Welldone!!!*) bgp - copy all well-known and optional transitive attributes for BGP VPNv4 (introduced in v7.9beta4);
*) netinstall-cli - improved device reinstall on failed attempt;
*) ssh - added support for Ed25519 key export and import in PKCS8 format;
*) timezone - updated timezone information from "tzdata2023c" release;
I'm not sure this is related to 7.9rc, however HotSpot is not designed for multigateway configuration.Having trouble with the Hotspot Login Page in PCC Load Balancing which often fails to appear or doesn't automatically redirect?
*) ssh - added support for Ed25519 key export and import in PKCS8 format;
ssh-keygen -t ed25519 -m PKCS8 -f my.key
Very strange, I never seen it before.I had it too on AX3 and for me this was solved with 7.9rc1. See couple of posts up where the reference is of sup ticket.
No more error after reboot.
yepp, same on beta4 and rc1well, please don't fix SNMP anymore :(
LibreNMS could not read port MAC address
ifPhysAddress
SNMP['/usr/bin/snmpbulkwalk' '-v2c' '-c' 'COMMUNITY' '-OQUst' '-m' 'IF-MIB' '-M' '/opt/librenms/mibs:/opt/librenms/mibs/mikrotik' 'udp:HOSTNAME:161' 'ifPhysAddress']
ifPhysAddress.1 = 0:0:0:0:0:0
ifPhysAddress.2 = 0:0:0:0:0:0
ifPhysAddress.3 = 0:0:0:0:0:0
ifPhysAddress.4 = 0:0:0:0:0:0
ifPhysAddress.5 = 0:0:0:0:0:0
I've got over 9000mW, thanks!For increase the TX power over regulatory limits, just alter this file and save it on
[...]
I've got over 9000mW, thanks!For increase the TX power over regulatory limits, just alter this file and save it on
[...]
APRILS FOOOOOOLS !!!! xDFor increase the TX power over regulatory limits, just alter this file and save it on
\lib\modules\5.6.3\kernel\net\wireless
where is also present cfg80211.ko
on flash device must be put on
\flash\lib\modules\5.6.3\kernel\net\wireless
@own3r1138 😂2023-04-01_06-27-06.png
it's in stock image :lol:nice one, but the utf-8 part gave you away (;
Thanks!
Still some issues with ssh keys:
*) ssh - added support for Ed25519 key export and import in PKCS8 format;
Export of hostkey in ed25519 works, reimport of the same key or other ed25519 keys will not import (tested on CHR(x86) platform):
ssh-keygen
$ node sshpk/bin/sshpk-conv --informat=pkcs8 --file=router_ed25519.pem --outformat=ssh --out=router_ed25519.pub --comment "admin@router"
$ file router_ed25519.pub
router_ed25519.pub: OpenSSH ED25519 public key
$
This is likely just lower rate limiting threshold. And can be ignored. I've seen the same on Juniper and Huawei, aggressive ICMP rate limiting.7.9rc1 on CCR2216 has an higher icmp ping loss compared to 7.8 and previews.
7.9rc1 respond with ~3x1000 loss
same hardware, same configuration but with 7.8 respond with ~3x10000 loss
just an order of magnitude higher.
Does*) ssh - added support for Ed25519 key export and import in PKCS8 format;
/user/ssh-keys/import
I agree with you, but I would like to controll this behavior to handle some corner case issues.This is likely just lower rate limiting threshold. And can be ignored. I've seen the same on Juniper and Huawei, aggressive ICMP rate limiting.7.9rc1 on CCR2216 has an higher icmp ping loss compared to 7.8 and previews.
7.9rc1 respond with ~3x1000 loss
same hardware, same configuration but with 7.8 respond with ~3x10000 loss
just an order of magnitude higher.
Looks like I'll stay on 7.8 for all production hardware, even in my home. Until 7.9 is stable.
And RC should never be run in production. I would say stay away from 7.x.0 and wait for 7.x.1 or higher. But it seems that MT like to add new function compare to make it stable. I was hoping on 7.8.1 be fore 7.9beta/rc, or at lest backport fixes broken in 7.8 to make 7.8.1 and try to reach the first long time release for 7.xLooks like I'll stay on 7.8 for all production hardware, even in my home. Until 7.9 is stable.
Well, this is a "release candidate" and MT had occasions where more than one RC happened before corresponding "full" release. Meaning that even MT considers this more a lab testing release than for production environment. Surely it's up to anyone to install rc or even beta in their production networks if they decide so.
*) ssh - added support for Ed25519 key export and import in PKCS8 format;
Ignore the first part of the above!Might be safer to provide a mechanism to export just the host public key (or encrypt the hosts keys with a passphrase).
ssh-keyscan router
V6 version is normalI'm not sure this is related to 7.9rc, however HotSpot is not designed for multigateway configuration.Having trouble with the Hotspot Login Page in PCC Load Balancing which often fails to appear or doesn't automatically redirect?
Don't push them too hard, they'll have the last laugh and tell us it will be in version 8!ipv6 fasttrack support is late
please mikrotik team
Copy & Paste???Tx Power for 2.4 GHz interface on hAP ax2 is limited to 15 dBm. Tested with different Country configured.
Moreover, right after changing a country 20 dBm is displayed for a while and then it falls to 15 dBm.
ipv6 was introduced 27 years ago. most competitors have fast-track support now.
Hmmm...:Non. Currently ed25519 keys for public key authentication are not supported.
*) ssh - added support for Ed25519 key export and import in PKCS8 format;
So they only added Ed25519 for "host keys", not for authentication? That's unfortunate...*) ssh - added Ed25519 host key support;
/int pri where .id=*8
Ah OK thanks rextendedWhat characters? "%*8"? Is interface with .id = 8
on terminal, for know the name:The bug is not the character, but the missing convertion from .id to name.Code: Select all/int pri where .id=*8
Please also fix the missing password parameter on /user ssh-keys private import private-key-file=xxx user=xxxAn additional fix for the "/ip/ssh/import-host-key" function will be available in rc2.
