Hi to Mikrotik's staff, one question, when will be support / download for RB5009 available ?
On download page is not a singe word about rb5009....

Thx.

I just received my RB5009, and it comes with 7.0.5 RouterOS.. when it checks for upate, it says latest stable version is 6.48.4 !

if i am not mistaken,

The new Mikrotik Devices (like the RB5009) will only support RouterOS v7 and higher.
At the moment there is no Public version of a "v7 Long-Term" or "v7 Stable" available

If you buy a RB5009 Device it will come with a non-Public v7 Stable.
I assume as soon as V7 is ready, you will be able to Update / Downgrade to the actual Public Stable.

The preinstalled 7.0.5 can be upgraded only to a better v7, currently I suggest upgrading to 7.1rc4 which is a much better version in all aspects.
v6 can't be installed on RB5009, the update checker is mistaken

Sorry Normis, the update checker is not mistaken, it Mikrotik who is just sloppy.

I stated this before and it was ignored, as useal.
This I get when click on the RouterOS current release download button on the RB5009 page.

I just received my RB5009, and it comes with 7.0.5 RouterOS.. when it checks for upate, it says latest stable version is 6.48.4 !

where did you buy your device? i couldn't find it

Nothing ?
Any change ?

Public Stable Release for version 7 will be within the next 6 weeks ..... so patience is a virtue ....

In Winbox 3.31 -- for System/Packages/Channel/ ...... select upgrade and THEN you will see the latest available for your specific device

Is there any support for L3 HW offload on RB5009? And, especially, NAT?

Is there any support for L3 HW offload on RB5009? And, especially, NAT?

+1

the question is how this was left behind? also the RB5009 have no way to use it's switch with different vlans without completely loosing fasttrack support.

Just FYI, since today I am using redundant power supply.

First power is standard 24V power supply from box, which came with rb5009.
Second power supply is 12V power supply connected to connector from side.

So far, so good.
I am write this here cause I do not find info about redundant power supply options on rb5009 anywhere.

Product page spends quite a few bits describing the 3 ways of powering RB5009.

I just migrated to a new RB5009UG+S+ from an RB3011. I have encountered a download speed performance issue that is easy to duplicate. When the RB5009 ether1 WAN port is "advertising" 2.5G full duplex and is connected to a 2.5G ISP modem with gigabit speed service, the data transfer speed drops by about half to 300 to 400 Mb/s. If I remove advertising 2.5G and only advertise 1.0G full duplex, the transfer rate increases to 650 to 850 Mb/s. CPU load runs about 1 to 10% in all scenarios. Bypassing the RB5009 and connecting directly to the ISP modem yields a fairly consistent 850Mb/s. There are no other switches, routers, or devices in the chain. The ISP modem is a Motorola MB8611.

I am running firmware 7.1. Basically a factory reset configuration with a fast-track firewall setting.

One of the reasons I opted for the RB5009 is the 2.5G WAN port. I didn't want to use an SFP+ module until I had a much faster Internet service speed available in my area.

...can be powered in 3 different ways
It's not saying anything about redundancy.

...can be powered in 3 different ways
It's not saying anything about redundancy.

You're right, it doesn't. So one can assume (yeah, mother of all ...) it's the same way as with other mikrotiks with more than single powering source: it will draw power from source with highest voltage. If that one fails, it'll use the one with highest voltage of the remaining sources. If multiple sources provide same voltage, device will draw power from all of them. Selection of power source with higher voktage is nothing magical: all power inputs are wired together over a diode (per powering source) to block egress current/voltage.

When the RB5009 ether1 WAN port is "advertising" 2.5G full duplex and is connected to a 2.5G ISP modem with gigabit speed service, the data transfer speed drops by about half to 300 to 400 Mb/s.

Probably some interaction between bufering in the WAN device (which has to smooth bursts exceeding 1 Gbps). If flow control is disabled on RB5009, then WAN device can only drop some frames and dropping frames affects TCP throughput a lot. Try to enable flow control on WAN port to see if it helps.

Is there any support for L3 HW offload on RB5009? And, especially, NAT?

+1

the question is how this was left behind? also the RB5009 have no way to use it's switch with different vlans without completely loosing fasttrack support.

have you checked this with latest V7.1 ??

https://help.mikrotik.com/docs/display/ ... troduction

https://help.mikrotik.com/docs/display/ ... Offloading

When the RB5009 ether1 WAN port is "advertising" 2.5G full duplex and is connected to a 2.5G ISP modem with gigabit speed service, the data transfer speed drops by about half to 300 to 400 Mb/s.

Probably some interaction between bufering in the WAN device (which has to smooth bursts exceeding 1 Gbps). If flow control is disabled on RB5009, then WAN device can only drop some frames and dropping frames affects TCP throughput a lot. Try to enable flow control on WAN port to see if it helps.

Thank you for the suggestion. I gave it a try but it did not have any affect. I still have slow speeds when ether1 is set to 2.5G. I think I will order an SFP+ module to see if it can provide proper performance on the cage port.

Probably some interaction between bufering in the WAN device (which has to smooth bursts exceeding 1 Gbps). If flow control is disabled on RB5009, then WAN device can only drop some frames and dropping frames affects TCP throughput a lot. Try to enable flow control on WAN port to see if it helps.

Thank you for the suggestion. I gave it a try but it did not have any affect. I still have slow speeds when ether1 is set to 2.5G. I think I will order an SFP+ module to see if it can provide proper performance on the cage port.

be aware copper rj-45 10g module generate a lot of heat, rb5009 uses passive cooling, adding this module will increase overall temperatures

I just migrated to a new RB5009UG+S+ from an RB3011. I have encountered a download speed performance issue that is easy to duplicate. When the RB5009 ether1 WAN port is "advertising" 2.5G full duplex and is connected to a 2.5G ISP modem with gigabit speed service, the data transfer speed drops by about half to 300 to 400 Mb/s. If I remove advertising 2.5G and only advertise 1.0G full duplex, the transfer rate increases to 650 to 850 Mb/s. CPU load runs about 1 to 10% in all scenarios. Bypassing the RB5009 and connecting directly to the ISP modem yields a fairly consistent 850Mb/s. There are no other switches, routers, or devices in the chain. The ISP modem is a Motorola MB8611.

I am running firmware 7.1. Basically a factory reset configuration with a fast-track firewall setting.

One of the reasons I opted for the RB5009 is the 2.5G WAN port. I didn't want to use an SFP+ module until I had a much faster Internet service speed available in my area.

I have a very similar problem, but in my scenario, I have 4 VLAN's and I'm using 2.5G port as an access port for LAN.
If it advertise as 1G then speed is OK, but when advetising 2.5G, speed is below 800MBits when connected to other 2.5G nic (tested with QNAP PCI and usb 2.5G).

Have 7.1 version installed. Still no hardware offloding option.

