gah... this is still an issue for me as well. RouterOS v6.43.12. I disable/enabled the dhcp client and it comes back online cleanly. Anyone out there have any luck working around this bug with mikrotik capability with cable modems?

It has to be a bug with the Atheros 8316. Tried, vlan > bridge > pppoeclient. Same result. Tried lower MTU, same result. Seems to work ok with the Atheros 7240, but CPU isn't fast enough to handle my load. If I remove the vlan, and rearrange my vlans so pppoe comes in on the ethernet interface untag...

tl;dr?

this bug exists in 5.19 through 5.22, and likely prior versions as well.

Please take pity on me as I've been troubleshooting an issue for 2 days straight. I've stripped every possible configuration out of my Mikrotik to narrow down the problem as much as I can. When pppoe-client runs over ethernet interface, everything works just fine. Unfortunately, when pppoe-client ru...

I'm in the same boat. I had a great multihomed home router, two internet connections, queue tree, QoS, the works. It ran great up until the day my ISP gave me more bandwidth than the router's CPU could handle. Averages around 80% CPU load, and will occasionally hit 100% and lockup for a few minutes ...

I'm really having a hard time here. I've tried redirect but it still doesn't work. I see this on the remote: 21:35:29.843898 ARP, Request who-has 192.168.2.2 tell 192.168.2.1, length 42 21:35:29.844051 ARP, Reply 192.168.2.2 is-at 00:14:bf:3c:1c:1b, length 28 On the Mikrotik, I see both the request ...

Drop invalid packets in firewall.

Duh. Thank you. =) Sometimes you need someone to point out the obvious. I already had a rule to drop invalid connections but it was in the wrong place. Everything is working properly after relocating it.

I've got an issue with the firewall leaking private IP addresses to the Internet. Quite regularly I'm seeing packets leave the public interface without the source address being translated. The packets that typically make it through w/o NAT are RST or FIN packets, although not always. Two questions. ...

I've run into a problem with the way switch chip implements VLAN tagging and I am searching for a work around. I expected that packets assigned a VLAN tag via '/interface ethernet switch rule' would be delivered to the appropriate VLAN interface on the RB. Unfortunately, this is not the case. Tagged...

if somebody can give us access to a client whos ISP supports these features, we can try to add them

I'm with Teksavvy, and they provide MLPPP service. I would love to test and get support for this. I'll email support.

To echo freethought, using external ADSL modems won't work with ML-PPP, so you can't bond multiple ADSL channels to improve bandwidth. You'll want to check that claim. Freebsd, Windows and Linux PPP implementations work out of the box with mlppp using external dsl modems. I for one would love to se...

Shouldn't the parent of pcq-dst be global-out, or more suitably, the interface the packets are leaving?

See http://www.mikrotik.com/testdocs/ros/2. ... flow31.jpg

What about historical data? Anyway to back that up?

Ive done some more testing and this looks more like a problem in the GUI. The state turns Red when 'rate > 3/4 * maxlimit' The functionality seems to work based on what the documentation says.

Hello, I believe I have found a problem with the Mikrotik bandwidth control. When the 'average rate' goes over max-limit / 2, the state turns Yellow. However the docs say the class should go yellow when actual rate goes over limit-at, and is less than max-limit. Is this a bug or documentation proble...

This works for me. I see you have not set the default route for one of the connections, what routes are added by automatically after you connect?

http://forum.openwrt.org/viewtopic.php?pid=61468#p61468

I have a work around until this bug is fixed properly. You can add this line to /etc/snmp/snmpd.conf:

view all excluded .1.3.6.1.2.1.25.2.3.1.7

/Chris

I've narrowed down the OID causing snmpd to crash. It is .1.3.6.1.2.1.25.2.3.1.7.2. Now to figure out why and a work around....

Hello Everyone, I'm trying to fix a problem I'm running into with the snmp agent used on openwrt. It seems, using the dude is causing the agent to hang. The two will seem to work for awhile until something triggers snmpd to use 98% cpu, and stop responding to snmp queries until the agent is restarte...

Perhaps a feature request is in order? I was hoping that adding the IP address a second time with a timeout of 1ms would fix the problem.

Unless you're doing MLPPP, it is rather simple. I'm crossing my fingers for MLPPP support soon.

Is it possible for a firewall rule to trigger removing an IP from an address list?

/Chris

Just a note, I'm using RouterOS v3.0rc13 on x86. With the exception of the scanning getting hung, the other three bugs were present in The Dude 3.0beta7

Hey Guys, First off, great product. Just wondering if someone can confirm if these are indeed bugs or just my setup. 1. PPP interfaces that reconnect with a new end point gateway are not discovered. In my case, The dude reports the old gateway as down but shows the PPP link still up. 2. Similar to n...

Here's how i got streaming working with mikrotik and wireshark: Mikrotik IP: 192.168.77.1 Workstation IP: 192.168.77.11 Fake IP address: 192.168.77.12 The stream is sent to a udp port on your workstation... but nothing is listening to that port... wireshark listens to the interface as a whole. What ...

I'm setting up 3.0rc11... I haven't been able to match 100% of unencrypted bittorrent traffic, never mind encrypted traffic.