Maybe time for a router upgrade?

[jasons6930]

Hey guys,

currently running a Draytek 2862 for modem and router duties but getting bored with that so fancy a change.

Been looking at the Microtik RB3011 and seems to be pretty good.

Is this better than the Ubiquiti Er4?

spec wise it seems to fall behind the ubiquiti but I guess that doesn’t always mean performance is worse?

cheers!

 

[coxhaus]

Let us know how it goes?

 

[jasons6930]

Will do!

I do have an er4 somewhere and I thought when I used it last, it didn’t seem to be that snappy so reverted back to the draytek.

I have seen the microtik and it seems like a reasonable price so I may give that a go, although the os looks a little daunting.

Will probably be back here for some serious help setting it up!

 

[ddaenen1]

Hey guys,

currently running a Draytek 2862 for modem and router duties but getting bored with that so fancy a change.

Been looking at the Microtik RB3011 and seems to be pretty good.

Is this better than the Ubiquiti Er4?

spec wise it seems to fall behind the ubiquiti but I guess that doesn’t always mean performance is worse?

cheers!

I am running an RB3011UiAS-RM since several months and am very happy with it. Runs like a charm, easy to set up (easier than my previous Ubiquiti ERL-3) a wide array of configuration options and functionality and last but not least, rock-stable. I am looking forward to see how it performs with my 1Gbps-line once they swapped out my modem but from the specs, it should have any issues dealing with that.

 

[Trip]

The ER-4 has a dual-core 1Ghz MIPS Cavium in it; the RB3011 a dual-core 1.4Ghz ARM, so they're roughly equivalent in processing power (the MIPS being better at most kind of routing-related math than ARM, so it usually takes roughly half again as much clock on an ARM chip to equal MIPS performance, depending on what's running on the box of course).

RouterOS is not for the faint of heart, but their GUI access through WinBox and/or the web UI makes it a lot more approachable when first starting out. Here's a great intro video on the basics.

 

[jasons6930]

The ER-4 has a dual-core 1Ghz MIPS Cavium in it; the RB3011 a dual-core 1.4Ghz ARM, so they're roughly equivalent in processing power (the MIPS being better at most kind of routing-related math than ARM, so it usually takes roughly half again as much clock on an ARM chip to equal MIPS performance, depending on what's running on the box of course).

RouterOS is not for the faint of heart, but their GUI access through WinBox and/or the web UI makes it a lot more approachable when first starting out. Here's a great intro video on the basics.

Thanks guys,

Trip, must admit i had a look at the video and i am far from a networking specialist, or expert, so i would end up trawling the internet for information, and there does seem to be a lack of it, even their own forums seem a little 'quiet'.

So, with that in mind, i think i may stick with Draytek, purchasing a switch or two and keeping the 2862 or even upgrading to the 2926 as i am looking at having a 100Mbps fibre installed, although having looked at the specs, performance wise it is identical to the 2862 i am running at the moment.

 

[coxhaus]

I would think the Cisco RV340 router would fall into this category. It works well for me. It is fairly easy to setup as it contains several wizards for easy setup.

Here is an Cisco emulator.
https://www.cisco.com/assets/sol/sb...mulator_v1-0-03-15/index.html#/GettingStarted

 

[Trip]

@jasons6930 - In what way didn't the ER-4 feel all that "snappy"? While using the web gui? Networking behavior of the endpoints themselves? Or other? I'm curious because I've had just the opposite takeaway from having deployed several of them now at customer sites.

I guess this comes down to what are you bored of with the Draytek, and does that equate to it doing anything sub-par that you want to address? On-board features/services, WAN connection quality, or something else?

 

[coxhaus]

I like running a layer 3 switch for local routing. It gives the router an easier job of just handling internet traffic. I am working on my daughter's Cisco L3 switch, SG500X-24, where the switch does all the QoS for 19 IP phones. She is running my old Cisco RV320 router.

 

[jasons6930]

@jasons6930 - In what way didn't the ER-4 feel all that "snappy"? While using the web gui? Networking behavior of the endpoints themselves? Or other? I'm curious because I've had just the opposite takeaway from having deployed several of them now at customer sites.

I guess this comes down to what are you bored of with the Draytek, and does that equate to it doing anything sub-par that you want to address? On-board features/services, WAN connection quality, or something else?

Hi Trip,

Sorry, when i mean snappy, i meant geneal usage, internet browsing, a bit of home networking.