I just migrated to a new RB5009UG+S+ from an RB3011. I have encountered a download speed performance issue that is easy to duplicate. When the RB5009 ether1 WAN port is "advertising" 2.5G full duplex and is connected to a 2.5G ISP modem with gigabit speed service, the data transfer speed drops by about half to 300 to 400 Mb/s. If I remove advertising 2.5G and only advertise 1.0G full duplex, the transfer rate increases to 650 to 850 Mb/s. CPU load runs about 1 to 10% in all scenarios. Bypassing the RB5009 and connecting directly to the ISP modem yields a fairly consistent 850Mb/s. There are no other switches, routers, or devices in the chain. The ISP modem is a Motorola MB8611.

I am running firmware 7.1. Basically a factory reset configuration with a fast-track firewall setting.

One of the reasons I opted for the RB5009 is the 2.5G WAN port. I didn't want to use an SFP+ module until I had a much faster Internet service speed available in my area.

+1, seeing very similar results, see below.



My service plan is 1GB/40Mb. Prior to discovering your post, I experimented with numerous other tweaks without much difference in results, from adding/removing fastrack to adjusting firewall rules. When running speed tests I noticed CPU was 60-80% with 2.5G advertising included, where as without I'm seeing 10-16%.

FYI, I recently migrated from Edgerouter POE5, where I was consistently getting 800-900+ Mb/s. Regardless of these results I still very much like the RB5009, it's got much more horsepower and performance feels snappier. Given this router is still really new, I'm optimistic that future fw will improve performance and make it even better/consistent experience. I'm hoping L3 offload is coming soon...

Probably some interaction between bufering in the WAN device (which has to smooth bursts exceeding 1 Gbps). If flow control is disabled on RB5009, then WAN device can only drop some frames and dropping frames affects TCP throughput a lot. Try to enable flow control on WAN port to see if it helps.

Thank you for the suggestion. I gave it a try but it did not have any affect. I still have slow speeds when ether1 is set to 2.5G. I think I will order an SFP+ module to see if it can provide proper performance on the cage port.

I'll be interested to see your results, I've got my popcorn ready...

Have 7.1 version installed. Still no hardware offloding option.

Welcome to the RB5009 unfinished product club.

I just migrated to a new RB5009UG+S+ from an RB3011. I have encountered a download speed performance issue that is easy to duplicate. When the RB5009 ether1 WAN port is "advertising" 2.5G full duplex and is connected to a 2.5G ISP modem with gigabit speed service, the data transfer speed drops by about half to 300 to 400 Mb/s. If I remove advertising 2.5G and only advertise 1.0G full duplex, the transfer rate increases to 650 to 850 Mb/s. CPU load runs about 1 to 10% in all scenarios. Bypassing the RB5009 and connecting directly to the ISP modem yields a fairly consistent 850Mb/s. There are no other switches, routers, or devices in the chain. The ISP modem is a Motorola MB8611.

I am running firmware 7.1. Basically a factory reset configuration with a fast-track firewall setting.

One of the reasons I opted for the RB5009 is the 2.5G WAN port. I didn't want to use an SFP+ module until I had a much faster Internet service speed available in my area.

+1, seeing very similar results, see below.



My service plan is 1GB/40Mb. Prior to discovering your post, I experimented with numerous other tweaks without much difference in results, from adding/removing fastrack to adjusting firewall rules. When running speed tests I noticed CPU was 60-80% with 2.5G advertising included, where as without I'm seeing 10-16%.

FYI, I recently migrated from Edgerouter POE5, where I was consistently getting 800-900+ Mb/s. Regardless of these results I still very much like the RB5009, it's got much more horsepower and performance feels snappier. Given this router is still really new, I'm optimistic that future fw will improve performance and make it even better/consistent experience. I'm hoping L3 offload is coming soon...

Try this:
Disable interface. Remove 2.5G full from Advertise options. Enable interface. It should work OK for 1G.
Now, disable it again and add 2.5G full in Advertise. Enable interface. It should negotiate 2.5G, and wonder if it make it faster.

Try this:
Disable interface. Remove 2.5G full from Advertise options. Enable interface. It should work OK for 1G.
Now, disable it again and add 2.5G full in Advertise. Enable interface. It should negotiate 2.5G, and wonder if it make it faster.

Thank you. Gave it a try, but still doesn't work.

Thank you for the suggestion. I gave it a try but it did not have any affect. I still have slow speeds when ether1 is set to 2.5G. I think I will order an SFP+ module to see if it can provide proper performance on the cage port.

I'll be interested to see your results, I've got my popcorn ready...

Just received a MikroTik S+RJ10 (RJ45) today. I'll get it installed and configured and post the results.

I'll be interested to see your results, I've got my popcorn ready...

Just received a MikroTik S+RJ10 (RJ45) today. I'll get it installed and configured and post the results.

Have you had a chance to test S+RJ10 ?

I'd like that CPU usage spike when accessing hosts data from the switch chip to fixed. I have an open ticket for it.
bridge/host/print or export causes spikes.
Today I didn't know what was eating the CPU again, turnes out it was a DHCP Server Leases window open in WinBox, window that by default has a "bridge port" column, and you guessed it, it gets its data from the switch chip.
So, for now, don't leave that window open in WinBox.

Just received a MikroTik S+RJ10 (RJ45) today. I'll get it installed and configured and post the results.

Have you had a chance to test S+RJ10 ?

Sorry, I've been out of town. Back now, will try to configure it later today.

Have you had a chance to test S+RJ10 ?

Well, bad news. I installed the S+RJ10 and I'm still getting the slower speeds. Behavior is exactly the same as using the Ether1 port. Also, it seems I cannot get the SFP+ module to negotiate at any speed other than 2.5G. I've tried to only advertise it as 1000M, but it makes no difference. It will still connect to the modem at 2.5G. My modem has only a single 2.5G port. Speeds are not configurable on the modem. Also, I have upgraded my RB5009 firmware to 7.1.1. Any ideas?

Really excited about the RB5009, I'm testing right now and the performance it's absolute over-killing compared to my OLD RB2011! Not sure if it's make improvement also from RB4011 that I've used elsewhere and I can't directly compare, but the gap it's impressive with the RB2011 managing dual 1Gbps WAN and some basic routing/firewalling.

Actually I was wondering how exactly the "multiple power source" works. I can safely plug my RB5009 connected to both DC jack input and POE on ETH1?
In this scenario, one power source is "active" power and other stay just "standby" ready for "fail over"? What will be the primary?

Thanks!

Have you had a chance to test S+RJ10 ?

Well, bad news. I installed the S+RJ10 and I'm still getting the slower speeds. Behavior is exactly the same as using the Ether1 port. Also, it seems I cannot get the SFP+ module to negotiate at any speed other than 2.5G. I've tried to only advertise it as 1000M, but it makes no difference. It will still connect to the modem at 2.5G. My modem has only a single 2.5G port. Speeds are not configurable on the modem. Also, I have upgraded my RB5009 firmware to 7.1.1. Any ideas?

I have been following this thread and signed up to relay that I am seeing the exact same thing with my RB5009. Xfinity 1.2gb connection, Motorola MB8611, eth1 port advertising 2.5gb and speedtests ~400Mbps. Configure eth1 to 1000M full and speedtests ~850Mbps. I also purchased a S+RJ10 and seeing, again, the exact same results. It will only negotiate 2.5Gb, and speedtests ~400Mbps. I’ve tried everything from the stock firmware to 7.2rc1. Very strange.

