Netgate SG-3100 "bans" access point
-
While it's "banned" I can access the Archer only directly: via any of its ethernet ports or from its WiFi.
It keeps its assigned ip (assigned from pfsense), unless I restart it.
From the rest of the net it's just dead.I got 22.05-RELEASE.
Archer is also on its latest firmware. -
@gich Please connect to another LAN port on the 3100 and try to ping the Archer and the pfSense LAN IP.
Assuming you set up VLANs on all the ports on the 3100 switch you can move ports.
Try swapping out cables.
-
@rcoleman-netgate The two aren't directly connected. There is at least an unmanaged switch in the middle.
Anyway tested different ports and cables.I think I'll move the Archer to a friend house, to test it with another router.
-
@gich Remove the unmanaged switch, then, too.
Troubleshooting includes replacing and swapping all things. There could be an ARP storm caused by the switch that is only effecting your AP.
-
@rcoleman-netgate I thought something that simple would be less likely to cause any problem.
It has a lot of things connected to it.
Anyway I'll try to bypass it. -
@gich There are 4 LAN ports on the 3100. If they're not changed from the default LAN interface you can plug into any of them.
But to state the 3100 "bans" a device that is multiple links down the chain is not likely at all. there would be a record in the system somewhere... unless the device in the middle is the one that is losing its place. MAC table corrupted, maxed out, ARP storm, switch is failing, etc.
The "simple" things are the ones most likely to cause a problem when they meet a challenge because they're simply not geared towards the expectations and performance of an active network. A cheap 8-port switch might have a 1k MAC table and that can fill in minutes or hours depending on how much is happening.
-
@rcoleman-netgate But it's not a new setup, everything is in place for more then 18 months.
I find odd that the thing that does the less complicated job starts to misbehave.
I thought since pfsense is the more complex piece, it was the most likely to have something gone wrong.Anyway, I'll try that to be sure.
-
@gich And hardware can (and will) fail. I would consider trying to reboot the intermediary switch next time instead of the pf and see if that resolves the issue.
-
@gich said in Netgate SG-3100 "bans" access point:
@rcoleman-netgate But it's not a new setup, everything is in place for more then 18 months.
I love when people say that. I must hear it 4 times a week.
I've been in the electronics field all my adult life. One thing never fails, you can take a piece of electronics right out of the box and have a resistor blow, or it could run "for 18 months" and have a resistor blow. Ya just don't know.
I always tell them, all electronic devices runs on smoke, once you let the smoke out, it won't work anymore!
-
@jarhead The switch has 7 more things plugged that have no problem.
I'm not saying it can't be the one that failed, but it's not at the top of my list.
I mean, if it was failing I'd expect something more noticeable.Restarting the switch did nothing.
Chronologically the last change was the update to pfsense, so that's why I'm here.
-
@jarhead Also remember that restarting pfsense, does put the Archer back in the game.
That's why I'd not focus elsewhere. -
@gich Without being willing to accept that something else might be causing the technical issue, or trying to do the recommended troubleshooting, will tend to others abandoning their attempts to assist you.
-
The only thing in a default pfSense install that could present like this is the sshgiard login protection.
When it appears to be 'banned' check Diag > Tables and look for any entries in the sshguard table. If the AP IP is shown there that would do it. It would get 'unbanned' after some time though.
Also that would only affect other devices connected to the AP if it was acting a as router and NATing all the traffic from wifi clients.Steve
-
@rcoleman-netgate Did you miss the "I'll try that to be sure" ???
I was explaining why I was pointed in another direction. -
There are some tests we can do to be sure. It 'feels' like a rogue dhcp server or IP conflict though.
-
So I tried to connect directly to the Netgate via another port. Nothing.
Then I removed the static ip on it and, while I was looking around, I had a glimpse on the ARP table of a "incomplete MAC".
WTF is that? Busted port? But it works when I connect the notebook.So since it was already planned: moved the Archer to a friend house, very basic setup, and it works fine for hours.
While this was going on I resetted the Netgate and reloaded the configuration just backupped.
Archer is back at home where it was before and going strong all night.Early to tell if this is definitive, since it might have worked for that long before, but I'm hopeful.
Still no idea if the "incomplete MAC" was real or a dream and what that might mean.
-
@gich said in Netgate SG-3100 "bans" access point:
Still no idea if the "incomplete MAC" was real or a dream and what that might mean.
Means it arped but didn't get an answer, like this if i ping an IP that is not actually there
? (192.168.9.33) at (incomplete) on igb0 expired [ethernet]
There was no answer to the arp, so its incomplete.
-
@gich said in Netgate SG-3100 "bans" access point:
So I tried to connect directly to the Netgate via another port. Nothing.
Ah, so you were unable to connect to the 3100 at all when this happens?
I assume you tried only one of the other LAN ports? The AP is connected to a LAN port also?
If you have not yet enabled the OPT port for local access I would do that. You can then try to connect via that and it doesn't rely on the on-board switch config. One thing that could explain a layer2 failure like this is of the switch config is changed somehow.
You can easily check that by running at the command line:etherswitchcfg
But to do that you need to have access to the 3100. The OPT port would give you that but you could also use the console directly.Steve
-
@stephenw10 I had actually tried a couple of ports on both ends.
After the reset of the Netgate I don't have the situation anymore so I can't investigate further.Note: I don't think it matters but one of the Netgate's port won't do gigabit anymore and goes only to 100.
-
Hmm, could be a physical port problem then. Just avoiding that port might prevent it happening again.
Steve
