Finally MokerLink 8x 2.5GbE 1x 10GbE Switch Review

22

MokerLink 2G08110GS Performance

In terms of the overall 2.5GbE performance, we were able to get fairly normal 2.5GbE speeds from each of the eight ports.

MokerLink 8x 2.5GbE 1x 10GbE Performance
MokerLink 8x 2.5GbE 1x 10GbE Performance

Performance was good, in the range of the other switches we reviewed looking at the average transfer rates when loaded. So far, we have been within 3% or so on all of the switches in this class so it does not feel like performance should be the key buying criterion in this price range.

One item we were able to test was we hooked up the A Quick Look at the Asustor Flashstor 6 FS6706T NAS to this unit and run SMB3 multichannel from the NAS over the switch and to a 10GbE client using a SFP+ to 10Gbase-T adapter. It was awesome to see that this worked so easily and is a good type of use case for a switch like this.

MokerLink 2G08110GS and Management Capabilities

Since we are going to be doing this as part of a series, we will note that this is an unmanaged switch. Also, we did not find PoE capabilities when testing this switch.

MokerLink 2G08110GS Power Consumption

Here is a quick shot of the 12V 2A power adapter. This is not a huge adapter by any means. These switches have either the English or Chinese writing on them. A fun bit is that this one is not a power adapter but instead a “ADAPTER POWER” due to the translation.

Mokerlink 8x 2.5GbE 1x 10GbE Switch 2G08110GS 2a 12V Adapter Power
Mokerlink 8x 2.5GbE 1x 10GbE Switch 2G08110GS 2a 12V Adapter Power

At idle, we saw 1.6W which is lower than many of the 3.8W-4.4W switches we have reviewed previously in the 5-8 port range.

Mokerlink 8x 2.5GbE 1x 10GbE Switch 2G08110GS 1.6W Idle
Mokerlink 8x 2.5GbE 1x 10GbE Switch 2G08110GS 1.6W Idle

Plugging in a port at 2.5GbE speeds we saw an incremental 0.7W of power consumption. That is in a more normal range.

Mokerlink 8x 2.5GbE 1x 10GbE Switch 2G08110GS 2.3W With One 2.5GbE Port
Mokerlink 8x 2.5GbE 1x 10GbE Switch 2G08110GS 2.3W With One 2.5GbE Port

Plugging in a 10Gbase-T to SFP+ adapter, that immediately starts to suck power, we got a paltry number.

Mokerlink 8x 2.5GbE 1x 10GbE Switch 10Gbase T Adapter Power 1
Mokerlink 8x 2.5GbE 1x 10GbE Switch 10Gbase T Adapter Power 1

This switch is the lowest power 2.5GbE option, and not just for 30Gbps of total networking switches, even down to the 12.5Gbps 5-port 2.5GbE options.

Final Words

To be frank, these switches are perhaps the best value for low-cost unmanaged fanless switches that have 2.5GbE and 10GbE ports. Of the eight units we have, across four brands, we have a few insights. First, the Horaco one from AliExpress saw a single instance where plugging in a SFP+ DAC reset the switch. We have not been able to reproduce it since, on this MokerLink or the other switches. Still, since they are the same architecture, we wanted to point it out. These are not perfect by any means, they are inexpensive and low power. We also wanted to note that MokerLink actually sent us one of the review units we have so there is a company behind MokerLink that has at least figured that out.

Mokerlink 8x 2.5GbE 1x 10GbE Switch 2G08110GS Front
Mokerlink 8x 2.5GbE 1x 10GbE Switch 2G08110GS Front

With that said, one of the biggest challenges is recommending one over the other. Of the three models we purchased on Amazon.com, the pricing was $119-129 or so depending on the discount, and some with free same-day Prime shipping. The Horaco from AliExpress took a long time to arrive (over 2.5 weeks) but was less expensive by ~$30.

If you want to learn more about this versus the other rebrands, feel free to take a look at the video while the reviews go online over the next few weeks.

What is clear, however, is that this offers something different using less power and at a relatively reasonable price.

Where to Buy

We purchased our unit on Amazon. Here is the affiliate link for this model.

Ultimate Fanless 2.5GbE Switch Roundup

You may have seen that we published the Ultimate Cheap Fanless 2.5GbE Switch Mega Round-Up. Here is the video for that one where this switch appeared.

We decided to do the round-up first and will be filling in with reviews that we had not published at that point, including this one. We are trying to keep everything in a single resource for folks there.

22 COMMENTS

  1. Are we ever going to get a 2.5 switch that I can set vlans and LACP? Even a simple web interface works.

  2. > we were able to get fairly normal 2.5GbE speeds from each of the five ports.

    Eight not five. :-)

  3. I don’t understand how one tests the speed of the 10GB port when there is only one of them. Where does it send the data? Don’t speed tests require a send and receive?

  4. @Eric Olson:
    > I don’t understand how one tests the speed of the 10GB port when there is only one of them.