Well, bad news. I installed the S+RJ10 and I'm still getting the slower speeds. Behavior is exactly the same as using the Ether1 port. Also, it seems I cannot get the SFP+ module to negotiate at any speed other than 2.5G. I've tried to only advertise it as 1000M, but it makes no difference. It will still connect to the modem at 2.5G. My modem has only a single 2.5G port. Speeds are not configurable on the modem. Also, I have upgraded my RB5009 firmware to 7.1.1. Any ideas?

I have been following this thread and signed up to relay that I am seeing the exact same thing with my RB5009. Xfinity 1.2gb connection, Motorola MB8611, eth1 port advertising 2.5gb and speedtests ~400Mbps. Configure eth1 to 1000M full and speedtests ~850Mbps. I also purchased a S+RJ10 and seeing, again, the exact same results. It will only negotiate 2.5Gb, and speedtests ~400Mbps. I’ve tried everything from the stock firmware to 7.2rc1. Very strange.

I also tried firmware 7.2rc1. Same results. The SFP+ cage and the 2.5G ether1 port are attached to two of the three 'SerDes' interfaces on the Marvell 88E6393X switch chip. So it seems that the only viable fix will be firmware changes. But, if anyone discovers some magical combination of settings to get it to work, please post.

I have been following this thread and signed up to relay that I am seeing the exact same thing with my RB5009. Xfinity 1.2gb connection, Motorola MB8611, eth1 port advertising 2.5gb and speedtests ~400Mbps. Configure eth1 to 1000M full and speedtests ~850Mbps. I also purchased a S+RJ10 and seeing, again, the exact same results. It will only negotiate 2.5Gb, and speedtests ~400Mbps. I’ve tried everything from the stock firmware to 7.2rc1. Very strange.

I also tried firmware 7.2rc1. Same results. The SFP+ cage and the 2.5G ether1 port are attached to two of the three 'SerDes' interfaces on the Marvell 88E6393X switch chip. So it seems that the only viable fix will be firmware changes. But, if anyone discovers some magical combination of settings to get it to work, please post.

Unless its a problem similar Intel had with I225-V. Then it means it cannot be solved with firmware updatend and CR5009 is a crap

Actually I was wondering how exactly the "multiple power source" works. I can safely plug my RB5009 connected to both DC jack input and POE on ETH1?
In this scenario, one power source is "active" power and other stay just "standby" ready for "fail over"? What will be the primary?

The way other RB devices with multiple power source options work (and I don't see any reason for RB5009 to behave differently) is that they use a simple diode on each power input (to block power flowing out) and then all the sources are connected together. So if you use two power sources, one say 24V and another one say 48V, then the common point of both power lines will have voltage of the highest source (48V). The diode on input will block power from escaping through ports with lower voltage (the 24V one). And the whole device power consumption will be drawn from the source with highest voltage (48V). If the highest voltage source fails, then the voltage on common point of power lines will drop to 24V and will start to draw power from the power source with s(now) highest voltage. As RB devices actually use DC-DC down converters (they actually need 5 or 3.3 or even lower voltages), this drop of supply voltage is unnoticeable (unless device has voltage monitoring sensors).

So yes, it is the way that one power source (with the highest voltage) is used for all power consumption, the rest of power sources are stand-by.

Things become interesting if different power sources come with (almost) same voltages. If power source doesn't have exact voltage regulation (almost all power sources have relatively poor voltage regulators), increased current draw means slightly dropped voltage. Wires between power source and power sink have some resistance and increased current means larger voltage drop. All in all it means that voltage from highest-voltage power source can drop to the value of second-highest power source. At this point device starts to draw power also from the second-highest-voltage power source (but keeps to draw majority of power from the first power source).
So in this case device draws power from both sources simultaneously, but amount of power is different (how different depends on a few factors mentioned above).
If both power sources are strong enough, they are still backup one to another, but not really in the "stand by" state.

...
So yes, it is the way that one power source (with the highest voltage) is used for all power consumption, the rest of power sources are stand-by.
...

Thanks very detailed explanation, I just need now to rethink my "power backup" plan, since I would like to have the the standard power jack AC/DC converter as my primary source and not POE.

+1 On the issue where the RB5009 only gives ~400Mbps with 2.5Gbps to a MB8611.

If I keep the modem>router on the 2.5Gbps, but connect the sfp+ via 10G to my switch I do get the full ~1.4Gbps however, so it seems the issue is not with the modem>router link over 2.5Gbps, but the 2.5g>1G interface within the router.

As much fun as beta testing hardware is, I was really hoping to use this as my daily router, but an untested 2.5G functionality makes me wonder what other bugs are to come. Is the CCR2004-16G-2S+ the next cheapest option if I'm trying to pass the full 1.4Gbps to my switch via sfp+?

+1 On the issue where the RB5009 only gives ~400Mbps with 2.5Gbps to a MB8611.

If I keep the modem>router on the 2.5Gbps, but connect the sfp+ via 10G to my switch I do get the full ~1.4Gbps however, so it seems the issue is not with the modem>router link over 2.5Gbps, but the 2.5g>1G interface within the router.

Same kind of problem here as well. WAN port is connected to ZTE MC7010 5G modem. Negotiating the link as 2.5Gbps leads to poor speedtest results: max 300 Mbps. Negotiating the port to 1Gbps helps in getting 1 Gbps through.

Topology: computer -> 1Gbps Ethernet -> switch -> SFP+ 10 gbps optical fiber -> RB5009 -> 2.5 Gbps Ethenet -> PoE injector for the modem -> ZTE 5G modem.

Previously the modem has worked fine on 2.5 Gbps link using Zyxel's switch on the other end. When I have time, I could try to add that switch to the path as it has multiple 2.5Gbps ports to see if it makes any difference.

I'm testing with brand new RB5009 using default settings and RouterOS 7.0.9.

I have just installed RB5009 to my office environment, I have fully updated my router (both OS and firmware, to the latest 7.1.1).

Router itself is connected over ether1 (2.5gbit connetion, verified) to my cable modem (Fritzbox 6660), and to my CRS326 through 10gbit spf+ port.

I am managing to get full 1gbit connection in this setup (I am yet to try with two computers at the same time to see what is my exact cable limit), but the point is that my 2.5gbit ether works and I do not have slowdowns like many people reported.

Anyone else seeing an issue where all ethernet ports suddenly go down, and then a few seconds come back up again? Causing internet to reset etc

I'm running ROS v7.1.1 (issue was also on 7.1) on various different hardware seemingly without this issue, yet on both of my RB5009's (one purchased in the US, and the other in South Africa), I'm seeing the exact symptom. It comes and goes. Most of the time the issue isn't there. When it comes, it often recurs several times in the space of a few hours before it disappears again for hours/days.