Yes, you copy-pasted your assumption as a reply to the reported and still-existing bug.Copy & Paste???Tx Power for 2.4 GHz interface on hAP ax2 is limited to 15 dBm. Tested with different Country configured.
Moreover, right after changing a country 20 dBm is displayed for a while and then it falls to 15 dBm.
viewtopic.php?p=993395#p993395
Same reply:
Is correct, 15 + 4,5 ~20dB...
You do not read the specs... or if you read it, you do not understand what you read, and this is not one assumption, is a fact.Yes, you copy-pasted your assumption as a reply to the reported and still-existing bug.
15:50:13 vrrp,info vrrp-security-dns now MASTER, master down timer
15:50:13 vrrp,info vrrp-security-dns now BACKUP (set by VRRP Group)
...
15:50:17 vrrp,info vrrp-security-dns now MASTER, master down timer
15:50:17 vrrp,info vrrp-security-dns now BACKUP (set by VRRP Group)
[admin@router-backup] /interface/vrrp> print
Flags: X, I - INVALID; B, F - FAILURE
Columns: NAME, INTERFACE, MAC-ADDRESS, GROUP-MASTER, VRID, PRIORITY, INTERVAL, VERSION, V3-PROTOCOL, SYNC-CONNECTION-TRACKING
# NAME INTERFACE MAC-ADDRESS GROUP-MASTER VRID PRIORITY INTERVAL VERSION V3-PROTOCOL SYNC-CONNECTION-TRACKING
...
12 B vrrp-mgmt-dns vlan-mgmt 00:00:5E:00:01:35 self 53 25 1s 2 ipv4 no
...
15 B vrrp-security-dns vlan-security 00:00:5E:00:01:35 vrrp-mgmt-dns 53 25 1s 2 ipv4 no
...
Same issue on my routers since 7.9beta4. I created SUP-112104 on Mar29. still "Waiting for support" status.With 7.9rc1 I suddenly see VRRP announcement traffic on interfaces that have a group-master (which is not themselves).
This is causing rapid address flapping
The monitor command will report transmitter (radio) power. As rextended pointed out, 15dBm is the appropriate transmitter power to use for an interface with 5 dB (rounded up) antenna gain, when operating on a channel, where regulatory requirements limit EIRP to 20dBm.Tx Power for 2.4 GHz interface on hAP ax2 is limited to 15 dBm. Tested with different Country configured.
Moreover, right after changing a country 20 dBm is displayed for a while and then it falls to 15 dBm.
Same thing on hap ac3. May be it connected with Wave2 packet (so no /interface/wireless more)Already know, already reported from me two times on another two rotueros version, and the exact procedure for fix that error, never fixed.
Then I broke down to rewrite it every time, they didn't care anyway.
On 7.5
viewtopic.php?p=955204#p955204
On 7.6
viewtopic.php?p=962556#p962762
But the MikroTik Staff never gave a damn about it.
I have a XPON module that works with 7.6, it still doesn't work in 7.9rc1*) sfp - improved system stability with some SFP GPON modules on RB4011, RB5009, CCR2004-1G-12S+2XS, CCR2004-16G-2S+, CCR2116-12G-4S+, CCR2216-1G-12XS-2XQ devices;
Hello, please open a support ticket regarding this.I’ve just found one possible bug, when activating ft=yes old iPads Air 2 running iPadOS 15.7.4 could not connect anymore (activating ft-over-ds do not help neither), always requesting for password and telling bad password. In my case valid authentication are wpa2-psk and wpa3-psk only (checked also with wpa-psk unsucessfully).
Yessss..........*) ssh - added support for Ed25519 key export and import in PKCS8 format;
is it possible to have superchannel where we will be able to set all parameters without any restictions?The monitor command will report transmitter (radio) power. As rextended pointed out, 15dBm is the appropriate transmitter power to use for an interface with 5 dB (rounded up) antenna gain, when operating on a channel, where regulatory requirements limit EIRP to 20dBm.Tx Power for 2.4 GHz interface on hAP ax2 is limited to 15 dBm. Tested with different Country configured.
Moreover, right after changing a country 20 dBm is displayed for a while and then it falls to 15 dBm.
The initial reading of 20dBm is erroneous. As the interface is starting up, the actual power figure is not yet available and the reading falls back to max regulatory EIRP.
if you want to import the key in interactive mode then for security reasons passphrase is asked when you execute the command.Yessss..........*) ssh - added support for Ed25519 key export and import in PKCS8 format;
Please also fix the missing password parameter on /user ssh-keys private import private-key-file=xxx user=xxx
Someone lost it along the way
*) snmp - fixed several OIDs that were returning empty values (introduced in v7.9beta4);
ifPhysAddress.1 = 6c:3b:6b:c2:37:a4
ifPhysAddress.2 = 6c:3b:6b:c2:37:a5
ifPhysAddress.3 = 6c:3b:6b:c2:37:a6
ifPhysAddress.4 = 6c:3b:6b:c2:37:a7
ifPhysAddress.5 = 6c:3b:6b:c2:37:a8
Send this link to support...............Yes and no.
Yes as in: I don't see it anymore either (already since rc1)
No as in: support confirmed me today they still haven't found the root cause of the problem (so inherently it is still there). So my ticket stays open.
Accidentally the already converted block was not deleted, old code and unused chunks were left from line 983 to line 995.
I'm very interested in this too, anyone have tried it?Is ovpn still unstable like in 7.8 or is that already fixed?
Done :)Send this link to support...............
viewtopic.php?t=194993#p993491
I'm very interested in this too, anyone tried it?Is OVPN still unstable like in 7.8 or is that already fixed?
Wow! What is Mikrotik waiting for? This is an important bug, it should have been fixed right away!
I'm very interested in this too, anyone tried it?
No, it has not been fixed. Instant crashes due to high CPU utilization.
I don't think it's so difficult to replicate... it happens to me in several routers. It happens when a tunnel gets disconnected for whatever reason, and then the router creates a new one (when there is a static -in interface, it creates a new dynamic one) but the process for the old tunnel seems to keep running in the background, maxing out a core of the cpu.@MT
We are not able to track this problem down/repeat it in our internal tests.
@rextended is this related to the WiFi error that happens on Audience with wifiwave2 confirmed also by @mkx?Send this link to support...............
viewtopic.php?t=194993#p993491
20dBm for all countries? No.The monitor command will report transmitter (radio) power. As rextended pointed out, 15dBm is the appropriate transmitter power to use for an interface with 5 dB (rounded up) antenna gain, when operating on a channel, where regulatory requirements limit EIRP to 20dBm.Tx Power for 2.4 GHz interface on hAP ax2 is limited to 15 dBm. Tested with different Country configured.
Moreover, right after changing a country 20 dBm is displayed for a while and then it falls to 15 dBm.
The initial reading of 20dBm is erroneous. As the interface is starting up, the actual power figure is not yet available and the reading falls back to max regulatory EIRP.
Regdb is filled up from such sources, but it could be outdated, for sure.kernel.org or any other source do not have any value at all.
Panama is 30dBm
The unique official source https://www.asep.gob.pa/?page_id=13116 Page 131
After few hours of uptime all wireless devices were losing ability to connect, to both 2.4 & 5GHz. Disabling WPA3, tripping wireless interface did nothing, only reboot helped. After second occurrence downgraded to 7.8.*) wifiwave2 - fixed key handshake timeout for re-associating client devices on 802.11ac interfaces;
Probably they use a version prior to this change (change that is now outdated and wrong based on your document).
Regdb is filled up from such sources, but it could be outdated, for sure.
Okay, let's take a look at https://www.asep.gob.pa/?page_id=13116 Page 131: it says 1W (30dBm) for the whole range from 2.4GHz to 2.4835GHz.
While 1ch is 2412 MHz (2401–2423) and 13ch is 2472 MHz (2461–2483), and taking into account the previous line - wifiwave2 says "no supported channels". What, Mikrotik?!
Wifi regulatory domain should be definitely fixed.
I’m experiencing the same issue on a Hap ax2, I didn’t find any solution except a reboot to solve the issue…hAP ax3. Funny thing that exactly on 7.9rc1 (2023-Mar-30 16:42) I first acquainted with "key handshake timeout" messages in my log:After few hours of uptime all wireless devices were losing ability to connect, to both 2.4 & 5GHz. Disabling WPA3, tripping wireless interface did nothing, only reboot helped. After second occurrence downgraded to 7.8.*) wifiwave2 - fixed key handshake timeout for re-associating client devices on 802.11ac interfaces;
You can control it via IP>Settings>ICMP RateI agree with you, but I would like to controll this behavior to handle some corner case issues.
regards
too easy... it is just now unlimitedYou can control it via IP>Settings>ICMP RateI agree with you, but I would like to controll this behavior to handle some corner case issues.
regards
/interface wifiwave2
set wifi1 channel=ch-2ghz configuration=common-conf disabled=no
set wifi2 channel=ch-5ghz configuration=common-conf disabled=no
/interface wifiwave2
set wifi1 channel=ch-5ghz configuration=common-conf disabled=no
set wifi2 channel=ch-2ghz configuration=common-conf disabled=no
I opened a support ticket for it SUP-113114, yes of course I tested to remove the network from the saved network list, but it didn’t help.
Hello, please open a support ticket regarding this.
Have you tried removing the network from the list of saved wifi networks on the iPad and adding it again after enabling ft?
'Adaptive' 802.11r appears to be a proprietary behaviour implemented to accomodate clients which get confused by 802.11r. The iPad should not require it.
nevermind. setting 10 minutes for tcp-established and sip-timeout works just fine. I guess the problem was with longer timeouts.i suggest setting default (factory) value of SIP TIMEOUT to 3 minutes (by default). hope we can see it in next releases.
Most routers use 3 minutes udp session timeout. when SIP timeout is higher than 3 minutes, keep alive packets or reconnections will not happen. and this will make your phone number unavailable.
tbh 10 seconds value will work even better if you add connection tracking timeout to SIP ALG. for example, i want SIP connection to stay in connection tracking table for 1 hours, and set sip timeout to 10 seconds so devices will send keepalive/invite etc each and every ten seconds.
thanks for reading , regards
I agreeNo, it needs to be specified that particular config is for non-AX devices.
And then a note indicating order of interfaces is swapped for AX.
Or the other way around.
But not simply change because then it will be wrong for AC3, Audience, ...
Still puzzles me why they changed that order. Much easier to keep it the same.
Confirmed :-/Hi,
tool ip-scan not work on 7.9
regards.
Hi.7.9rc2 working on ax2 without any issue
I have none of your problems, I have one hAP ax3 and two hAP ax2 with capsman which work perfectly.Longer version
Hap AX3 has issues with 7.9rc1&rc2 with capsman. Datapath fails to dynamically add to bridge on remote APs, and comms go nowhere unless the wifi interface is manually added to the bridge port, and bridges uplink (trunk) port is set to disable hardware offloading. Manually bridging wifi/virtual ap to bridge creates issues also. If the CAPsMAN configuration is updated, the remote CAPs will re-create their wifi interface names using the next available wifi numbers. This means you have to go back and add the new interface names to the bridge. Rebooting the remote CAP will result in it getting its original wifi names, and the bridge ports working again. => Dynamic bridge port assignment not working on remote CAPs, If CAPsMAN is running on an AP, that AP will work with dynamic binding.
I also have the problem of 'incorrect password' after a period of time. Phones (Samsung) and Laptop (Win11) all show disconnected with invalid password. Re-entering the credentials (WPA2) does not correct the problem. Only solution is to reboot the CAPsMAN AP and then the remote APs. Once done, systems can re-connect (Win11 automatically, Samsung Phone you have to tap the SSID in the WiFi list and then it will attempt to re-connect).