I ran a throughout test, albeit from my iphone to my Macbook pro and these are the results..

IPv4




Ipv6



To me, the throughput doesn't look that good, considering my network set up?


With regards to the ER4, i had it set up and running on the "default" settings so i'm not sure if there was something i should have done to speed things up somewhere.

 

[Trip]

So right off the bat, the biggest issue with your testing is that you had multiple hops of wireless in the mix, which is a no-no from the simple standpoint of testing a wired router. When doing so, you want to cut out wireless, plus as many unnecessary wired hops in the path as possible. Just knowing that alone, I'm 99% sure the performance depredation you're attributing to the ER-4 was being caused by the iPhone and its radio and/or the AP/router and its radio, and/or your MBP and its radio, if it wasn't wired into ER-4, and/or a switch connected to the ER-4 if you had one.

I would retest with 1 host directly wired to the ER-4, running an iperf server on the host, then use the iperf client on the ER-4 (found under "Tools" in the GUI) to run your test. (Or if you want client control or better reporting, run iperf client on the host, and server on the ER-4). If all is setup properly, I would predict full gigabit wire speed and practically zero latency or jitter.

If you're thinking the ER-4 was performing poorly because it was connected up to all the other devices, you'd have to test speed through the ER-4 and each successive device, added back into the network one by one, and do so then against the DrayTek or whatever router in place of the ER-4. After doing so, I suspect you'd find it's one or all of the wifi clients/stations in the network path that is/are causing the performance issues, and doing so to an equal degree across all routers, regardless of which router you have in place. I could be wrong, but that's really the only way to know for sure.

 

[jasons6930]

Thanks for that.

I can't do too much testing at the moment due to the network being used by my other half with work etc.

However, I did do a quick retest, this time using a MacMini that is wired, and my MBP that is wifi and things have improved a reasonable amount!

Although, still a fair loss..

This is from the MacMini as the server & MBP as the client...



I should have the place to myself tomorrow, so i will try and do a fully wired test, but the only other thing I have with a network port is an ageing Asus laptop!

 

[coxhaus]

Trip, how can you tell there are multiple hops in the wireless? I don't see it but I am not familiar with this testing. And yes multiple wireless hops are not good.

It seems like to me he is just testing layer 2 so the ER4 layer 3 is not kicking in unless he is testing across the internet.

It takes me 2ms to cross my local LAN using wireless. I start on the wireless to my L3 switch which routes me to my router on a different network. This is layer 3, 2ms. Layer 2 is I start with the wireless AP continue into a POE+ layer2 switch then to my L3 switch then across layer3 and layer2 into my router. Wired is less than 1ms. Wireless is my slowdown. My router is not involved unless I go to the internet.

My gear is all Cisco small business equipment. I am using ping to test with.

 

[jasons6930]

It's probably because I was testing using an Iphone and a Macbook Pro which are both Wifi only.

 

[coxhaus]

I have ping on my iPhone but I have blocking in place because my iPhones run in the guest network and they can not access my LAN network. Let me think about this and maybe I can try laptop to iPhone both on the wireless.

Is there a way to see the IP in the iPhone?

 

[coxhaus]

I have a strange thing happening. I put my laptop on the same SSID as my iPhone7+ and I ping. I get less than 10ms when I ping from my iPhone7+ to my wireless laptop on the same AP. If I ping from my laptop to my iPhone7+ it is about 50ms, This makes no sense to me. I rebooted both my wireless APs and it is the same result. My iPhone7+ has 650AC connection and my laptop has 702AC connection on the same wireless AP.

 

[jasons6930]

Sorry was away from my computer.

Anyway, I can only put this down to the nic in the phone and or MacBook as Trip suggested.

Your IP address of your iPhone should be under the wireless network you are connected to

 

[coxhaus]

Sorry was away from my computer.

Anyway, I can only put this down to the nic in the phone and or MacBook as Trip suggested.

Your IP address of your iPhone should be under the wireless network you are connected to

I assume this is the iPhone7+ but I would like someone else to confirm it. Yes, only layer 2.

 

[jasons6930]

The IP address location is the same for all Iphones, however, i am using an iphone 8

 

[coxhaus]

The IP address location is the same for all Iphones, however, i am using an iphone 8

Could you run a ping test both ways and report the latency. I can't run speed test any more as my laptop is locked down tight.