Surely I don't somehow have the same network hardware fault on both my RB5009's? Could it be something with my configuration that's common to both routers that's triggering this ? Or am I not the only one seeing this issue ? It would be great if Mikrotik could acknowledge this issue and whether it's software/hardware-related so that anyone facing this issue would know where we stand

I have just installed RB5009 to my office environment, I have fully updated my router (both OS and firmware, to the latest 7.1.1).

Router itself is connected over ether1 (2.5gbit connetion, verified) to my cable modem (Fritzbox 6660), and to my CRS326 through 10gbit spf+ port.

I am managing to get full 1gbit connection in this setup (I am yet to try with two computers at the same time to see what is my exact cable limit), but the point is that my 2.5gbit ether works and I do not have slowdowns like many people reported.

rulleeeee (post #37) tested a similar setup to yours with the same good performance. It seems the problem is getting reduced performance when connecting through one of ethernet ports on the RB5009. If you are able, could you please try connecting directly to the RB5009 with your test computer instead of the CRS326? If so, please let us know if you see a performance drop or not.

I have just installed RB5009 to my office environment, I have fully updated my router (both OS and firmware, to the latest 7.1.1).

Router itself is connected over ether1 (2.5gbit connetion, verified) to my cable modem (Fritzbox 6660), and to my CRS326 through 10gbit spf+ port.

I am managing to get full 1gbit connection in this setup (I am yet to try with two computers at the same time to see what is my exact cable limit), but the point is that my 2.5gbit ether works and I do not have slowdowns like many people reported.

rulleeeee (post #37) tested a similar setup to yours with the same good performance. It seems the problem is getting reduced performance when connecting through one of ethernet ports on the RB5009. If you are able, could you please try connecting directly to the RB5009 with your test computer instead of the CRS326? If so, please let us know if you see a performance drop or not.

I'm actually having the issue even if I'm using a 1Gbps port in the switch attached via SFP+. For now I've just had to remove the 2.5Gbps advertising and lock it at 1Gbps so all devices can at least get that speed. Sort of sucks to leave ~300Mbps unused though. Tried everything I can think of, so at this point I think Mikrotik would need to look into this and probably update firmware to solve this issue.

Hi,
I just received my RB5009, sadly can confirm that when connecting to cable modem WAN port which is 2.5g, I lose approximately 50% on my bandwidth. from 1gig down to 500Mbs. Removing the 2.5g off ethernet1 and having it auto negotiate the connection at 1gb with the cable modem will give me a full ~950Mbps download. Very easy to reproduce, night and day results.

I noticed there are multiple people here with this issue, has anyone posted a bug report? viewtopic.php?t=152006

Does anyone know if the same issue happens on 2.5GB+ LAN connections?

Does anyone know if the same issue happens on 2.5GB+ LAN connections?

In my tests, 2.5GB has isssues on ether1 and also on SFP+. 10GB works fine on SFP+, 1GB works fine on all ports.
Never tested 5GB on SFP+ due to lack of a device supporting it.

While ether2-8 go to 1GB ports of the switch chip, SFP+ and ether1 are connected to 10/5/2.5/1 GB serdes ports.
I seems like 2.5GB is not working properly on those serdes ports, affecting ether1 and SFP+.
This is most likely a problem of the very complex switch chip serdes configuration, and fixable with future ROS releases.

Openwrt is working on RB5009 support and is facing similar problems with chip SoC serdes config complexities.

The RB4011 had similar issues with the SFP+ port fixed with later ROS updates.

I posted a bug report for this problem here; viewtopic.php?t=182691

Having a very weird issue where upload bandwidth is very limited (about half or less), but only when Fast Path is enabled: viewtopic.php?t=182694

Also starting to debug an issue where just IPv6 traffic is dropping around 50% traffic from my Synology NAS, but another host on macOS doesn't drop any packets to the same speed test host. Very odd,

Having a very weird issue where upload bandwidth is very limited (about half or less), but only when Fast Path is enabled: viewtopic.php?t=182694

Also starting to debug an issue where just IPv6 traffic is dropping around 50% traffic from my Synology NAS, but another host on macOS doesn't drop any packets to the same speed test host. Very odd,

Okay, stuff is getting weird. But basically I think I have the same issue as everyone else with the 2.5G link, with some twists.

2.5G to 2.5G
- 10G host: 2.2Gbps down, 700Mbps up -- about 35% speed up, this is new and only happens on the 10G hosts
- 1G host: 500Mbps down, 1Gbps up -- about 50% down, same as many others here

Force 1G to 2.5G
- 10G host: 1Gbps down, 500Mbps up -- what is going on, even forced to 1G I am getting about 50% up?!
- 1G host: 1Gbps up, 1Gbps down -- back to normal, as others has reported


I also still have issues with IPv6 being very slow and dropping packets. I know it is the RB5009 because bypassing it with the same exact host results in 2.2Gbps/2.2Gbps.

I bought the RB5009 only to be able to use 2G internet service I am paying for. Direct connection between cable modem 2.5G port and 2.5G port of my PC gives me that 2G download speed. The RB5009 2.5G port gives only 220Mbit/s (connection 2.5G full). Looks like they "mistakenly" labeled that ethernet1 port as 2.5G and enabled its advertisement as 2.5G full in setup. The only possible "fix" I see is to glue 1G sticker over 2.5G label under that port, remove 2.5G advertisement in setup (as default) - only one small problem here: the price, it's too much for 1G router.
Now, I have the question: Can you guys recommend the router with at least four 1G ports (with LACP bonding ability) and ONE 2.5G copper or SFP+ cage WORKING as 2.5G. Netgear MS510TX better than RB5009 - it gives ~570Mbits at least, but not 1G also.

The only possible "fix" I see is to glue 1G sticker over 2.5G label under that port, remove 2.5G advertisement in setup (as default)

Except even at 1G, I still see 10G hosts on my network at half speed. Clearly it's not specifically the 2.5G port, but rather some sort of wider spread issue with buffers/timings or something to do with mismatched speed interfaces.

I have big problems with turning on(booting) the device!
It turns on the third or fourth time.
I tried different power supplies and connected via POE.

The symptoms are:
Power supply - The blue LED lights up, and the green LED on the ETH2 port starts to dimly burn. Then nothing happens - the device is frozen! Similarly, on the POE, the orange LED lights up on the ETH1 port, and the green LED starts to dimly burn on the ETX2 port. Then nothing happens - the device is frozen! Green LED starts to glow dimly on the ETH2 port - both without a patch cord and with a patch cord!

At the same time, all ports are in the "down" status! There are no links, netinstall and winbox do not see anything because the ports are in the down!

After 3-4 times the device normally starts and works!

What does the dim green LED on the ETH2 port mean?
How to solve the problem?

I tried different firmware, used netinstall - nothing helped P.S.
Sorry for Google translate.

Every time I had these "dim leds" symptoms on MIkrotik hardware, it was because of bad caps or PSU problems (undervoltage, unstable voltage under load, etc.)
It would be very unfortunate if brand new RB5009 was already suffering from bad caps...
As these are hardware problems, doing netinstalls or changing anything on the software side can't fix it in any way.

I have tried different power supplies, also I have tried POE.

