Question is really simple…it’s a fact that other vendors are looking more in the 2.5G since more and more domestic equiment now support it, even new minipc comes with 2.5G eth or most of the Access Point for WiFI 6E or 7
So…what about a device with
4 X 2.5G
2 X 10G SFP+
1 X 10G Base-T (eth, or even two of them)
First 5009 release was back in the 2021, it looks like it’s a great moment for the successor, otherwise I do believe most of us will be forced to go to other vendors 
It’s a fact that other vendors (Ubiquity for eg) is ahead of Mikrotik, especially in wireless segment. (While Mikrotik is stuck at WiFi6, other vendors have several WiFi7 APs in their lineup)
For 300 bucks or so UCG Fiber is a great device, they have poe on their LAN ports, fiber, 10G eth, ssd for cameras. IDS/IPS routing at 5Gbps i think.
With the biggest packet size and it doesn’t feature TLS Inspection 
The price is competitive, but the features as a router are lackluster. ^^
Just SFP for LAN.
Tbh the device will be more interesting once the unofficial block diagrams drop.
But I got a bit of a feeling that it’s just a repackaged UDM with the same issues again and again
But a new and improved RB5009 would be great
Well, one problem with Ubiquiti i think is, they release product no matter what even if software side is not ready for release.
Their U7 Pro AP had a problem with 2.4 GHz radio i think. Don’t know of they fixed that.
I’m fine with the current port selection on RB5009, but I wouldn’t mind all 2.5Gbps and an extra 10Gbps or SFP+.
What I would really love from mikrotik is HW offloaded PPPoE.
Where I live, PPPoE is sadly still popular. Currently I only have 1Gbps, and have been waiting on my ISP for over a year now to expand their 10Gbps offering to my street, and while it doesn’t looks like that’s coming very soon, I’m probably gonna switch to the UCG fiber when that happens as that has hw offloaded PPPoE.
Also Banana Pi BPI-R4 Pro Router board jump n the competition for a more “appropriate” number and types of port speeds.
Any news from Tik? Really looking forward for a successor of my RB5009, I don’t want to move away from routeros, but the port gap with competition is now relevant.
You could add a CRS310-8G+2S+IN to your setup, link it to the RB5009 at 10Gbps, and have 8 more 2.5Gbps and an extra SFP+ port.
The CRS310-8G+2S+IN uses 13W doing nothing, with no links up.
when you plug something, goes up to 18W. Also the FAN is so noisy.
The Gateway Fiber uses ~20W fully loaded if PoE out is not used.
an new RB5009-like with faster ports (maybe less ports, but faster) would be well received.
my RB5009 sometimes struggles to route/NAT PPPoE 8Gbps WAN (no CPU is fully loaded), on IPv6 it normally gets ~7.6Gbps
I recently ordered a BPI-R4 to test since I’m too impatient to wait for the upcoming Pro and also since as far as I can tell, the only difference between the Pro and non-Pro version is the additional 10GBASE-T and 2.5GBASE-T ports. Otherwise it’s got the same CPU and all I care about is the two SFP+ cages anyway. I’ve seen claims that the BPI-R4 can route full 10 Gbps (~9.4 Gbps after overhead) but I’m assuming that’s only in one direction at a time. I’ll be shocked if I’m actually able to push 10 Gbps in and out of it simultaneously, and that’s really what I’m after for my use case.
The only thing mikrotik needs is HW offloaded PPPoE (ok, maybe not the only thing, but this would be a big one). Or at least multi-threaded PPPoE implementation.
There are many of us that still have to authenticate through PPPoE, but sadly it’s still single threaded in routeros.
I just bit the bullet and bought a RB5009Upr. I needed the SFP(+) port to replace my ISP ONT and one PoE out to power an AP (U7-Lite), initially I wanted to buy the new hex S 2025, but I cancelled the idea once I saw many people complaining about the port speed issues. Also dual-core did not really attract me, it would be a downgrade from my current AC2.
Currently running a 1Gbps connection, supposedly until the end of the year I’ll be able to get 10Gbps (it is available a couple of blocks away from me), but wouldn’t surprise me if it takes my ISP more to extend the infrastructure on my street. The RB5009 should be enough for my current connection even with QoS, advanced firewall. I’m not sure that even with fasttrack it would be able to saturate the 10Gbps, so I just hope that Mikrotik comes with some solution for PPPoE until next year, I really don’t want to switch to Ubiquiti.
So please mikrotik, do something with PPPoE. Multi-threaded PPPoE would be nice, as the RB5009 I just bought would also benefit from that. Or at least future devices with HW offloaded PPPoE.
+1 for an RB5009 successor with more 2.5Gbps ports. The UCG Fibre is a serious competitor.