I also have an Audience, and the firmware does not support vlan assignment on it. Hence it cannot be part of the CAPsMAN network at present.
So, in summary: 7.9rc2 lack of dynamic datapath bridge binding makes CAPsMAN unworkable (Any changes to CAPsMAN config result in remote APs regenerating their interface names and hence no longer being bridged to the network). 'Incorrect password' issue after a period of time, with a reboot of APs being the only solution.
I bit of work is needed before I can have a stable WiFi with FT roaming across AX3 unfortunately. I do not know if Mikrotik will ever support vlan assignment on the Audience, and I may have to drop it from the network (If it can't be managed by CAPsMAN it cannot participate in FT roaming).
Please also fix the missing password parameter on /user ssh-keys private import private-key-file=xxx user=xxx
Someone lost it along the way
Already know, already reported from me two times on another two rotueros version, and the exact procedure for fix that error, never fixed.
Then I broke down to rewrite it every time, they didn't care anyway.
On 7.5
viewtopic.php?p=955204#p955204
On 7.6
viewtopic.php?p=962556#p962762
But the MikroTik Staff never gave a damn about it.
I have none of your problems, I have one hAP ax3 and two hAP ax2 with capsman which work perfectly.
You probably misconfigured capsman wave 2, I don't use dynamic interfaces for wifi.
Are you seriously?What's new in 7.9rc3 (2023-Apr-12 15:53):
*) tools - fixed "ip-scan" (introduced in v7.9beta4);
*) user-manager - fixed process startup after booting (introduced in v7.9beta4);
not "containers", but particular ones imported that are OCI schema based. This is due to change in dockerhub.
Are you seriously?
containers are broken, but there is no fix.
importing remote image: adguard/adguardhome, tag: latest
was unable to import, container 71e2cc5c-abf7-415c-b63a-45036240aa61
this problem was already in 7.8beta, why not fix it in 7.9beta?not "containers", but particular ones imported that are OCI schema based. This is due to change in dockerhub.
Multiple workarounds are already posted in forums, proper "fix" will come with next beta version.
This is due to change in dockerhub.
Multiple workarounds are already posted in forums
$ docker pull adguard-home:latest
$ docker save adguard/adguardhome:latest -o agh-broken.tar
$ brew install skopeo # or apt, or dnf, or…
$ skopeo copy docker-archive:agh-broken.tar docker-archive:agh-fixed.tar
$ scp agh-fixed.tar myrouter:
$ ssh myrouter
> /container/add file=agh-fixed.tar interface=veth1 …plus any other parameters that please you
Fair enough I didn't know this yet.as in any other RouterOS menu - items has to be printed before they can be accessed by using their list number.
seems to be solved with RC4There is a bug with Radius Server and User Manager causes Radius Timeout error for clients trying to connect.
mtxrInterfaceStats SNMP subtree now shows some wild values for interfaces:What's new in 7.9rc2 (2023-Apr-05 13:56):
*) snmp - fixed several OIDs that were returning empty values (introduced in v7.9beta4);
MIKROTIK-MIB::mtxrInterfaceStatsRxBroadcast.1 = Counter64: 57407172828463104
MIKROTIK-MIB::mtxrInterfaceStatsRxBroadcast.2 = Counter64: 28725071988129792
MIKROTIK-MIB::mtxrInterfaceStatsRxPause.1 = Counter64: 13366149
MIKROTIK-MIB::mtxrInterfaceStatsRxPause.2 = Counter64: 6688077
MIKROTIK-MIB::mtxrInterfaceStatsRxMulticast.1 = Counter64: 209053238165504
MIKROTIK-MIB::mtxrInterfaceStatsRxMulticast.2 = Counter64: 7706661682675712
MIKROTIK-MIB::mtxrInterfaceStatsRxFCSError.1 = Counter64: 48674
MIKROTIK-MIB::mtxrInterfaceStatsRxFCSError.2 = Counter64: 1794347
MIKROTIK-MIB::mtxrInterfaceStatsRxAlignError.1 = Counter64: 161838662680576
MIKROTIK-MIB::mtxrInterfaceStatsRxAlignError.2 = Counter64: 244851790577664
MIKROTIK-MIB::mtxrInterfaceStatsTxExcessiveCollision.1 = Counter64: 1679
MIKROTIK-MIB::mtxrInterfaceStatsTxExcessiveCollision.2 = Counter64: 25998
MIKROTIK-MIB::mtxrInterfaceStatsTxMultipleCollision.1 = Counter64: 22605
MIKROTIK-MIB::mtxrInterfaceStatsTxMultipleCollision.2 = Counter64: 9449
MIKROTIK-MIB::mtxrInterfaceStatsTxSingleCollision.1 = Counter64: 5102
MIKROTIK-MIB::mtxrInterfaceStatsTxSingleCollision.2 = Counter64: 18099
MIKROTIK-MIB::mtxrInterfaceStatsTxExcessiveDeferred.1 = Counter64: 1402
MIKROTIK-MIB::mtxrInterfaceStatsTxExcessiveDeferred.2 = Counter64: 3429
MIKROTIK-MIB::mtxrInterfaceStatsTxDeferred.1 = Counter64: 475
MIKROTIK-MIB::mtxrInterfaceStatsTxDeferred.2 = Counter64: 34
MIKROTIK-MIB::mtxrInterfaceStatsTxLateCollision.1 = Counter64: 6303
MIKROTIK-MIB::mtxrInterfaceStatsTxLateCollision.2 = Counter64: 2
Point of the problem is that accidentally partial offline upgrade removes wifiwave2 package and this makes Reset Configuration act wrong. Expected result would be that even without wifiwave2 package, after Reset Configuration router gets default configuration, but without wireless functionality.kalamaja, just use check for updates. it is there for beginners, so situations like this are avoided. manual uploading of files should not be done anymore, unless you are a seasoned professional and part of this forum :)
This feature would be nice to mitigate the removed route cache when upgrading from ROS 6.ipv6 fasttrack support please
I'm facing the same issue on an HAP Ax2, it's only appearing using v7.9 from beta till rc2 (didn't test rc3 until now). I'm using static wifi interfaces with vlans. Do you need a support ticket for this issue ?CAPsMAN AX3/Audience issues with 7.9RC1 and 7.9RC2
- Incorrect password [Using WPA2] occurs after a period of time across multiple devices (Samsung Phones/Win11 etc), only resolved by rebooting APs[/
I also have the problem of 'incorrect password' after a period of time. Phones (Samsung) and Laptop (Win11) all show disconnected with invalid password. Re-entering the credentials (WPA2) does not correct the problem. Only solution is to reboot the CAPsMAN AP and then the remote APs. Once done, systems can re-connect (Win11 automatically, Samsung Phone you have to tap the SSID in the WiFi list and then it will attempt to re-connect).