Since I am in Russia, it is possible that they will not change it for me under warranty. Then I'll take a look at the board.

Now I solved the problem like this:
On the switch, Mikrotik set up Power Cycle POE towards the router. with a timeout of 3 minutes.
And now, if the router hangs when loading, then the switch overloads it, here is an example of a log from the switch: As you can see, it loaded from the third time!

P.S.
Sorry for Google translate.

Hello people

I'm new here, and i want to purchase a router, RB5009UG+S+IN.. The problem is that i've read on multiple sites that it has problems with the 2.5G port?? Does anyone have a update on that ??
Can it be fixed (firmware update) or do i have to look for something else ??

thank you.

regards

So far I've tried a few suggestion from Mikrotik on the RB5009 re: 2.5G, such as some queue changes, but nothing has fixed the issue with me.

I'd steer clear of the RB5009 right now due to what seems like issues with mixed link speed routing, not just specifically the 2.5G port but also related to 1G <> 10G.

Mikrotik support told me to try the following:
No improvement. I still flap like a bird if I view the DHCP leases.

Mikrotik support told me to try the following:

Code: Select all

 /queue type set ethernet-default pfifo-limit=300
/queue interface set [find where queue!=no-queue] queue=ethernet-default

No improvement. I still flap like a bird if I view the DHCP leases.

Same thing here, for IPv6 dropped packets and the issues with >1G connections. Did nothing.

Guys, have you tried to use something better than CAT5e cables? I'm using CAT8 cable with a shielded connection. It was fairly hard to fit in, but only using this cable I could get stable 1Gbps connection between ISP' router, and my home router. With a standard CAT5e cable, I had 200-300mbps speedtest results at most, ceteris paribus. IMHO, 2.5Gbps connections should have higher requirements for connection quality.https://drive.google.com/file/d/1l37YDS ... sp=sharing

Indeed, I did a lookup for info on cabling.
https://www.cablematters.com/blog/Netwo ... t-ethernet
https://www.neweggbusiness.com/smartbuy ... ings-know/
It appears that Cat6 cable is a must have, to get anything above 1Gbps. However, you should take into account EM interference at your specific location. It could just happen to be very high, and CAT6A/CAT8 cables would have to be used (like in my case).

Hello again:)

I'm using Cat 6 cables and is about 3 meter from the modem..So i think i should be good right??

regards

Hello again:)

I'm using Cat 6 cables and is about 3 meter from the modem..So i think i should be good right??

regards

I had a 3m CAT6 cable, and at some point after having some issues with a router, I've tried to switch my cable as well. At that point it didn't solve the issue, but afterwards, I found that using CAT8 cable increased my Internet speeds to 800/900Mbps on 1Gbps connection. Previously on CAT6 cable I had 300/200mbit limit. Which is ridiculous, because its supposed to hold up to 10Gbps. I don't know what kind of interference I'm having around the router, but it'd be interesting to find that out. If someone knows what kind of equipment is better to use to measure EM fields, would be glad to get a hint =). I live in a kilometer range from high voltage power lines, and 3km from local electrical grid station.
Back to the subject of a 5009 router problem - if one person got it working at above 1Gpbs on 2.5G connection, and the rest could not, this could be something everyone is doing wrong. I haven't got this router yet, but was planning to purchase one to have 2.5G LAN with low deployment cost. Could someone please test all their devices used in setup separately and using CAT8 cables? Thank you.

CAT6 alone does not say anything about shielding.

Hello people

I'm new here, and i want to purchase a router, RB5009UG+S+IN.. The problem is that i've read on multiple sites that it has problems with the 2.5G port?? Does anyone have a update on that ??
Can it be fixed (firmware update) or do i have to look for something else ??

thank you.

regards

Hello,

I have a RB5009 in production and works perfectly, I got it right when it came out and just swapped my RB4011 with it.
I have 11d of uptime, no issues with the 2.5g port or the 10g SFP+(that is connected to a CRS305), also no link flapping at all.
I think that the issues that some people are having with the 2.5G port is most probably issue with the another end, as I have seen the issues most happen with Cable Modems, I have seen issues with Broadcomm based devices that have weird behaviours when connected at 2.5G. I have a 2.5G AX access point that was having some slow speed (client upload capped to 100mbps) when the 2.5g port was linked at 2.5G, that did not happened when the same port was set to 1G. Note that this happened to when I was still using the RB4011 and it was connected to the CRS305 and the AP connected to the same port using a MK SFP+ to RJ45 10g module.

Hi
I have just set up my router, But there is a weird problem. on Ethernet2 it can't advertise 1000m half or full. i have connected my 1g device to it they communicate with 100m and when i moved my device to other port the link was up with 1000m. when i forced the 1g advertise to eth2, no link was established.
is my device defected ?

Hello again:)

I'm using Cat 6 cables and is about 3 meter from the modem..So i think i should be good right??

regards

I had a 3m CAT6 cable, and at some point after having some issues with a router, I've tried to switch my cable as well. At that point it didn't solve the issue, but afterwards, I found that using CAT8 cable increased my Internet speeds to 800/900Mbps on 1Gbps connection. Previously on CAT6 cable I had 300/200mbit limit. Which is ridiculous, because its supposed to hold up to 10Gbps. I don't know what kind of interference I'm having around the router, but it'd be interesting to find that out. If someone knows what kind of equipment is better to use to measure EM fields, would be glad to get a hint =). I live in a kilometer range from high voltage power lines, and 3km from local electrical grid station.
Back to the subject of a 5009 router problem - if one person got it working at above 1Gpbs on 2.5G connection, and the rest could not, this could be something everyone is doing wrong. I haven't got this router yet, but was planning to purchase one to have 2.5G LAN with low deployment cost. Could someone please test all their devices used in setup separately and using CAT8 cables? Thank you.

I just tried a CAT 8 cable, and the only difference was my wallet was $8 lighter. Same poor (480Mbps) download performance with a 2.5Gbps link. Changing the router-to-modem link to 1Gbps gives me 900Mbps downloads.

Y'all having issues with your 2.5g WAN connections are posting info only about one-half of the puzzle, and your omission of any discussion concerning other aspects involved in determining compatibility is telling.

The speed of the data connection your ISP is providing to you, in this case 2.5g, is a necessary, but not sufficient condition to establish a 2.5g connection through your 2.5g port on your MT router.

What standard is your ISP using? There is a ton of information on the internet, going back several years, concerning compatibility issues with ISP's 2.5g data service, especially with respect to 2.5g GPON (but not limited to 2.5g GPON).

When such compatibility issues arise, the most common result appears to be routers falling back to negotiating a 1g connection. This happens to match perfectly the circumstances that several people in this thread claim to have experienced.

Below are various discussions, some quite of extensive, concerning router compatibility issues with 2.5g connections. The first link is to the granddaddy of all effort to make personal routers work with an ISP's 2.5g service.

https://www.dslreports.com/forum/r32230 ... 57810S-NIC

https://community.ui.com/questions/Goog ... 026884f8f0

https://www.reddit.com/r/googlefiber/co ... _my_unifi/

Y'all having issues with your 2.5g WAN connections are posting info only about one-half of the puzzle, and your omission of any discussion concerning other aspects involved in determining compatibility is telling.

