unstable LAN

RouterOS General
1

We use mikrotik equipment for years ( all the time updated to current stable version ) and before COVID-19 crises some situations like that we attributed to poor connection to ISP.
Before crises all of us worked directly in firm but COVID-19 put us in situation working from home.
When we have started to use VPN (l2tp with ipsec) connection and RDP at the begging everything went smoothly, but after few days some of user started to complain on freezing RDP and lost connection.
What have I found?
VPN connection never broke!
All of us always stayed connected with VPN on mikrotik but RDP became very problematic.
If I have rebooted mikrotik everything went smoothly for few days again.
So, I started to investigate local LAN and found weird things.
This devices with problems started to have a lot loosen pings ( tool/ping resault more than 30% loose ).

After that I start tool/netwatch and got that :

https://1drv.ms/u/s!AjA-Ktm4yR9FkJ9GJx7NLSrJIZbGhw?e=lXJYMY

Only reboot of mikrotik or disable/enable port on mikrotik helped to work normally.
When ping is stable everything going like charm.
On one port we have connection trough another switch but few devices ( PC, laptop, Synology NAS… ) is directly connect to mikrotik.
Does anyone know the reasons why a directly or trough the switch connected device to mikrotik port on RouterBOARD 3011UiAS would have lost ping like that?
All cable is original (not a homemade) and I switched them for new one.

2

In the logs of the router, do you see any “link down” “link up” messages from ethernet ports lately ?
Since the latest update to 6.47 stable, my RB3011 device is seeing certain ethernet ports flip / flop very regularly resulting in a missed “ping” here and there.
At the moment I don’t seem to “suffer” at the application level, but …
I’m not sure which hardware models might be affected, you did not post any info on the model of your device.

3

I wrote model in post …but screenshot is here:

About log … nop…everything is normal … i just have few link down/up when I disable/enable …

4

Hmm, what a coincidence you are also running RB3011 …
But even more strange you cannot find any such events happening (link down/up) in the logs. So perhaps it is something else…
When did you upgrade ? Was there also a problem before the upgrade?

5

When you have ping issue peel in Tools>Profile> CPU Total (start) and chek that CPU load usage is ok, maybe it’s overloaded in some moments.

6



I have automatic upgrade on stable version … so, in 24-48 hours when new stable version become online… upgrade go on …
try to make similar test on your RB3011…

/system logging action
add disk-file-count=5 disk-file-name=NetWatch name=NetWatch target=disk

/system logging
add action=NetWatch prefix=netwatch topics=warning

/tool netwatch
add comment=Google down-script=":log warning \"Internet is down\";" host=8.8.8.8 interval=1s timeout=100ms up-script=\
    ":log warning \"Internet is up\";"
add comment=device_xxx down-script=":log warning \"device_xxx is down\";\r\
    \n\r\
    \n" host=x.x.x.x interval=1s timeout=10ms up-script=":log warning \"device_xxx is up\";"

change host=x.x.x.x for some of your IP in your LAN and put few more like that, for test purpose …

I notice a problem at beginning of February …

7

thanks for your tip …but this is something what I check in first second …

you can see here…
https://1drv.ms/v/s!AjA-Ktm4yR9FkJ9QaJtdJMt8r6QY9g?e=e9rDcD

are you still suspicion about cpu ?
I forgot to tell that I make enable/disable ports before 2-3 hour …so now everything goes very smoothly …

8

unfortunately…
:frowning:
it’s not just a coincidence, it’s the rule …
I produce same behavior on two places more …
one is RB3011 like first one …and another is on:

:frowning: