Very unstable operation of ospf with md5 auth:

so, how do we do this recursive routing? It was a question here, last post was #604?

Same as before, but now target scope needs to be greater than scope, while in RouterOS v6 you could set target scope to the same as scope. Raise your target-scope by 1 compared to what it was in ROS6 and it should start working.

Thanks. Simple, but in no way obvious.
Regards.

Very unstable operation of ospf with md5 auth:

OSPF Auth is broken in 7.1

viewtopic.php?t=180847

Good news and bad news here. No MetaROUTER in v7. MT [..]

Thanks for your answer, I'll wait until containers show up on a stable release then, to upgrade the RB2011UiAS.

Router OS 7.1 on an RB3011iUAS-RM - OpenVPN in UDP mode will cause random lock ups of VPN clients requiring disconnect/reconnect to pass traffic, this has ben observed from the Android OpenVPN client, Linux OpenVPN package and Windows OpenVPN client (all latest versions)

I also observed this on 7.1beta4 and I believe I posted about it but cant immediately see.


Dec 12 20:06:41 aruma-RB3011-gw <ovpn-orionvpn>: terminating... - peer disconnected
Dec 12 20:06:41 aruma-RB3011-gw orionvpn logged out, 3660 643745 284909 2920 2107 from 45.XXX.XXX.XXX
Dec 12 20:06:41 aruma-RB3011-gw <ovpn-orionvpn>: disconnected
Dec 12 20:06:45 aruma-RB3011-gw connection established from 45.XXX.XXX.XXX, port: 44268
Dec 12 20:06:45 aruma-RB3011-gw recvd P_DATA packet, dropping
Dec 12 20:06:45 aruma-RB3011-gw recvd P_DATA packet, dropping
Dec 12 20:06:45 aruma-RB3011-gw recvd P_DATA packet, dropping
Dec 12 20:06:45 aruma-RB3011-gw : using encoding - AES-256-CBC/SHA1
Dec 12 20:06:45 aruma-RB3011-gw orionvpn logged in, 10.45.XXX.XXX from 45.XXX.XXX.XXX
Dec 12 20:06:45 aruma-RB3011-gw <ovpn-orionvpn>: connected


It can take as little as a two hours to occur or as much as 2 days.


The behaviour does not occur with TCP method set in the Mikrotik OpenVPN Server settings.

Can others confirm similar unreliability?

The performance benefits are good but the unpredictable unreliability makes it unusable.

Kind Regards,
Jim.

error system critical "error while running customized default configuration script: bad command name wireless (line 977 column 25)" is still present in stable.

I've originally renamed default interfaces, dunno if the error makes any harm
Later I've restored original names wifi1/2/3 - the issue is still there.

The original names are wlan1, wlan2 etc.
Indeed when renaming these some issues can occur which they apparently cannot get under control (amazing...)

Router OS 7.1 on an RB3011iUAS-RM - OpenVPN in UDP mode will cause random lock ups of VPN clients requiring disconnect/reconnect to pass traffic, this has ben observed from the Android OpenVPN client, Linux OpenVPN package and Windows OpenVPN client (all latest versions)

I also observed this on 7.1beta4 and I believe I posted about it but cant immediately see.


Dec 12 20:06:41 aruma-RB3011-gw <ovpn-orionvpn>: terminating... - peer disconnected
Dec 12 20:06:41 aruma-RB3011-gw orionvpn logged out, 3660 643745 284909 2920 2107 from 45.XXX.XXX.XXX
Dec 12 20:06:41 aruma-RB3011-gw <ovpn-orionvpn>: disconnected
Dec 12 20:06:45 aruma-RB3011-gw connection established from 45.XXX.XXX.XXX, port: 44268
Dec 12 20:06:45 aruma-RB3011-gw recvd P_DATA packet, dropping
Dec 12 20:06:45 aruma-RB3011-gw recvd P_DATA packet, dropping
Dec 12 20:06:45 aruma-RB3011-gw recvd P_DATA packet, dropping
Dec 12 20:06:45 aruma-RB3011-gw : using encoding - AES-256-CBC/SHA1
Dec 12 20:06:45 aruma-RB3011-gw orionvpn logged in, 10.45.XXX.XXX from 45.XXX.XXX.XXX
Dec 12 20:06:45 aruma-RB3011-gw <ovpn-orionvpn>: connected


It can take as little as a two hours to occur or as much as 2 days.


The behaviour does not occur with TCP method set in the Mikrotik OpenVPN Server settings.

Can others confirm similar unreliability?

The performance benefits are good but the unpredictable unreliability makes it unusable.

Kind Regards,
Jim.