The links you provided are not at all applicable to this issue. Those links are about using Google 2.5g over fiber. The RB5009 problem is with the 2.5g copper port. The 2.5g port doesn't have any issues connecting at 2.5g and will run at full speed if you test from a PC on the SFP+ port connected at 2.5/5/10g. The problem is that any 1g devices on your network will get around 500mbps. Mikrotik has confirmed the issue on their side and is working on a fix.

Y'all having issues with your 2.5g WAN connections are posting info only about one-half of the puzzle, and your omission of any discussion concerning other aspects involved in determining compatibility is telling.

The links you provided are not at all applicable to this issue. Those links are about using Google 2.5g over fiber. The RB5009 problem is with the 2.5g copper port. The 2.5g port doesn't have any issues connecting at 2.5g and will run at full speed if you test from a PC on the SFP+ port connected at 2.5/5/10g. The problem is that any 1g devices on your network will get around 500mbps. Mikrotik has confirmed the issue on their side and is working on a fix.

Where did Mikrotik confirm the issue and state that they are working on a fix? I was hoping that was the case but did not find anything official.

The links you provided are not at all applicable to this issue. Those links are about using Google 2.5g over fiber. The RB5009 problem is with the 2.5g copper port. The 2.5g port doesn't have any issues connecting at 2.5g and will run at full speed if you test from a PC on the SFP+ port connected at 2.5/5/10g. The problem is that any 1g devices on your network will get around 500mbps. Mikrotik has confirmed the issue on their side and is working on a fix.

Where did Mikrotik confirm the issue and state that they are working on a fix? I was hoping that was the case but did not find anything official.

They confirmed it in a support ticket with forum user wojo, post is here: viewtopic.php?p=915477#p915477

Mikrotik support has some alpha firmware builds (dated Feb 25) that seem to fix the issue. I'd expect something to be released in the dev channel before too long.

I tried 4 different versions dated Feb 25.. Version 7.99. Unfortunately the issue was not resolved for me with that version, did anyone else have better luck?

I tried 4 different versions dated Feb 25.. Version 7.99. Unfortunately the issue was not resolved for me with that version, did anyone else have better luck?

I did have one PC with a 2.5gbps Asus USB network adapter that was acting a little strange. Enabled flow control and updated the Realtek driver and it started acting a bit more consistent. Here's my results after working that out (just pasting what I sent to Mikrotik)

I have tested all four versions and the first three appear to fix the issue. The fourth appears to be a bit better than the current stable firmware version but does still have reduced speed when going from the 2.5gbps WAN to a 1gbps PC.

ISP bandwidth is 1.2gbps but can sometimes go a little higher (1.3 or even 1.4)



7.1.3

1GB ISP / 1GB PC

949.13mbps down

41.83mbps up



2.5GB ISP / 2.5GB PC

1281.56mbps down

41.91mbps up



2.5GB ISP / 1GB PC

Lower than expected - 377.87mbps down

41.91mbps up





Test fw v1

2.5GB ISP / 1GB PC

939mbps down

41.88mbps up



2.5GB ISP / 2.5GB PC

1302.56mbps down

38.62mbps up



Setting speeds to advertise on sfp plus port not working



Test fw v2

Approx the same as test v1



Test v3

Approx the same as test v1



Test v4

2.5GB ISP / 1GB PC

Lower than expected - Varying, between 600 and 815mbps down

41.89mbps up



2.5GB ISP / 2.5GB PC

1300mbps down

Are you plugging the 1G clients directly into the RB5009? When I do that the new FW is a bit better, but my typical setup has the RB5009 connected via 10G DAC to my switch, the 1G clients in there saw no noticeable improvement while the 10G clients in the same switch still got ~1400Mbps.

Are you plugging the 1G clients directly into the RB5009? When I do that the new FW is a bit better, but my typical setup has the RB5009 connected via 10G DAC to my switch, the 1G clients in there saw no noticeable improvement while the 10G clients in the same switch still got ~1400Mbps.

For the the above tests I had each device connected directly to the RB5009. I do have a switch connected to the RB5009 with multiple 1G links using LACP. I just tested with my PC in the switch and it seems to give the full 1G. I don't have a 10G switch at home unfortunately.

I've tested all four 7.99's. I use ether5,6,7,8 as LACP. I tried 1G and 2.5G on ether1. Looks like all builds do not increase speed above 1G in any settings. (Use different computers at the same time, It's combined speed). 1G +2G(LACP) gives 450Mbits + 450Mbits - total < 1G
V1: 2.5G - 937Mbits
1G - 938Mbits
V2: 2.5G - 935Mbits
1G - 938Mbits
V3: 2.5G - 860Mbits
1G - 935Mbits
V4: 2.5G - 425Mbits
1G - 933Mbits

7.2rc4 2.5G - 230Mbits
1G - 945Mbits

Direct cat6a 30" 2.5G NIC to cable modem 2.5G port gives 2040Mbits

Really interesting that some people are not seeing it fixed.

For me V1, V2 and V3 gave full 2.3Gbps both ways. V4 was a tad slower, 800-900Mbps start and it would go up from there.

Really interesting that some people are not seeing it fixed.

For me V1, V2 and V3 gave full 2.3Gbps both ways. V4 was a tad slower, 800-900Mbps start and it would go up from there.

What is your setup? What are your ports?
I've tested ether1 connected to 2.5G(full duplex) cable modem (2G service), and two 1G NIC computers on ether 2 and ether3. Running my ISP speed test on one and Google speed test on other at the same time. Speed drops on a first computer when I started second test on a second computer. Combined speed never exceeded 1G on V1,V2, 800M on V3 and 400M on V4

Really interesting that some people are not seeing it fixed.

For me V1, V2 and V3 gave full 2.3Gbps both ways. V4 was a tad slower, 800-900Mbps start and it would go up from there.

What is your setup? What are your ports?
I've tested ether1 connected to 2.5G(full duplex) cable modem (2G service), and two 1G NIC computers on ether 2 and ether3. Running my ISP speed test on one and Google speed test on other at the same time. Speed drops on a first computer when I started second test on a second computer. Combined speed never exceeded 1G on V1,V2, 800M on V3 and 400M on V4

My topology is:

2.5G AT&T UVerse ---copper---> 2.5G -- RB5009 -- 10G SFP+ ----10G DAC-----> CRS238 10G SFP+

then off that switch, a mixture of

- Direct 10G SFP host
- Direct 1G host
- 10G SFP+ to 10GBaseT (over Cat5e) to 10G SFP+ on another CRS238 10G or TPLink 10G switch
- ... and that CRS238 or TPLink to other 10G or 1G hosts

In all situations, I'm getting a full 2.3Gbps with V1/2/3 both ways via AT&T.

