is independent of the switch.

Try running etherswitchcfg at the CLI to see how the switch should return. Then run it again from the console if you can replicate the issue. Assuming the console is still responsive.

Cabledude

@stephenw10 said in SG-2100 unresponsive after powering down nearby Synology NAS:

Where were you testing from when it failed?

Hi Steve, thanks for your insights. We all lost internet connection so I went troubleshooting various nodes in my network. As DHCP was down I wired my laptop to some LAN ports of the SG-2100, including the LAN1 which has all the VLANs. No IP.

One possibility is that the switch lost its config. I have seen that given a big enough power surge. But that would only affect LAN side devices since the WAN in the 2100 is independent of the switch.

The WAN also went down, at least the WAN side lost ping capability according to UptimeRobot. So do you still think this could be a switch config lost issue? Looks bigger to me. Look at this:

I rebooted the unit at 19:32
UptimeRobot reports Down: 19:16 / UP: 19:37
Log entries missing from 19:00 - 19:32
I am certain that all operation was normal well after 19:00.
So there are logs missing from 19:00 - 19:16

Try running etherswitchcfg at the CLI to see how the switch should return. Then run it again from the console if you can replicate the issue. Assuming the console is still responsive.

The unit has operated normally since the event. Nevertheless, I ran etherswitchcfg from CLI and I see ports and VLAN configs that look okay.

I haven't tried to replicate the issue any further.

stephenw10

@Cabledude said in SG-2100 unresponsive after powering down nearby Synology NAS:

The WAN also went down, at least the WAN side lost ping capability according to UptimeRobot. So do you still think this could be a switch config lost issue?

No, not assuming that the WAN is still using the WAN port (mvneta0). That is independent of the switch.

Hard to say what happened without logs or console output from the time it was happening.

Cabledude

@stephenw10 I understand, no logs, no clues. Well then in this case, given no logs, there's not much we can do now.

Just one thing: do log entries tend to go missing more often, when units become unresponsive? In this case, there must be missing log entries from around 19:00 for a couple of minutes at least.

Thanks,

pete

stephenw10

It depends why they stopped repsonding and how unresponsive they actually are.

If it has some network issue I would expect it to keep logging as normal.

If it had some drive or filesystem error it may continue to function for some time but be unable to log anything.

It can lock up hard and stop logging or responding entirely but that is far less common. Usually some low level OS bug.