HI, for me Openvpn over UDP is completely broken. Tested yesterday between an CCR1009 (server) and RB4011 (client) and the moment the connection is established both router would reboot. Changing back to TCP solved the issue.

Let's hope this would be fixed in coming versions.

Regards.

I think v7.1.1 and 7.2(testing) and 6.50(testing) are gonna be fun changelogs to read once they arrive.

what is going on with IP Cloud in 7.1? I can see 2 different dns-names in Winbox and CLI, when hitting F5. One is the actual SN of the device, like it always was. The other one looks like this 455a01d2eaf030.sn.mynetname.net. The problem here is that it's always random which of these dns-names are getting updated with the actual ipv4 and ipv6 information. So I can't just use one of these names in my domain dns settings, because there's a 50% chance that it will hold outdated ip addresses after I get a new dynamic one from my ISP.

I think v7.1.1 and 7.2(testing) and 6.50(testing) are gonna be fun changelogs to read once they arrive.

indeed! they can also mark 7.1.1 as "very stable" and 7.1.2 as "extremely stable"

Tell me who tested the speeds through WireGuard and OpenVPN UDP protocol? write down the speed that you got?

WireGuard over LTE (~40Mbps down/~20Mbps up, 31ms ping)
WireGuard over Fibre (~900Mbps down/~465Mbps up, 3ms ping)

OpenVPN over LTE (~38Mbps down/~20Mbps up, 34ms ping)
OpenVPN over Fibre (~870Mbps down/~435Mbps up, 5ms ping)

OpenVPN is about the same for me, no noticeable difference for what i'm doing (personally prefer WireGuard)

Since what version OpenVPN supports UDP protocol ?

Since what version OpenVPN supports UDP protocol ?

7.0 beta3

Edit:
changelog:

Release 7.0beta3 --- 2019-10-23
## General information

+) Based on Kernel 4.14.131
+) New CLI style which is more similar to API commands (v6 commands still supported)
+) OpenVPN UDP protocol support added
+) New NTP client and server implementation, both now included in RouterOS main package
+) removed individual packages, only bundle and a few extra packages will remain
+) ipv6 is now built into RouterOS main package and is always enabled
-) BGP is disabled right now, until further notice
-) MPLS is disabled right now, until further notice
-) Not all packages have been published at the moment
-) Winbox does not show all features, use CLI for most functionality

Thanks @holvoetn,

I was looking for it on the stable 6.x versions changelog, so its available only on ROS 7 ...

Thanks @holvoetn,

I was looking for it on the stable 6.x versions changelog, so its available only on ROS 7 ...

Probably has to do with the change in Linux kernel from ROS 6 to 7.
I doubt the old ROS 6 kernel had support for it or it was too difficult to squeeze it in.

Thanks @holvoetn,

I was looking for it on the stable 6.x versions changelog, so its available only on ROS 7 ...

Probably has to do with the change in Linux kernel from ROS 6 to 7.
I doubt the old ROS 6 kernel had support for it or it was too difficult to squeeze it in.

Maybe...
The old manual https://wiki.mikrotik.com/wiki/Manual:Interface/OVPN indicates UDP on OpenVPN as unsupported but the newer https://help.mikrotik.com/docs/display/ROS/OpenVPN does not list it as unsupported, so it is supported.. it just won't mention the ROS versions supporting UDP...

I got the hEX RB750Gr-3 installed the ver 7.1, with netinstall and get the title "RouterOS 7.1 (testing)".
Using file: https://download.mikrotik.com/routeros/ ... -mmips.npk

I think the title should be "RouterOS 7.1 (stable)".

When i try to change to stable it says there no upgrade.

//Best regards
Patrik

Probably has to do with the change in Linux kernel from ROS 6 to 7.
I doubt the old ROS 6 kernel had support for it or it was too difficult to squeeze it in.

Maybe...
The old manual https://wiki.mikrotik.com/wiki/Manual:Interface/OVPN indicates UDP on OpenVPN as unsupported but the newer https://help.mikrotik.com/docs/display/ROS/OpenVPN does not list it as unsupported, so it is supported.. it just won't mention the ROS versions supporting UDP...

I could be wrong but most of the material on that "newer" help side was created when ROS7 came out.
But I 100% agree it is, to say the least, not clear at all.
E.g. for User manager there are already 3 documentation pages to be found. One on the older wiki platform, already 2 on the new environment for different versions.

Everything related to documentation needs a quite big overhaul with clear indication what works where or we're going to have a lot of similar conversations all over the place.

