Important: this behavior is not officially supported, even in RouterOS v7.
Both devices are running the current version of OS’s. My plan is to use the sfp-sfpplus1 interface on the RB4011 as a trunk interface to the switch for implementing VLANS. I am not seeing any errors on the swOS statics page.
I haven’t started the VLAN exercise yet. I’m just working the hardware to get familiar. I’m using the RB206SG switch in the lab for testing. The goal is to connect the RB4011 to a Mikrotik CSS326 using the SFP interfaces with the DAC cable.
Excuse me, but I don't really understand your doubts, you already have the DAC, Mikrotik officially lists it as compatible with BOTH the devices you have, you have it working.
At this point you ask ChatGPT?
Try making an iperf test and check what throughput you have, if it works, it works.
Go on, if - for any reasons - it won't work in your setup, THEN it will be time to find an alternative.
Thanks for the input. I had reservations mentioning ChatGPT’s output. Foraging for MikroTik devices to support a small business was a path I chose against the advice of some network professionals I know. When trusted sources are waiting in the wings with an “I told you so” response, ChatGPT can be one resource to help address confirmation bias.
Getting it to work and discovering later it was not the ideal path is something experience has taught me to avoid. My use of MikroTik has been a journey of discovery that has exceeded expectations. RB206GS is the fourth MikroTik device in a series of network devices I have purchased. They may be good, but I recognize I am not.
A few posts in the forum described challenges using SFP28 1M DAC. Not with this specific equipment, but challenges non-the-less. In a past life there was a team of people that supported networking. I relied on their knowledge to further the company's network. That safety net is gone.
I have successfully challenged ChatGPT in the past to change its output. I failed this time using the resources I had. I’m long in the tooth and have never liked the try until you get it working approach and assume it is right. It is so much easier to do things on paper first before building on an idea.
I used the compatibility chart to purchase the DAC. It didn’t work out of the box; I had to discover the command to get packets flowing. I posted to the forum to shore up my confidence before implementing VLANs with this interface as the trunk.
I wasn’t aware of the iperf test. It rings a bell (it may not be the right bell) from my early days at Sun Microsystems operating system. I will check it out and add it to my bag of tricks.
The MikroTik quick start guides refer to https://192.168.88.1 as the browser link. https has never worked for me. The landscape that supports the (s) has transformed in the past five years and looks to change even more. Did the secured link function correctly previously?
SFP sockets (and the various devices you can plug in them) are actually a PITA, they are essentially "dumber" than good ol' plain ethernet where auto-negotiation always worked, so the progress is that now, in most cases, you need to set/force a given speed.
The RB4011 has a SFP+ cage (and can go up to 10 Gb), the RB260GS has a SFP (without plus) cage, so it maxes out at 1Gb and represents the bottleneck.
The good thing is that, depending on the traffic/speed you expect to go through the link, you can later upgrade the RB260GS to a switch with a SFP+ cage (CSS610?) keeping your DAC (that will work much faster).
Running an iperf test (possibly between two devices connected to the router and to the switch (running iperf tests "internally" between the two Mikrotik devices may be misleading due to the CPU usage) is only a way to make sure that you actually get the 1 Gb speed you expect.
The naysayers will anyway find a way to criticize your choices, let them be.
I recall early in my network days or was it early in Cisco networking equipment auto-negotiation was something only the new kids on the block turned on. Nail-up-the-speed was the moto of senior guys. Buried deep in the bowls of an ISP I worked for is a device that has auto off for all ports. With such a mix of legacy and new gear it was a constant battle to connect network elements.
I will expose my naivety by inquiring about the term cage. It sound more than my hole in the box to plug an SFP into idea. I have increased my knowledge about DAC when a senior took the time to respond. Hopefully I don’t take away from my next ask.
Can you point me to the iperf testing you are familiar with? I’m hunting and pecking and haven’t found a reliable source.
If you like to call it "port" or "connector" instead of "cage" you are welcome, but the thingy (the component) is commonly called "cage": https://no.rs-online.com/web/p/pluggable-io-connector-accessories/1676250
I personally prefer calling it cage to distinguish it from (ethernet, USB and serial) ports, the difference is that besides the cable or fibre link, until the cage is populated (by a module) it does pretty much nothing.
And when you buy for 150 bucks (say) a CRS305 as a zoo you still need more than that for the beasts needed to populate its four cages.
I tested it and it works. The XS-DA0001 DAC is listed as compatible by MikroTik for both devices, the link comes up at 1G, and I can pass traffic without errors. If iperf shows stable throughput, I would just use it and move on with the VLAN setup. If problems appear later, then I would switch cables.
My formal educational background is electronics. I moved into system administration when the career path was in it’s infancy. I never actually realized the transition until resolving a SCSI bus termination issue. Hey this bus is 100MHz RF transmission path. I know how to terminate to reduce VSWR and reflections.
Your cage picture is BANG ON!
I have to manage my time to invest in iperf and keep the VLAN implementation on schedule.
The DAC seemed pretty straight forward. High speed copper whats not to work. Chatgpt is still convinced it is not supported no matter how much I have provided.
During last night IIHF junior hockey game I dabbled in iperf3. I didn’t realize (never looked) how simple it is to install and use. I can acknowledge why my cry for documentation seemed trivial.
The test on the new DAC failed the 1G test. I discovered a 100M bottle-neck in my home network because of an old switch. Replacing the switch solved that issue.
Actually - like many Linux tools - iperf has countless possible options, It Is the basic use as a transfer speed checker that Is simple.
Learning the available options for more advanced uses Is far from trivial.
you still need more than that for the beasts needed to populate its four cages.