Yes, I hope they provide a solution quickly. Movistar TV in Spain is pixelated.
Disabling a entry in /ip/dns/addlist don’t freeup dns cache size. I’m running rc1 on two rb5009.
[admmikrotik@router70] /ip/dns> adlist/print
Flags: X - disabled
0 url="https://raw.githubusercontent.com/StevenBlack/hosts/master/hosts" ssl-verify=no match-count=317 name-count=161549
[admmikrotik@router70] /ip/dns> print
....
cache-size: 40960KiB
cache-used: 18286KiB
....
[admmikrotik@router70] /ip/dns> adlist/disable numbers=0
[admmikrotik@router70] /ip/dns> adlist/print
Flags: X - disabled
0 X url="https://raw.githubusercontent.com/StevenBlack/hosts/master/hosts" ssl-verify=no match-count=317 name-count=161549
[admmikrotik@router70] /ip/dns> print
....
cache-size: 40960KiB
cache-used: 18284KiB
....
Did your BGP config survive during update? Do you have multiple routing tables?
On mine, the BGP config was lost together with multiple routing table config.
VLANs stopped working in 7.15, but now work again in 7.16rc1, however:
The /import of scripts that used to work fine for years now fail! It seems, due to line continuations not being respected.
If a necessary argument is on a subsequent line, the import fails. For example, I used to specify:
/interface wireless cap
set bridge=bridge-boss discovery-interfaces=bridge-boss enabled=yes \
interfaces=wlan2,wlan1
and that now fails; I have to change it to (this, oddly, doesn’t happen on all line continuations, only some of them):
/interface wireless cap
set bridge=bridge-boss discovery-interfaces=bridge-boss enabled=yes interfaces=wlan2,wlan1
This is only one example; any multi-line scripts fail on import, eg.:
#line 34..35
add add-default-route=special-classless default-route-distance=88 interface=\
vlan88-backhaul script=":if ( \$bound != 0 ) do={\r\
#line 36
\n # Attempt DNS update, in case Internet is available now\r\
Script Error: expected command name (line 36 column 5)
Yes and yes, but just 3 tables.
Inconsistency on syntax of RouterOS it is a pain in the…
Ex:
/routing/bgp/connection/print where address-families → ip, ipv6, vpnv4, vpnv6
/routint/route/print where afi → ip, ipv6, vpv4, vpn6
Another thing that is getting blurry for me is the concepts of RIB and FIB in RouterOS.
It is clear that /ip/route/ and /ipv6/route/ are the same thing, just with a hidden address-family filter.
In winbox, by counting routes, this becomes even more evident.
And this same table seems to be the /routing/route/ table.
But how can I print the differences between RIB and FIB?
Isn’t there the concept of RIB-ONLY?
The Intel AX adapters are hot garbage. Last week I literally had a laptop with an AX200 card sitting next to one with a Realtek 8852CE. The Realtek was connecting to my cAP ax in the next room at 1201 Mbs (max rate) and the AX200 was connecting at 300 Mbs. It also randomly decides to connect to 2.4Ghz on a far AP even when sitting underneath a cAP ax, which requires turning WiFi off and back on to fix. This is in Windows 11 23H2 connecting to cAP ax units running ROS 7.15.1.
I do not blame Mikrotik for this, as I have a ton of Wireless devices (Apple, Android, Google Home, Nvidia Shield, Chromecasts, etc) and it’s ONLY the devices with AX cards that act weird.
We have noticed that all complaints relating to client connection to mikrotik 802.11ax AP are related with these intel chips. It seems that there is a known bug in the linux kernel, specifically relating to this intel chip. Like I said somewhere else in the forum, we are investigating if we can somehow work around this known issue
Only active routes are in the FIB.
Well, it is probably not just a Linux issue with these Intel AX cards. Kaldek is running Windows. It is more likely their Intel device firmware (firmware is proprietary binary regardless of OS/platform AFAIK)
I have ZERO issues with ax200/210 cards on Windows 10 or 11 with the latest drivers. The only problem I ever had was with the Known sleep bug which was fixed a fair few iterations back. Of course I am using 7.16RC1 not 7.15. something…
The ax200 card connected to the cap-wifi1 is in another room with the door closed and still getting 1201 up down with internal antennas.
I was talking about linux kernel in the mikrotik AP, not the client side.
Is there any known reason why Routing->Tables and Routing->BGP configuration is lost on upgrade?
This was also reported earlier in beta by someone else.
Please do not release before this is solved, risk of lock-out of remote routers!
hum, 
of using dated and EOL 5.5.x kernel.
Here in Europe (Germany) I’ve had an issue with all AX2XX and BE200 Cards ive owned with any channel above 124.
Above Channel 124 they only want to connect to 5GHz wifi if the channel width is a max of 20mhz.
Otherwise they connect with limited wifi speed.
I’ve had similar issues running a Qualcomm WCN685x. Which I think is an AMD Version.
The AMD RZ600 Qualcomm worked fine tho.
And a realtek card worked fine aswell.
Other devices dont seem to have any issues either. (specifically Phones with AX)
Back then I opened a topic on Intel forum where I reported most of my findings.
https://community.intel.com/t5/Wireless/AX210-and-BE200-only-5Ghz-AX-with-20mhz-Channel-width/m-p/1550533#M51074
P.s.: I tried 3 different intel Cards with 3 different APs.
Mikrotik hap AX3, AX2 and ZTE MF281 (only supports AC) which is rebranded by Deutsche Telekom
It may be related to incorrect country/region specific information.
There are several manufacturers where updating that info is a mess.
Probably the feeds of correct information are not that clear or are sometimes in error.
Not really true, since we apply lots and lots of patches
The more patches you need, the more complicated it becomes (and of course it needs much more time) to upgrade the kernel.
Not sure if that’s good or bad.
risky strategy.. https://youtu.be/_yWhsynnxEg?t=1211