Today afternoon, we started to experience network issues: bad voip phone quality, latency and disconnections while using winbox to router.
Trying to diagnose I discovered a severe packet loss (30%) on ping from outside to router and inside network but also from inside network to the router. In the morning everything was ok, no configuration changes after.
Disabling hardware offloading on a bridge port solved the issue.
One of our lan switch (ubiquiti usw-pro-48-poe) is connected to the sfp+ ports of the router (Mikrotik CCR2004-16g-2s+) by two dac cable in bond/vlan trunk configuration. The router is connected to an “outside” switch by eth1+eth2 in bond/vlan trunk configuration. This outside switch is connected to few AP for wireless link also to our ISP. The bridge port with hw offload on was the eth1+eth2 bond.
Not understand if it is normal that hw offloading can cause this issue or there is something wrong elsewhere.
Not understand why no issue befoore today noon: for sure I didn’t enabled hw offload at noon.
Should I leave hw offload disabled? Or should I change/check something else?
Thanks
Same switch, similar issues here. What FW version? I’m on 7.10. Thinking of reverting
Downgrade to 7.8 on it’s own made no difference. Disabling HW offload on all active bridge ports made the difference though.
7.10, but since few days, and the issue appeared only yesterday, and affect both bridge port, even the one without hw offload.
Did you downgrade Router OS or RouterBoard?
Downgraded ros to 7.8 with no change. I had to completely disable HW Offload on every bridge port to clear up the issues. May try upgrading to 7.10 again to see if it stays stable with offload disabled.
Hi
I have exactly the same issue! FW 7.13
This problem occurs when traffic is between the vlans and to external interface. Interesting thing is if you disable some of interfaces no matter it is included in a vlan or bridge or even active, this issue disappears for a while / between 1-5 minutes/ and then again. I do also tried with different uplink interfaces and speeds, as copper, dac 1G, 10g, SFP, 1g, 10g - problem is still there
In the same vlan problem doesn’t persist!
my config is: CCR2004-16G2S+ configured as CoreSwitch with several vlans on one bridge interface assigned to all eth interfaces /exept eth1,2 as they are for ISPs/. Uplink is 10G with DAC to FS S3900-48TS6 10G switch
I have tried different configurations for vlans on bridges or standalone even LACP for uplink on different interfaces without any success
On my opinion there is a batch with faulty CCR2004…
Finally I have replaced the CCR2004 with one HEX-S for a while and by the way HEX is working like a charm with the above configuration…
Do you have any new experience with this issue?
UPDATE!
In my case problem was with communication between Mikrotik and other Devices - FS 10G Switches
Problem was solved as I rised the priority of my Core Bridge interface to 0
interface/bridge/set BR-XXX priority=0