Current setup is an RB5009UPr+S+IN with a wAP AX via PoE. Also have a CSS318-16G-2S+ on port 8.
Thanks to fibre networks in the UK mandating the use of a Ethernet connection to an ONT device (I understand that networks in Europe provide or allow the use of an SFP module to connect your fibre?) I’ve inadvertently designed my first MikroTik network with a fine bottleneck.
The 10G SFP+ module isn’t ideal for a passively cooled device and also WiFi 7 AP will be bottlenecked by a 1G port. Ideally this router needed a 2nd or more 2.5G ports for future expansion.
Thankfully I’ve not managed to stretch my 900Mbps connection. Yet.
yes mikrotik please get a successor to rb5009 for 10gbps pppoe capability and lesser ports ,but only 2.5 and sfp+
Even MT made a successor of RB5009 as long as it use the same more or less chipset it can’t give you a 10G performance as far as the PPPoE is concern, because it is single threaded by nature without drastically rewrite PPPoE I think this will be forever, most of the ISP that we know offering more than 2GB of bandwidth use DHCP/IPoE (no PPPoe overhead)
The processor used in the RB5009 has a versatile packet parser that can probably be programmed to handle PPPoE “in hardware”, but there are several other processors used in MikroTik equipment that can do that.
It seems RouterOS does not yet take advantage of that, maybe at some time the PPPoE can be offloaded just like some devices can offload the routing to a switch chip.
“cheap white boxes” offered by Fiber ISPs often do use that offloading, because they run the software provided by the chip manufacturer (much less versatile than RouterOS, but far better performance).
Well If there were such feature/capability in the hardware what’s stop them from implementing it in their software stack that important features is non optional from a start, there were lot of missing opportunities from Tiks because the ship has sailed, anyway there’s nothing we can do from here aside from being surprise when it’s done, this situation is can be compared to the elusive VTI features that a lot of folks here including me is waiting to be realized 
hahahaha
RouterOS has started out as a plain Linux running on i386 PC boards and gradually has been extended and migrated towards dedicated router chipsets. There is only so much you can do with a limited team of programmers, and priorities will always have to be set.
I have also asked in the past why there is no VDSL support anywhere in the product line, not even support for other manufacturer’s VDSL SFP, but the reply was always along the lines of “VDSL? we do not have that here! we all have fiber! why bother with VDSL?”, ignoring the fact that it is quite popular in other countries where technilogical development started before it did in the Baltics, and that therefore have legacy infrastructure that is not immediately replaced. I still do not have fiber at home and we also do not have it at one of our work locations.
Similarly, the use of PPPoE which is natural here (both over VDSL and over fiber) due to the situation of having multiple ISPs operating over a smaller number of infrastructure networks (VDSL by the incumbent telecom and fiber by them and a couple of new fiber companies) may well be not happening there. “why do you want PPPoE? who uses that? ask your ISP to stop using it!”.
While the ISPs themselves deploy routers from other companies like AVM, ZTE etc, which have similar chipsets as MikroTik have, but are running PPPoE at very high speeds (like 2.5Gbps) without a sweat.
We know that MT consist of small development team and it has been discussed for quite some time now, what I don’t get is why they can’t expand it to the point the all of these backlogs in development can be fulfilled?.
If funding is the issue they have a lot of means to gather funds e.g by offering a paid commercial support with contracts with companies or vars (value added resellers) they can be creative the customers can understand this and I’ve seen a few times many people ask including me
Leverage their partnership with the whole supply chain to help them bring to speed all those development task they need and discussed their roadmaps this is not their first rodeo so they can do it for sure
Outsource some of their development task that’s beyond or outside of their specialties and pay them in cash, stocks and royalties or whatever means to compensate this partner
In my humble opinion, MikroTik operates more like a boutique workshop than a full-scale tech enterprise. They build reliable and clever products, but seem reluctant to grow into a larger ecosystem where more innovation, support, and services could be delivered by scaling their team and resources.
How I wish they broaden their horizon, they have so much potential specially here in Asia they just need to tweak their goal.
I guess the showstopper is the same as in “why can’t they implement bridge L2HW offload for devices with Qualcomm switch chips (AR8xxx and QCA8337)” … they lack developers at the time of introduction of something new (either new device model or new ROS feature) and they don’t do it at later stage (and this is probably s business decission). And in case of some new feature (like PPPoE offload to hardware) likely hooks in kernel/software stack/UI are missing as well, which means more work to support first devices …
They have some open vacancies. Surely they would hire a super skilled senior C developer to work on their backlog. But maybe none exist in Riga.
Yes on top of that even a seasoned developer from other vendor(s) wants to work with them it will be very hard because part of the requirement is the developer must speak their language and that is massive undertaking on itself, anyway it is what is