CRS212 loses 25% of all RX traffic. The TX traffic works perfectly.
I have a CCR-1016 and RB2011, and put in the middle a CRS212.
Without the CRS212 everything works perfectly, no packet loss between CCR1016 and RB2011, but with CRS212, traffic is a mess, and RB2011 not recive a lot of packets.
I tried several CRS212 units with firmware 6.24 and 6.27, even using the default configuration (the only modification was SFP1 as master and auto-negotiation disabled on the SFP modules). I also tried with several modules SFP (all Mikrotik) and others optic fiber cables, and I’m sure that CRS212 is the problem.
A simple test bandwith between CRS212 and my core CCR-1016 (10MB RX/10MB TX), loses 25% of RX!, and customers have many problems to use of internet. Bandwith test between CCR-1016 an the RB2011 works fine.
I bought many CRS212 units and I urgently need to use it.
What is happening??
Make sure you are using CRS212 as switch.
Ethernet ports configuration has one masterport configured. For switch configuration you should get wire speed.
If you are using other configuration besides switch (like routing/VPN/firewall). Perhaps you need to change default queue on used interfaces, by default at /queue type only-hardware-queue is configured, set it to ethernet-default.
Yes, I’m sure CRS212 is working as switch. Only sfp1 is master, and all others are slave of sfp1.
No queues, no firewall or bridge rules, no high cpu, no routes
I tried default config, with only these changes:
SFP1 is the master port
Auto negotiation in SFP module is OFF (set to 1gb)
I have seen further that:
Some SFP interface is marked as R some moments (with no cable connected…)
Ethernet SFP tab in window show and delete constantly information about SFP module
I always lost 25% of packets… if 10MB, then I lost 2.5Mb, if 100, then I lost 25MB
Check all the links between the CRS and other equipment. This sounds a lot like there could be a duplex mismatch.
Make sure all ports are the correct speed/duplex on both the CRS and on the attached devices.
I check all links/duplex, and is correct.
If I put a RB2011 before, so, RB2011 has the main input optic cable and after connect a RJ45 from RB2011 to CRS212 master-port (with SFP RJ45 module) then all works fine with no packet loss.
Why if I put a RB2011 before and connect CRS2112 with RJ45 all works fine?
What is the difference between RB2011 and CRS212 with the same SFP module??
I try several SFP modules, and the same module that works in RB2011 fine, lose packets in CRS212
The config in CRS212 not affect, however is (SFP1 and SFP2 are uplinks):
/interface ethernet
set [ find default-name=ether1 ] name=ether1-master-local
set [ find default-name=sfp1 ] master-port=ether1-master-local name=sfp1-slave-local
set [ find default-name=sfp2 ] auto-negotiation=no master-port=ether1-master-local name=sfp2-slave-local
set [ find default-name=sfp3 ] auto-negotiation=no master-port=ether1-master-local name=sfp3-slave-local
set [ find default-name=sfp4 ] auto-negotiation=no master-port=ether1-master-local name=sfp4-slave-local
set [ find default-name=sfp5 ] auto-negotiation=no master-port=ether1-master-local name=sfp5-slave-local
set [ find default-name=sfp6 ] auto-negotiation=no master-port=ether1-master-local name=sfp6-slave-local
set [ find default-name=sfp7 ] auto-negotiation=no master-port=ether1-master-local name=sfp7-slave-local
set [ find default-name=sfp8 ] auto-negotiation=no master-port=ether1-master-local name=sfp8-slave-local
set [ find default-name=sfp9 ] auto-negotiation=no master-port=ether1-master-local name=sfp9-slave-local
set [ find default-name=sfp10 ] auto-negotiation=no master-port=ether1-master-local name=sfp10-slave-local
set [ find default-name=sfpplus1 ] auto-negotiation=no master-port=ether1-master-local name=sfpplus1-slave-local
/interface ethernet switch port
set 0 isolation-leakage-profile-override=0
set 1 isolation-leakage-profile-override=0
set 2 isolation-leakage-profile-override=1
set 3 isolation-leakage-profile-override=1
set 4 isolation-leakage-profile-override=1
set 5 isolation-leakage-profile-override=1
set 6 isolation-leakage-profile-override=1
set 7 isolation-leakage-profile-override=1
set 8 isolation-leakage-profile-override=1
set 9 isolation-leakage-profile-override=1
set 10 isolation-leakage-profile-override=1
set 11 isolation-leakage-profile-override=1
/interface ethernet switch port-isolation
add port-profile=1 ports=sfp1-slave-local,sfp2-slave-local type=dst
/ip address
add address=172.16.0.32/24 interface=ether1-master-local network=172.16.0.0
/ip dns
set allow-remote-requests=yes servers=172.16.0.1
/ip route
add distance=1 gateway=172.16.0.1
Hello, I have exactly the same problem with the CRS 125. Ethernet1 is as uplink routing eth2-20 other ports as HW switch. Approximately 60MB starts from packet loss.
RB2011 worked without a problem, because I had run out of ports replaced with CRS and there was this problem, which from the outset, ever since version 6.12?
Config I sent to the slides on the table and repair. Now try RC19 but it looks just as bad is version RC20 any change that would like to try?
The CRS network have two switches and both have the same problem, thus avoiding problems in the cables and the like. Both Sitch only routed traffic, have no firewall rule or queue, conection tracking is turned off.
Hello, solves this problem from MikroTik anyone? After a while a new version 6.28 but nowhere chabgelogu I do not see anything that would fix for CRS, can somebody please finally do something about it or should I have to buy another HW?
Thank you very much.
Quitting already have patience, communication with the technical support is close to zero. First, I advise to load the latest version of RC, it did not help, work to change the default queue on ethernet, also nothing and then change the value pfifo 500 also nothing. I bit her and queue off completely, but nothing about the problem somewhere else, right?
Ccr1036 with bgp and 10g upstream → crs212 via copper sfp+
Crs212 has 8 1gig fiber customers running on it and every week or 2 it has an issue sending customers traffic and it needs a reboot, config is just 10 vlan tagged on sfp+ and 1vlan tagged on each 1gig port
No routing, no bridges, just passing vlans. Tried 3 crs and also upgraded firmware with same result.
) using 2 PC without problem or packet loss.