But the post you are referring to, are about dynamic interface/bridge, not about authentication issues, like stated in posts 114, 97 and 95.See post 117 and 121 above.
It will be fixed in a later version.
MAC_ADDR@wifi1 connected, signal strength -44
MAC_ADDR@@wifi1 reauthenticating
MAC_ADDR@@wifi1 reauthenticating
MAC_ADDR@@wifi1 connected, but was associated
MAC_ADDR@@wifi1 disconnected, key handshake timeout, signal strength -44
loops the above logs
[Some_User@MikroTik] > /interface/wifiwave2/actual-configuration print
0 name="wifi1" mac-address=MAC-ADDR arp-timeout=auto radio-mac=MAC-ADDR
configuration.mode=ap .ssid="Some_SSID" .country=Some_Country
security.authentication-types=wpa2-psk .encryption=ccmp .passphrase="Some-PASS" .disable-pmkid=yes
channel.frequency=5180 .band=5ghz-n .width=20mhz .skip-dfs-channels=all
Not saying it shouldn't work, per se. But any reason to change the band? e.g. the default, .band=5ghz-axCode: Select all.band=5ghz-n
same issue with ax or any other band. Was set to n to replicate the known working config of the hap ac on ros 6.xNot saying it shouldn't work, per se. But any reason to change the band? e.g. the default, .band=5ghz-axCode: Select all.band=5ghz-n
same issue with the same exact log output *disconnected, key handshake timeout*Does it happen in stable v7.8?
Now...there is a lot of functionality that's missing in wifiwave2...so very well may not work in v7.8.
Same issue. Only by enabling wpa-psk it connects immediately ...What if you disable ccmp ?
Just checked, I don't have it enabled on my setup (wpa-psk2 and psk3).
:global sqrt do={
:for i from=0 to=$1 do={
:if (i * i > $1) do={ :return ($i - 1) }
}
}
:global addNatRules do={
/ip firewall nat add chain=srcnat action=jump jump-target=xxx \
src-address="$($srcStart)-$($srcStart + $count - 1)"
:local x [$sqrt $count]
:local y $x
:if ($x * $x = $count) do={ :set y ($x + 1) }
:for i from=0 to=$x do={
/ip firewall nat add chain=xxx action=jump jump-target="xxx-$($i)" \
src-address="$($srcStart + ($x * $i))-$($srcStart + ($x * ($i + 1) - 1))"
}
:for i from=0 to=($count - 1) do={
:local prange "$($portStart + ($i * $portsPerAddr))-$($portStart + (($i + 1) * $portsPerAddr) - 1)"
/ip firewall nat add chain="xxx-$($i / $x)" action=src-nat protocol=tcp src-address=($srcStart + $i) \
to-address=$toAddr to-ports=$prange
/ip firewall nat add chain="xxx-$($i / $x)" action=src-nat protocol=udp src-address=($srcStart + $i) \
to-address=$toAddr to-ports=$prange
}
}
Likely worth mentioning specifically... but ROSE now seems to mount SMB in MacOS 13.3 in v7.9rc. That was not working in v7.8. Thanks!*) rose-storage - various stability fixes;
The wifi interface is only automatically added to the appropriate bridge VLAN, when datapath.vlan-id is configured on the cAP.I use a CHR as CAPsMANv3 with 7.9RC3 and a hAP AX2 as CAP with 3 SSID´s, two of them with VLAN tag (Guest and IoT).
But in none of my configs (I´ve tried several onces) the automatically add VLAN-tagged interface to the appropriate bridge VLAN is working.
Does the issue persist if you set security.disable-pmkid to 'no'?device can't connect with wpa2-psk ccmp while it worked fine with ros 6.x on hap ac same config.
Hi FToms, the VLANs ARE added to the bridge: but I can´t change the VLAN ID in a dynamic interface: How should it work in your opinion?The wifi interface is only automatically added to the appropriate bridge VLAN, when datapath.vlan-id is configured on the cAP.I use a CHR as CAPsMANv3 with 7.9RC3 and a hAP AX2 as CAP with 3 SSID´s, two of them with VLAN tag (Guest and IoT).
But in none of my configs (I´ve tried several onces) the automatically add VLAN-tagged interface to the appropriate bridge VLAN is working.
We are yet to implement this functionality when datapath.vlan-id is only configured on the CAPsMAN.
Yes same issue with or without disable-pmkid set.Does the issue persist if you set security.disable-pmkid to 'no'?device can't connect with wpa2-psk ccmp while it worked fine with ros 6.x on hap ac same config.
If so, please open a support ticket regarding this.
On my CCR2116 with Noctua PWM fans.The fan control settings don't seem to do anything.
Not quite following you, can you elaborate a little?Disable vlan bridging on ap bridge.
That's how it worked for me after feedback from support.
Hi holvoetn,Apologies, bad wording from my side.
On AP device, disable VLAN filtering.
Winbox / Bridge / tab VLAN, uncheck VLAN Filtering.
This worked for me.