Wireguard bug: When you put peer with a ddns as address, peer does not come up when the first resolve fails. it keeps disconnected forever until a reboot, please FIX

Currently running 7.1 on my RB4011. Today I see all ports of switch 2 losing link every 6 minutes. Is this a known issue? Otherwise I have to investigate if my configuration has issues. edit: after downgrading to 6.49.2 (following this guide https://help.mikrotik.com/docs/display/ ... g+RouterOS) the issue is still present. Do I need to reset my configuration and import an old backup after the downgrade?

Just upgraded a mAP lite from 6.48.6 to 7.1. While this worked, I made an export before upgrading and compared with an export after.

This lines of config were added:
while these ip/ipv6 settings look legit, the others before are strange.

I mean WTF, this device (RBmAPL-2nD) does not even have a LTE interface. What is this?
I never configured anything for OSPF or BGP.

The configuration-migration seems to behave very strange. Seems like it adds some defaults for no apparent reason.

And now look at my configuration that was active on v6 before upgrade. This device is just a wireless-ethernet-bridge:

I've seen those parts too on map and maplite.

I found a problem with drop packet as invalid in forward in ipv6 fw.
When i have a mangle rule in postrouting with new packet mark in a queue that's exist
then all packets mark as invalid.
I try to change the names of queues or packets and i test it with simple and tree queues.
The results is the same.

Please confirm that.

I got the hEX RB750Gr-3 installed the ver 7.1, with netinstall and get the title "RouterOS 7.1 (testing)".
I think the title should be "RouterOS 7.1 (stable)".

When you are worried about that kind of textual things, it is not yet the time for you to test 7.1 !

It is indeed a bit chaotic on the moment. When 7.2 is out more thing will drop in place and 'order' will nomalise.

RouterOS 7.1 does not work properly with boards equipped with 32MB of RAM (e.g. RB411)

RouterOS 7.1 does not work properly with boards equipped with 32MB of RAM (e.g. RB411)

Code: Select all

00:01:06 system,error,critical router was rebooted without proper shutdown
00:01:07 system,error,critical kernel failure in previous boot
00:01:07 system,error,critical out of memory condition was detected
00:01:25 interface,info wireguard1 link up
00:01:44 wireguard,debug wireguard1: Interface created
00:01:55 ssh,error Corrupt host's key, regenerating it! Reboot required!

Working fine for me on an old SXT G-2HnD which only has 32Mb of RAM:

i still have the METAROUTER menu item on my v7 upgraded map lite. wasnt that TRASHED????? did they keep it for mipsbe platform????

i still have the METAROUTER menu item on my v7 upgraded map lite. wasnt that TRASHED????? did they keep it for mipsbe platform????

I see this menu on my hAP ac as well. There still seem to be some options in the GUI for things that no longer exist, like the checkbox to turn route caching on or off. There is no reason for that check box to exist any longer as there is no route caching to turn on or off, and unchecking it disables fastpath/fasttrack. So it could be a zombie/dummy option like that. Similarly I see in IP->Firewall->Service Ports that there are three invalid helpers (DCCP, SCTP, and UDPlite) - probably they have been removed but the menu options are still there.

I suspect this is being done to try to keep most of the codebase between v6 and v7 the same until they can focus on v7 development and allow them to diverge, otherwise there is more work involved in porting fixes backward or forward.

it was said at least one time directly by MT staff, that metarouter is not included in v7. so first I thought this is some winbox bug - but i see it on CLI too.

DOM/DDM on some SFPs not work anymore after upgrade to 7.1

Bildschirmfoto vom 2021-12-07 17-09-40.png

Bildschirmfoto vom 2021-12-07 17-06-04.png

I've reported the same issue via support 6 days ago, also on RB760iGS. For reference, the ticket number is SUP-68029. I've yet to receive any response, so you might try opening new one too ans sending them supout.rif file as it might help them narrow the issue (though from what I've seen the problem is universal to all RB760iGS and possibly related to i2c so they should be able to readily reproduce it).

I found a problem with drop packet as invalid in forward in ipv6 fw.
When i have a mangle rule in postrouting with new packet mark in a queue that's exist
then all packets mark as invalid.
I try to change the names of queues or packets and i test it with simple and tree queues.
The results is the same.

Please confirm that.

Also i try with new queue exclusive for ipv6 mangle and i check all chains.
The problem appear in prerouting, postrouting and forward.
Not in input or output.

Anyone know when RBLHGGR will be supported? Every-time I have tried to test it, unit fails to boot and have to reset it with v6 software.

RouterOS 7.1 does not work properly with boards equipped with 32MB of RAM (e.g. RB411)

Code: Select all

00:01:06 system,error,critical router was rebooted without proper shutdown
00:01:07 system,error,critical kernel failure in previous boot
00:01:07 system,error,critical out of memory condition was detected
00:01:25 interface,info wireguard1 link up
00:01:44 wireguard,debug wireguard1: Interface created
00:01:55 ssh,error Corrupt host's key, regenerating it! Reboot required!

i got this on hap ac2 and cap ac, controlled whit capsman. But sorry, cant provide supout any more.