    It’s an upstream port, so that you don’t have a 2.5G bottleneck when using more than one port at max. speed.

  5. @Eric Olson: You send multiple streams of data in parallel. One stream on each 2.5G port means you only need four streams across four ports to get 10G of bandwidth all up. Likewise if you send four streams in on the 10G port, with each stream destined to a device on a different port, the switch will split the data across multiple 2.5G ports.

    In this case there would be eight 2.5G ports transferring 20 Gbps of data, so easily enough to saturate the single 10G link to find out its true top speed.

  6. Out of curiosity(and since you guys seem like good people to ask, given that this very article rockets to the top of the search results, even vs. vendor part pages or product family marketing materials, if I try to chase either realtek part number); how much diversity is the maker of a switch like this even given to choose from in terms of suppliers of switching silicon?

    Is this fast enough(albeit still relatively slow and unmanaged) that more or less everyone associated with switching silicon has some offering? Is it a literal handful of models from realtek and maybe something from Marvell?

    From reading this review series the board shots show some different layouts, which leads me to believe that we aren’t seeing literally the same reference design shoved into different boxes with more or less penny pinching on power capacitors and ethernet magnetics; but the fact that all the switch chips are just toasty enough to require a heatsink makes it hard for me to tell how much difference lies beneath: Are these within ~3% of one another because there are a bunch of options but unmanaged switching at this speed is a mature and solved problem so there just aren’t any gotchas; or do they perform largely the same because they are mostly the same switch chips with slight variations depending on the production values of the supporting circuitry?

  7. @fuzzyfuzzyfungus
    The switch in this review, as well as the Horaco, the Nicgifa, and the sodola are all using the same switch chip inside them. They are all 8+1 unmanaged switches.

  8. “This might be the biggest thing to come out of Wuhan in years.”

    This switch might not be hot, but this writer sure is.

  9. The English here – or at least within the first few paragraphs (I read no further) – is, sad to say, abysmal.

  10. Of these cheap 2.5gbe switches, this is the first to actually catch my attention. Many homelabbers likely have a free sfp+ port, so switches like these are a cheap way to have uncorked 2.5gb ports for cheap

  11. How does this switch handle lower link speeds? Namely plugging a 1Gbps device to one of the 2.5Gbps ports and saturating the 1Gbps link from the uplink port?

    I have a bad experience with cheap 1Gbps switches (namely Cisco SG100D-08 V2) where a 100Mbps device would make the switch unusable when the 100Mbps device downloads data at speeds near 100 Mbps from a 1Gbps device in another port.

    I assume these issues are buffer related and switches with larger buffers wouldn’t be affected by this?

  12. Got the Mokerlink version just last week here in Germany. 139,- Euro (incl. VAT/Salestax +Shipping). From FS, i have a SFP+ RJ45 Nbase-T transceiver for 30m (suitable for Netgear), works flawlessy at 2.5G to desktops and 10G to an Intel X550. Did not test it at 5G (but have no need anyway for that).
    I also checked with a DAC-Cable on a XS716T’s SFP+ Port, also working at 10G as it should.

    Would love to see 2x 10G instead 1x in addition to the 8x 2.5G, but hey: cooling is passive, power draw very low and it’s good value for money

    Nice little basic switch. Replaces an old simple DGS-1008D here.

  13. does anyone know if the 10G port supports 1G and/or 2.5G SFP modules, would it autodetect the speed?
    Also would it support the Magic SFP moca 2.5 module?

  14. Any suggestions on SFP+ cabling to connect this to a netgear managed switch (L2+) 10G SFP+ port? (short run <2M)

  15. Has anyone had any luck using an optical module on the 10G SFP+ port? I have two of these switches, one works great with a direct attach cable but the other one I can’t seem to get to work despite trying 2 different brands of optical module (one of them was a Brocade.) Just wondering if it might be picky about what brands it will work with.

  16. Hello StH / Patrick. Please examine the newest (it’s user manual shows v1.0 from April 2023) Mokerlik 10G120GSM which is a 12 port SFP+ (those also support SFP) and says it can handle 240Gbps bandwidth, and it’s an L3 managed fiber switch… sells for about $270 USD on AMz currently.

    I’ve not seen ANY reviews for it, but on paper it looks awesome for a homelab. I’m considering it over the Mikrotik CRS309-1G-8S+IN (8 SFP+) managed switch.

    THANKS BRO!

  17. I came across information, firmware 1.9 for Xike Stor branded switch actually supports LACP. Anyone knows where to get this firmware version or the one from other brands?

  18. @StH / @Patrick,

    Managed versions of these are now available, and for very low prices, all things considered. For example, I got the Horaco version on Aliexpress for about $61. At least for the Mokerlink version, they just put an “M” on the end of the model number

LEAVE A REPLY

Please enter your comment!
Please enter your name here

This site uses Akismet to reduce spam. Learn how your comment data is processed.