But I disabled that workaround and usage of capsman2 and will wait for a next version where this is implemented properly.
2.5Gbps Fixed rate on GPON SFP (Huawei MA5671A) works but PPPoE Scan is not working (shows no entries) was working on 7.7 stable on RB5009.
Yeah, I don't know if the fan controller has some static RPM -> PWM% mapping, but I got some 250-6000 RPM fans and they don't really behave as I would expect them to. The change in RPM is not linear with the temperature target at all and the fan behaviour is very similar to how it was with non PWM fans with the RPM jumping from ~2.7k RPM to 5k RPM when the temp hits 59 (default target of 58). I tried setting a min fan speed percent of 10% and the fans turned off at around 55 degrees. I now have the min fan speed at 40% and they seem to be staying on at 2.7kish.On my CCR2116 with Noctua PWM fans.The fan control settings don't seem to do anything.
I have confirmed that there are some levels of rpm.
0% = 0~3000 rpm (automatic?)
20% = 0~3000 rpm (same to above?)
40% = 3000~3500 rpm
60% = 4000~4500 rpm
100% = 4700~5000 rpm
/routing bgp vpn
add disabled=yes export-route-targets=xxx:1004 import-route-targets=xxx:1005 label-allocation-policy=per-vrf redistribute=connected route-distinguisher=xxx:1004 vrf=vrf-MPLS_TEST
/routing bgp vpn
add disabled=yes export.route-targets=xxx:1004 import.route-targets=xxx:1005 label-allocation-policy=per-vrf route-distinguisher=xxx:1004 vrf=vrf-MPLS_TEST
Confirmed this is indeed broken. DHCP failing for a multitude of devices in my environment with the "DHCP offering lease without success" log messages. It's happening to a multitude of devices, notably:Looks like the DHCP options sequencing issue that broke Google Nest Doorbells has been reintroduced in 7.9. Please resolve!
I'm wondering about the same thing. Could supporting only old and obsolete cipher suites for SSTP be considered as a bug? Probably not :(Any idea if Mikrotik will add support of TLS v1.3 to the SSTP server/client?!
I'm also facing problems with OpenVPN on 7.8, and i really don't believe it's being triggered by high CPU usage. I'm facing that on a CCR1009 which barely exceeds 7-8%, so there's really a lot of spare CPU time and, even then, the problem is happening, just as described on several posts on the 7.8 release. Openvpn interface shows connected but doesn't flow traffic. Restarting the interface doesn't solve the problem, only rebooting the routerboard seems to solve it.No, it has not been fixed. Instant crashes due to high CPU utilization.
The problem happens when a tunnel gets incorrectly disconnected (because the internet connection drops off or whatever reason). The router is not capable of killing the connection (that's why you see the interface still as "connected") and you cannot kill it either. A new dynamic interface is created but the old connection uses a full core from the CPU and, after some variable time, the router gets restarted.I'm also facing problems with OpenVPN on 7.8, and i really don't believe it's being triggered by high CPU usage. I'm facing that on a CCR1009 which barely exceeds 7-8%, so there's really a lot of spare CPU time and, even then, the problem is happening, just as described on several posts on the 7.8 release. Openvpn interface shows connected but doesn't flow traffic. Restarting the interface doesn't solve the problem, only rebooting the routerboard seems to solve it.No, it has not been fixed. Instant crashes due to high CPU utilization.
Haven't tried 7.9rc on the places i'm facing the problem, but not good hearing it's not fixed :(
.What's new in 7.9rc4 (2023-Apr-24 16:34):
[ .... ]
*) ovpn - improved system stability for Tile devices;
[ .... ]
Confirm, fixed.What's new in 7.9rc4 (2023-Apr-24 16:34):
*) snmp - fixed several OIDs that were returning incorrect values (introduced in v7.9beta4);
/interface bridge
add admin-mac=48:A9:8A:48:5D:3F auto-mac=no frame-types=\
admit-only-vlan-tagged name=bridge vlan-filtering=yes
/interface vlan
add interface=bridge name=home vlan-id=77
/interface wifiwave2 channel
add band=2ghz-ax disabled=no frequency=2412 name=2ghz-ch01 width=20mhz
add band=5ghz-ax disabled=no frequency=5180 name=5ghz-ch36 width=20/40/80mhz
/interface wifiwave2 datapath
add bridge=bridge disabled=no name=home vlan-id=77
add bridge=bridge client-isolation=yes disabled=no name=guest vlan-id=87
/interface wifiwave2 security
add authentication-types=wpa2-psk,wpa3-psk disabled=no name=home wps=disable
add authentication-types=wpa2-psk,wpa3-psk disabled=no name=guest wps=disable
/interface wifiwave2 configuration
add country=Spain datapath=guest disabled=no mode=ap name=guest security=\
guest ssid=guest
add country=Spain datapath=home disabled=no mode=ap name=home security=home \
ssid=home
/interface wifiwave2
set [ find default-name=wifi2 ] channel=2ghz-ch01 configuration=home \
configuration.mode=ap .tx-power=15 disabled=no name=wifi-2ghz
add configuration=guest configuration.mode=ap disabled=no mac-address=\
4A:A9:8A:48:5D:45 master-interface=wifi-2ghz name=wifi-2ghz-guest
set [ find default-name=wifi1 ] channel=5ghz-ch36 configuration=home \
configuration.mode=ap disabled=no name=wifi-5ghz
add configuration=guest configuration.mode=ap disabled=no mac-address=\
4A:A9:8A:48:5D:44 master-interface=wifi-5ghz name=wifi-5ghz-guest
/interface bridge port
add bridge=bridge frame-types=admit-only-vlan-tagged interface=ether1
add bridge=bridge frame-types=admit-only-untagged-and-priority-tagged \
interface=ether2 pvid=77
add bridge=bridge frame-types=admit-only-untagged-and-priority-tagged \
interface=ether3 pvid=77
add bridge=bridge frame-types=admit-only-untagged-and-priority-tagged \
interface=ether4 pvid=77
add bridge=bridge frame-types=admit-only-untagged-and-priority-tagged \
interface=ether5 pvid=77
/interface bridge vlan
add bridge=bridge tagged=bridge,ether1 vlan-ids=77
add bridge=bridge tagged=ether1 vlan-ids=87
/ip dhcp-client
add interface=home use-peer-dns=no use-peer-ntp=no
/ip dns
set servers=1.1.1.2,1.0.0.2
With what this "/task" submenu deals with?Other changes since v7.8:
*) console - added "/task" submenu (CLI only);
.*) webfig - added inline comments;
Being a person that uses the webfig much more than winbox, this change:
.
