Hey.
I was just wondering, if either the red L009 or black 5011 routers are a full metal enclosure/case? Or are they metal lower with an upper plastic case?
Thanks,
Gavin.
Full metal. Sorry, genuinely thought so. Thanks to @antonsb for the correction!
Even if it wasn’t, Mikrotik actually uses “proper” plastics for their enclosures, so they don’t crack, fall apart, lose flexibility etc. even after many years of use. In direct sunlight they do acquire a yellowish tint, but they are not actually damaged.
Upper covers are from plastic
2 Likes
They have a (thick) metal base that doubles as heatsink and a plastic upper half cover, see:
https://www.servethehome.com/mikrotik-rb5009upr-s-in-poe-router-mini-review-marvell-arm/
and here (around 6 minutes)
https://www.youtube.com/watch?v=Qvg1LQonuhM
you can see how NOT to open a RB5009 (do not remove the board from the metal base ( or you should change the thermal pads with new ones)
2 Likes
Cheers!
I have an bit of an RF environment near an FM transmitter (not high power, <20 Watts RF) where I prefer to keep things metal. I’m replacing an RB2011 that’s still going strong, but time for a little more juice. Looks like it’ll be the 4011 then instead.
Thanks.
Well, I would rather shield a RB5009, the 4011 is a bit dated, though of course if you are replacing a 2011 it is anyway more than enough, see also:
Just my 2 cents. The rb5009 is an altogether better device. I have deployed both, and both function correctly and reliably - it’s not that one is bad, but that the other is better.
The sort of sheet metal used for these cases doesn’t really provide much in terms of RF immunity, and I wouldn’t be the least bit surprised if actually the rb5009 had better tolerance if tested.
Just a note: for the rb5009, the rack mount ears have to be ordered separately. Not that they are especially expensive, but it can be a bit infuriating to forget it.
Don’t forget about L009, it’s made as pure replacement for 2011 so it should get you more power already.
And exact same case as RB5009 (but different color). You can easily combine both using the same rackmounts (I have a setup like that at home with RB5009 as router and L009 as switch).
But obviously as a router, RB5009 is at a different level 
I bit the bullet and bought an RB4011 ethernet router, and an RB4011 with the 5GHz and 2GHz wireless interfaces. So far, the wireless one is a headache…
First thing I did was update the firmware to 7.19.3 stable - the latest release.
The 5GHz Wifi is enabled and working as an AP and broadcasting SSID. However, DHCP server says “–interface not running”. This is via Wireless. I tried installing Wifi-qcom-ac and configure the 5GHz interface in there instead as an AP. The interface is enbled and broadcasting the SSID.
Changed DHCP server interface from wlan2 to wifi, and it reports the same, “–interface not running”. Except the Interface IS running. Something up with the firmware by the looks of it straight off the bat.
If I setup devices to use static IP and settings etc… I can connect to the 5 GHz card and the Internet is working fine. If I switch to DHCP, it won’t work. As the RB4011 reports under DHCP “–interface not running”, which it IS! I’ve set the pools and DHCP network settings. But it keeps saying interface not running.
Post your configuration, it could be both a small detail in it preventing the DHCP to work correctly or some hardware issue, so you’d better find out whta the problem is before the time for returning it expires.
Since the RB4011 is an oldish device it is unlikely that the 7.19.3 has been tested extensively on it, so it could be also an OS version issue, personalIy I would downgrade to 7.14.3, which is a release that has been around since a bit of time and reported as having wi-fi working fine (while later releases have been reported as giving issues with roaming/disconnections).
There’s a general “thing” about wireless AP interfaces: they aren’t considered running (the “R” flag in the interface display) until they have at least one client connected to them. (They of course still broadcast their SSID and are ready to receive a connection, it’s just that they’re not considered “running”.)
This has all sorts of side-effects. You are experiencing one of them. One way to solve it is to create a bridge with only the wifi interface in it, and put the dhcp server on the bridge (bridges are always considered running)
The other is that this behavior of wireless interfaces can be altered with the command
/interface wifi set disable-running-check=yes
(The same command exists both for the legacy wireless and the new wifi-qcom* interfaces.)
Note: This is one of the few config options only available via the CLI.
1 Like
Quick update. I finally got it running. I had to manually connect to the interface from a wi-fi client, using manual/static IP settings. With a client connected, it finally came up as running. After re-connecting the client with DHCP this time, the DHCP Server is now running. It appears it was bad default behavior not allowing the DHCP Server to set until a client was already connected to the interface manually.
No.
As explained by lurker888, this is standard behavior in ROS.
Wifi interfaces are not “running” until a client is connected to it.
lurker888 gave you 2 options to circumvent this so-called “problem”.
Cool this is good to know, thanks.
The DHCP Server allowed me to set it for the 2.4 GHz interface without a client connected.
Same with my Hap ax2, and older RB951Ui (not using bridge mode).
This is the first time it hasn’t and only on that interface, a little inconsistent.
The 4011 certainly smokes the 2011. 1-3% CPU @ 100 Mbps. It was about 73-80% on the 2011 @ 100 Mbps and around 7% @ 16 Mbps with a simple queue. When fast.com went from down to upstream test (measuring ping), it would always max out the CPU even slowing down the bandwidth to 16 Mbps.
Happy days. Thanks fella’s.