My topology is:

2.5G AT&T UVerse ---copper---> 2.5G -- RB5009 -- 10G SFP+ ----10G DAC-----> CRS238 10G SFP+

then off that switch, a mixture of

- Direct 10G SFP host
- Direct 1G host
- 10G SFP+ to 10GBaseT (over Cat5e) to 10G SFP+ on another CRS238 10G or TPLink 10G switch
- ... and that CRS238 or TPLink to other 10G or 1G hosts

In all situations, I'm getting a full 2.3Gbps with V1/2/3 both ways via AT&T.

My test shows switching between ether1 and rest of 1G ports not working correctly.
Any way, it's great, they fixed SFP+ and ether1 communication. Time to put copper module into SFP+
BTW, what SFP+ module do you have in there? I heard Mikrotik's one is running too hot.

My topology is:

2.5G AT&T UVerse ---copper---> 2.5G -- RB5009 -- 10G SFP+ ----10G DAC-----> CRS238 10G SFP+

then off that switch, a mixture of

- Direct 10G SFP host
- Direct 1G host
- 10G SFP+ to 10GBaseT (over Cat5e) to 10G SFP+ on another CRS238 10G or TPLink 10G switch
- ... and that CRS238 or TPLink to other 10G or 1G hosts

In all situations, I'm getting a full 2.3Gbps with V1/2/3 both ways via AT&T.

My test shows switching between ether1 and rest of 1G ports not working correctly.
Any way, it's great, they fixed SFP+ and ether1 communication. Time to put copper module into SFP+
BTW, what SFP+ module do you have in there? I heard Mikrotik's one is running too hot.

Fiber Mall 10G SFP+ to RJ45 for Cisco SFP-10G-T-80 Compatible, 10GBASE-T SFP+ Copper RJ-45 80m Transceiver Module

Can find on Amazon. They are the 80m capable Broadcom chips. No drop outs running up 3 floors on Cat5e with zero issues at full 10Gbps.

I'll test a bit more on the 1Gbps ports. I've only done that with a cascaded switch down to 1Gbps.

Oh do you have flow control on? That's huge for me on mixed speed networks. Without it I see a massive reduction in speeds.

Fiber Mall 10G SFP+ to RJ45 for Cisco SFP-10G-T-80 Compatible, 10GBASE-T SFP+ Copper RJ-45 80m Transceiver Module

Can find on Amazon. They are the 80m capable Broadcom chips. No drop outs running up 3 floors on Cat5e with zero issues at full 10Gbps.

I'll test a bit more on the 1Gbps ports. I've only done that with a cascaded switch down to 1Gbps.

Oh do you have flow control on? That's huge for me on mixed speed networks. Without it I see a massive reduction in speeds.

Thanks, I found it.
Flow control is off by default on all the 5009 ports. On 7.2cr4 "on" or "off" makes no difference, so I left it "as is" in all the tests. Probably will need to retest.

Yes, I run this command after each reboot - "/interface ethernet switch set switch1 cpu-flow-control=yes" - but Rx and Tx flow control for each port of 5009 gives no difference in speed.
I have 48 ports 1G switch connected with 4ports LACP to 5009 (1' shielded cat 6a cables), 2 servers with 2 1G LACP ports, All LAN communications are as expected, but only internal link in 5009 between ether1 and other 1G ports are limited to 1G total (like connection on 2.5 is actually 1G , or communication speed from ether2,3,4,5,6,7,8 to ether1 is limited to 1G)

Tech Support sucks. Looks like hardware is not fixable. Time to sell this piece of *** on Ebay and never, ever buy anything from mikrotik!

Sell it, sell it!.

Tech Support sucks. Looks like hardware is not fixable. Time to sell this piece of *** on Ebay and never, ever buy anything from mikrotik!

Did you get official reply from Tech support ?

Tech Support sucks. Looks like hardware is not fixable. Time to sell this piece of *** on Ebay and never, ever buy anything from mikrotik!

Did you get official reply from Tech support ?

Yes, I opened the case, but they pretend something else is wrong. Like LACP not giving more than 1G regardless 4 ports used for it. OK! So, I use ether1 as a regular port, not WAN port, and directly connect 3 computers to ether2, 3 and 4. And computer with proven working 2.5G port connected to ether1. All the same. Port ether1 is nor 2.5G port. It's 1G port. They refuse to accept their hardware is not tested well and is actually an alfa version. I was downloading movie files from ether1 connected computer to three others at the same time, Combined speed is less than 45MB (360Mb/s), when 2.5G removed from ether1 advertisement - speed goes up to 113MB(904Mb/s). And my HD is M.2 Samsung 980 PRO SSD 2TB PCIe NVMe Gen 4.
So, I am done with this ***
So much easier to deal with American company, I am back to EdgeOS.

(PS they removed 5009 from Amazon....)

Running 7.2rc5 my 10G hosts are doing well, 2.3Gbps bidirectional on my 2Gbps over-provisioned AT&T link.

However, my 1G hosts are doing about 780Mbps down, and 950Mbps up. Not sure why the download is impacted. Happens on many hosts, and even with /interface/ethernet/switch/set 0 cpu-flow-control=no. May be my ISP fault (even though I can get up to 2.3Gbps), researching...

Either way a huge improvement.

UPDATE: In general, my 1G speeds are lower. 780-800-ish Mbps, and I'm pretty sure this is due to the upgrade. Uploads are not impacted.

With latest 7.2rc5 firmware on 1G hosts download speed (with 2.5G advertisement OFF) is 935Mb. And download speed on 1G hosts (with 2.5G advertisement ON) is 794Mb. I do not measure upload, because my ISP plan is: 2Gb down and only 75Mb up. Running two or three hosts at the same time gives the same combined speeds. Same cable from modem connected to PC NIC instead of RB5009 gives beautiful 2Gb download!

With latest 7.2rc5 firmware on 1G hosts download speed (with 2.5G advertisement OFF) is 935Mb. And download speed on 1G hosts (with 2.5G advertisement ON) is 794Mb. I do not measure upload, because my ISP plan is: 2Gb down and only 75Mb up. Running two or three hosts at the same time gives the same combined speeds. Same cable from modem connected to PC NIC instead of RB5009 gives beautiful 2Gb download!

Ah you are seeing the same issue with 1G download that I am. Good to know I'm not alone. Working with Mikrotik on this, open a ticket and they may request info (stats counters, etc).

For me, everything fine. 2 PC on GBit of CRS326, a third one with 2.5 Gbit on 5009, CRS and 5009 connected via sfp+. Throughput about 1.8x Gbit as expected. Tested with original factory fw, about 8 weeks ago. Seems not to be a general hw problem. Also the ductus sounds like pebkac.

DNAT

Are you plugging the 1G clients directly into the RB5009? When I do that the new FW is a bit better, but my typical setup has the RB5009 connected via 10G DAC to my switch, the 1G clients in there saw no noticeable improvement while the 10G clients in the same switch still got ~1400Mbps

My topology is still 2.5G AT&T Fiber <---> RB5009 <---10G DAC---> CRS328 <---> 1G/10G clients.

Are you plugging the 1G clients directly into the RB5009? When I do that the new FW is a bit better, but my typical setup has the RB5009 connected via 10G DAC to my switch, the 1G clients in there saw no noticeable improvement while the 10G clients in the same switch still got ~1400Mbps.