*) webfig - added inline comments;
.
is terrible. I'd love being able to at least choose the old behavior and having the comments on a different line instead of inline. Please consider having this as an option and not fixed inline the way it's now on 7.9rc4. For those who don't have very big screens/resolutions, this is terrible.
Use packet sniffer to get radius traffic and open created file with Wireshark.Radius client configuration with VRF doesn't work, Radius-server and Radius-client is CHR-7.8
/radius/print detail
Flags: X - disabled
0 service=login called-id="" domain="" address=10.20.2.17@VRF-OAM secret="******" authentication-port=1812 accounting-port=1813 timeout=300ms
accounting-backup=yes realm="" src-address=10.20.2.19 protocol=udp certificate=none
ping 10.20.2.17 vrf=VRF-OAM
..............
07:43:53 radius,debug timeout for 0d:5d
radius server is configured and tried check radius server with the user manager user, it is working
can someone please advice
Same here, I love inline comments!! We still need to be able to move the comment column to the right or somewhere else :)Being a person that uses the webfig much more than winbox, this change:
.
*) webfig - added inline comments;
.
is terrible. I'd love being able to at least choose the old behavior and having the comments on a different line instead of inline. Please consider having this as an option and not fixed inline the way it's now on 7.9rc4. For those who don't have very big screens/resolutions, this is terrible.
+1
Thanks for your response, this resolved by uncheck the accounting backupUse packet sniffer to get radius traffic and open created file with Wireshark.Radius client configuration with VRF doesn't work, Radius-server and Radius-client is CHR-7.8
/radius/print detail
Flags: X - disabled
0 service=login called-id="" domain="" address=10.20.2.17@VRF-OAM secret="******" authentication-port=1812 accounting-port=1813 timeout=300ms
accounting-backup=yes realm="" src-address=10.20.2.19 protocol=udp certificate=none
ping 10.20.2.17 vrf=VRF-OAM
..............
07:43:53 radius,debug timeout for 0d:5d
radius server is configured and tried check radius server with the user manager user, it is working
can someone please advice
Use NTRadPing to create requests for testing Radius server.
+100Being a person that uses the webfig much more than winbox, this change:
..*) webfig - added inline comments;
is terrible. I'd love being able to at least choose the old behavior and having the comments on a different line instead of inline. Please consider having this as an option and not fixed inline the way it's now on 7.9rc4. For those who don't have very big screens/resolutions, this is terrible.
impossible to monitor traffic on offlaoded vlan.What are you doing?
First you free up the CPU, then you reuse it to count packets?
....
MT, when will see statistics/counters on offloaded vlan traffic?
I liked the inline comments, but it gets a little harder with longer ones. Maybe a compromise, and insert a line brake if the comment is too long?I'd love being able to at least choose the old behavior and having the comments on a different line instead of inline. Please consider having this as an option and not fixed inline the way it's now on 7.9rc4. For those who don't have very big screens/resolutions, this is terrible.
Obviously not just yet ;-)is time for v7.9 [stable] !!
.What's new in 7.9rc5 (2023-Apr-28 11:52):
[ ... ]
If you read the post for the MT staff, you vill see that RC are only to fix broken stuff. For change in function you need to wait for next version like 7.10 beta.Please review our previous comments on the latest webfig inline-comments ...
{ :local defgw [/ipv6 route get number=[find dst-address="::/0"] value-name=gateway ] ; :put $defgw }
{ :local defgw [/ipv6 route get number=[find dst-address="::/0" and active] value-name=gateway ] ; :put $defgw }
I found 2 bugs.7.9rc4 Regression:
Using RouterOS 7.9rc4 in IP -> ARP list or IP -> DHCP Server -> Leases tab, “unknown” is reported in “Bridge Port” column for all hosts connected to wifi (both Webfig and Winbox).
In Bridge -> Hosts -> Interface column field is empty for wifi hosts.
With RouterOS 7.9rc3 it worked correctly.
Now they call it "inactive" ("invalid" was) and does not count as a SCARY RED state.The v7.7 shows unreachable routes with RED color in the "/ip route" However, the latest 7.9rc5 release doesn't!
Same here, tried 7.9RCs to fix reboots, but after some time wifi got lost. With 7.9RC5 so far no problems.Problems with v7.9rc4 on hAP ax^3. Using the ax^3 as AP only. I have now rolled back to v7.8.
1) The system has crashed and restarted once.
2) Two times all clients registered with the AP have disappeared, and no clients can register or register without an AP reboot.
Can anybody try this on its hAP ax3?Bug #2:
If I open Webfig on hAP ax3 and click on:
Wireless -> Radios → wifi1 (5GHz radio) Webfig hangs, and I need to reload webpage again.
If I click on Wireless -> Radios → wifi2 it works.
Using Winbox it works for both radios.
On hAP ax2 it works for both wifi1 and wifi2 in Webfig and Winbox, so the problem is in hAP ax3.
Please can someone try this on its hAP ax3?
Thanks.
I can repeat this problem on my ax3; wifi1 locks up, wifi2 doesn't.Bug #2:
If I open Webfig on hAP ax3 and click on:
Wireless -> Radios → wifi1 (5GHz radio) Webfig hangs, and I need to reload webpage again.
If I click on Wireless -> Radios → wifi2 it works.
Using Winbox it works for both radios.
On hAP ax2 it works for both wifi1 and wifi2 in Webfig and Winbox, so the problem is in hAP ax3.
Please can someone try this on its hAP ax3?
Thanks.
Thanks. I opened a support ticket for it (SUP-114533).
I can repeat this problem on my ax3; wifi1 locks up, wifi2 doesn't.
What's new in 7.9rc5 (2023-Apr-28 11:52):
*) console - fixed password prompt (introduced in v7.9beta4);
*) lte - improved system stability when changing LTE interface configuration during network scan with MBIM modems (introduced in v7.8 );
*) wifiwave2 - fixed a compatibility issue when using OWE authentication (introduced in v7.8 );
EdPa & strods, Can you provide any feedback on when we might see a proper long term release? Omitting features that are still being developed & debugged would be quite acceptable (even preferred), it's more important to have a bug-free release (at least free of anything you are in any way aware of) than having new functionality that is still being hammered out. These features can be added to a future long term release once they have matured and are no longer being tweaked and fixed every release.If you experience system reboots with this or actually any other RouterOS release, then please send supout files from your routers (from the latest RC and generated after the router has rebooted by itself at least once) to support@mikrotik.com. Users in the forum can not see in supout files the actual reason for the reboots. The correct way how to figure out what caused the reboot is to contact support directly as soon as possible.
/ip/dns/static/add name=a-device.lan type=AAAA address=fd11:1111:1111:1::2 address-list=to_be_filtered_ipv6
/ip/dns/static/add name=a-device.lan type=A address=192.168.1.2 address-list=to_be_filtered_ipv4
/ip/dns/static/add name=another-device.lan type=AAAA address=fd11:1111:1111:1::3 address-list=to_be_filtered_ipv6
/ip/dns/static/add name=another-device.lan type=A address=192.168.1.3 address-list=to_be_filtered_ipv4
/ipv6/firewall/address-list/add list=to_be_filtered_ipv6 address=a-device.lan
/ipv6/firewall/address-list/add list=to_be_filtered_ipv6 address=another-device.lan
/ip/firewall/address-list/add list=to_be_filtered_ipv4 address=a-device.lan
/ip/firewall/address-list/add list=to_be_filtered_ipv4 address=another-device.lan
/ip/firewall/address-list/add list=only_allowed_destination_ipv4 address=allowed-url.com
/ip/firewall/filter/add action=reject chain=forward src-address-list=to_be_filtered_ipv4 dst-address-list=!only_allowed_destination_ipv4
Having a v7 LTS would likely mean that v6 becomes unsupported which (if certain core functionality would be still missing that is actively used on v6) would put v6 (LTS) users who are forced to stay on that in a quite unpleasant situation of having to run production network(s) on unsupported software version. Anyway if one check v6 LTS changelogs than can see that even on LTS versions there are other than improved and fixed changes.... Can you provide any feedback on when we might see a proper long term release? Omitting features ... would be ... preferred ... it's more important to have a bug-free release ...
Does this apply only if both the CAP and the CAPsMAN device are running 7.9? I've got two cAP ax units running 7.9rc5 and my RB5009 is running 7.8 still due to the DHCP issues introduced in 7.9rcX, so I left it on 7.8. I don't have any automated addition to bridges on the CAPs.*) wifiwave2 - automatically add a VLAN-tagged interface to the appropriate bridge VLAN;
I am also suffering from the CCR2216 instability. My supplier asked Mikrotik and their answer was this:on 7.9rc4 and 7.9rc5 running on x86 I see a very good improvment about cpu usage about routing (bgp with multiple fullroutes).
Now I don't see any more one core loaded at 100% for all the time.
Still not the same on CCR2216 where one core is still full about rotuing.
Particularly on CCR2216 I still see a slight packet loss (compleained by online player), packet loss not present on hw-offloaded traffic. So I think there still something to get polished in handling switch-cpu traffic.
regards
If the issue only appears with enabled L3HW, please keep this option disabled for now.
The next RouterOS version 7.10 will include various fixes and improvements for L3HW. When the version will be available, I would appreciate if you could share your feendback. The 7.10 release should be available in few weeks.
many thanks about the feedbackIf the issue only appears with enabled L3HW, please keep this option disabled for now.
The next RouterOS version 7.10 will include various fixes and improvements for L3HW. When the version will be available, I would appreciate if you could share your feendback. The 7.10 release should be available in few weeks.
Works for me...On 7.9rc5 the IPv4 firewall address list does not resolve local static addresses, therefore those can not be referenced for filter rules for example.
[admin@MikroTik] /ip/firewall/address-list> print where list="test"
Flags: D - DYNAMIC
Columns: LIST, ADDRESS, CREATION-TIME
# LIST ADDRESS CREATION-TIME
12 test 7E-AF-4F-3C-B4-CA.dhcp-guest-vpn.MikroTik.eworm.net may/02/2023 11:40:30
;;; 7E-AF-4F-3C-B4-CA.dhcp-guest-vpn.MikroTik.eworm.net
13 D test 192.168.21.250 may/02/2023 11:40:30
Please open a support ticket regarding this and attach a supout file generated when the issue is observable.7.9rc4 Regression:
Using RouterOS 7.9rc4 in IP -> ARP list or IP -> DHCP Server -> Leases tab, “unknown” is reported in “Bridge Port” column for all hosts connected to wifi (both Webfig and Winbox).
In Bridge -> Hosts -> Interface column field is empty for wifi hosts.
I opened a support ticket for it (SUP-114532).Please open a support ticket regarding this and attach a supout file generated when the issue is observable.7.9rc4 Regression:
Using RouterOS 7.9rc4 in IP -> ARP list or IP -> DHCP Server -> Leases tab, “unknown” is reported in “Bridge Port” column for all hosts connected to wifi (both Webfig and Winbox).
In Bridge -> Hosts -> Interface column field is empty for wifi hosts.