I tried both: 1G clients directly to RB5009 1G ports (ether2,3,4) or to 1G switch connected with a LACP team of 4 ports (ether5,6,7,8). Results ~same.
And BTW, 5009 ether1 does not see modem 2.5G as 2.5G but as 1000M, but connection reported as 2.5G, full duplex on both devices.
Do you ground 5009 and how? (it is not in the rack)

Today I've got Wiitek SFP-10G-T. (It supports 1G,2.5G,5G,10G)
First, I connect my PC with 2.5G NIC to SFP+ at 2.5G,full. Download speed test (ether1 advertised as 1G) shows same ~900Mbs.
When I turned ether1 advertisement to 2.5G,full, my download speed drops to 95Mb/s (new firmware 7.2rc6) What was that?
Next step: SFP+ as WAN port, connection to modem as 2.5G, full. Ether1 as regular bridge port, 2.5G,full, connected to PC 2.5G NIC - download speed test shows ~2G!!!
Now, the final test: same PC connected to ether2 (should be 1G) Oops! Download speed on all 1G ports is only ~740Mb/s Tested multiple times - same result, sad...

This looks like congestion problems if download traffic ingresses with 2.5GB at ether1 and gets bridged (switched) towards the ether2-8 1GB ports.
As the bridge ingress port is faster than the egress port, the switch chip queue for the egress port overruns and packets get dropped.
This causes packet loss and issues with TCPIP windows sizes and ACKs and reduces the effective bandwith.

Has anyone checked if the traffic stats on 2.5GB ether1 towards the cable modem shows RX drops?
And for TX drops on ether2-8 towrds 1GB clients?
If there are, it could be solved by allowing bridge to produce backpressure towards the incoming download using flow control:
This allows the RB5009 to apply back pressure towards the modem in case the 1GB output queue towards the client on ether2 is overrun by incoming 2.5GB downlad traffic on ether1.

This is how the same problem is often solved on 100GB core switches aggregating 10GB link to/from floor switches.

Code: Select all

         ---------------------------------Download--------------------------------------->
         CableModem<---2.5GB--->[ether1].RB5009bridge.[ether2]<---1GB--->LANclient
             |                     | 
Flow RX      on                   off
Flow TX      off                  on
         <---------back pressure--------

This allows the RB5009 to apply back pressure towards the modem in case the 1GB output queue towards the client on ether2 is overrun by incoming 2.5GB downlad traffic on ether1.

This is how the same problem is often solved on 100GB core switches aggregating 10GB link to/from floor switches.

Unfortunately my cable modem (FAST3896U) has everything (like firewall, wifi6, router, 4 ether ports) but no setting for flow control. Love to try, but...

Today I've got Wiitek SFP-10G-T. (It supports 1G,2.5G,5G,10G)
First, I connect my PC with 2.5G NIC to SFP+ at 2.5G,full. Download speed test (ether1 advertised as 1G) shows same ~900Mbs.
When I turned ether1 advertisement to 2.5G,full, my download speed drops to 95Mb/s (new firmware 7.2rc6) What was that?

I have a similar problem with 1Gb devices and WiFi APs connected through the SFP+ port on the RB5009. I'm running 7.2.3. I'm using the 2.5Gb port for my cable modem and have the SFP+ connected to a CRS326-24S+2Q-RM with fiber. I got a little more speed... about 200Mbps. Playing with flow control helps a little, but ~200 is the best I've come up with. I forget which settings worked best... I tried "all" of them, then punted and plugged some stuff into the 1Gb ports on the RB5009. 10Gb hosts and my laptop using an Asus 2.5Gb USB adapter and an S+RJ10 get full internet speed, about 1.4Gbps. If I gimp ether1 to 1Gb the WiFi and computers connected at 1Gb work as they should, but of course that also limits the 10Gb machines and 2.5Gb laptop to 1Gb.

Next step: SFP+ as WAN port, connection to modem as 2.5G, full. Ether1 as regular bridge port, 2.5G,full, connected to PC 2.5G NIC - download speed test shows ~2G!!!
Now, the final test: same PC connected to ether2 (should be 1G) Oops! Download speed on all 1G ports is only ~740Mb/s Tested multiple times - same result, sad...

Disabling hardware offload in the bridge settings for the 1Gb ports seems to help, at least with my simple setup. Worth a try if you haven't tried it already. I'm getting close to 900 from the 1Gb ports on the RB5009 from the Internet and nearly full speed locally. OTOH your Internet connection is faster than mine, so maybe that makes the problem worse?

The thing that really irks me about this problem is I can't run a >1Gbps trunk line up to the second floor and hang a switch with 1Gb devices on it without gimping their internet speed, downgrading my service/setting ether1 to 1Gbps and gimping everything, or throwing out a router that's only a few months old. I have two existing lines (going to different rooms), but adding more would involve a bunch of drywall work.. some of it on a ceiling.

.... and an S+RJ10 get full internet speed, about 1.4Gbps......

Thank you for the tip (hardware offload) - will try it today at night.
Just curious, what is the temperature of the S+RJ10 module during the test? I bought Wiitek SFP-10G-T and It is not made by MikroTik but it's temperature is 43C max.

Hotspot till have a problem.

Just curious, what is the temperature of the S+RJ10 module during the test? I bought Wiitek SFP-10G-T and It is not made by MikroTik but it's temperature is 43C max.

Load doesn't actually seem to do much to it's temperature. My S+RJ10 idles around 47-48C and only goes up to 50C whacking it with repeated local speed tests or file transfers. I'm sure I could get it a little hotter with a load generator. My setup is maybe a little out on the cool end for S+RJ10 temps outside of a datacenter. It's all alone in a group of 8 ports in my CRS326-24S+2Q-RM, which has 3 cooling fans. The port count is overkill for me, but 8 ports wasn't enough and I thought the 2 QSFPs were worth the price difference over the CRS317.

I thought of one more thing I should try - removing ports from the bridge on the RB5009. Maybe I'll get to it this weekend. If that "fixes" internet speed to 1Gb hosts connected through the SFP+ port I'll be quite happy. My target setup is the RB5009 only has 2 connections: 2.5G to the modem and 10Gb SFP+ to the CRS326. The idea here is to disable all hardware acceleration and just make the CPU do all the work. As long as it can manage a file download (aka single TCP connection) at 1.4Gb I'll be happy.

If that doesn't work I think I'm just going to make 2 VLANs for computers instead of one. The >1Gb VLAN would connect to the RB5009 using the SFP+ port and everything else would use a 1Gb port.

Well, I tried turning off Hardware offloading of ports in the bridge.... SFP+ 2Gb down 75Mb up (cable modem), 2.5Gb NIC on ether1 - 2Gb down, 75Mb up (No difference with HW offloading on or off), on ether2 with HW offloading on - 790Mb down. HW offloading off - 800Mb+ down! BUT only first test, second one back to 790Mb down. HW offloading on again - speed goes to 800Mb+, but only first test, then goes back to 790Mb. So, no luck...