Hi,

The new NTP client in 7.1 is having a very hard time synchronising the clock on an RB750Gr3.
Is this a problem with the new NTP client or is it showing an issue with the hardware clock in the device?
Regards,
Nigel

v7 consumes more RAM. on v6 my map lite has ~42mb free on fresh boot. on v7 it is only 25mb. maybe the 5.6 kernel caches more aggressively? or is it userspace that eats all the memory?

Hi,

The new NTP client in 7.1 is having a very hard time synchronising the clock on an RB750Gr3.
Is this a problem with the new NTP client or is it showing an issue with the hardware clock in the device?

Did you have a problem with 6.x software? I have no problem with my 7.x clients (not RB750Gr3) syncing using unicast on he client side.

no kvm on X86? why?

I wanted to update all my routers, because MT said it is stable. I was in the middle of another project, so I didn’t have time. I’m glad I didn’t update.

You made a good decision.

Is the 7.1 upgrade from 6.49.2 going to require any configuration changes, or is it a simple upgrade?

As always, first do an export of your current configuration, save it remote and then perform the upgrade. And to answer your question, yes it is a simple upgrade.

The new NTP client in 7.1 is having a very hard time synchronising the clock on an RB750Gr3.
Is this a problem with the new NTP client or is it showing an issue with the hardware clock in the device?

Code: Select all

[admin@MikroTik-Malthus] > /system/ntp/client/ print
         enabled: yes
            mode: unicast
         servers: nz.pool.ntp.org
      freq-drift: 415.449 PPM
          status: synchronized
   synced-server: nz.pool.ntp.org
  synced-stratum: 3
   system-offset: -1126.627 ms

That drift looks worrisome - that's, ~45% error?

That said, I did have some issues with NTP after upgrading (two IP's were hardcoded instead of DNS names, from which it failed to sync even though they were responding). Try adding multiple servers (0.nz.pool.ntp.org, 1.nz.pool.ntp.org, 2.nz.pool.ntp.org) and use /system/ntp/monitor-peers to check up on peers. Your current server is stratum 3 (most should be 2?) as well..

does 7.1 stable support DAC cable in rb4011?

I noticed the following when trying to log in using ssh key:
1. copy id_rsa.pub file to ROS
2. /user/ssh-keys/import public-key-file=id_rsa.pub user=admin
3. try to log in via key => still asks for password
Turns out that current OpenSSH client needs PubkeyAcceptedKeyTypes=+ssh-rsa set. Then it works.

So after some testa I had to do a rollback to V6 due to issues with OSPF.
It gets stuck in ExStart state on some links. Had the same issue on two diffrent routers. Same config but had diffrent results. Also interfaces that run over a L2TP (or anything PPP) do not convert over to a Interface Template but that has been said before.
So overall only issues with OSPF and on routers where I do not run OSPF no issues as of yet!

I have same NTP issues as reported earlier in this thread.

Specify NTP servers by IP-address fail (Waiting...).
Specify DNS name seem to work.

I found a problem with drop packet as invalid in forward in ipv6 fw.
When i have a mangle rule in postrouting with new packet mark in a queue that's exist
then all packets mark as invalid.
I try to change the names of queues or packets and i test it with simple and tree queues.
The results is the same.

Please confirm that.

Also i try with new queue exclusive for ipv6 mangle and i check all chains.
The problem appear in prerouting, postrouting and forward.
Not in input or output.

I also have this issue since at least 7.1rc3:
viewtopic.php?f=1&t=178695

Is the 7.1 upgrade from 6.49.2 going to require any configuration changes, or is it a simple upgrade?

As always, first do an export of your current configuration, save it remote and then perform the upgrade. And to answer your question, yes it is a simple upgrade.

It heavily depends on the features you are using on the router. On a typical NAT router, yes it is a simple upgrade. On a router that has BGP and/or OSPF, it is a painful experience at least, maybe even a disaster will result.
There are other such cases of features that no longer work acceptably.
But for the typical home user, that is not an issue.

I have same NTP issues as reported earlier in this thread.

Specify NTP servers by IP-address fail (Waiting...).
Specify DNS name seem to work.

NTP works fine on my RB4011 and my hAP ac2, with literal IP addresses.

The new NTP client in 7.1 is having a very hard time synchronising the clock on an RB750Gr3.
Is this a problem with the new NTP client or is it showing an issue with the hardware clock in the device?

Code: Select all

[admin@MikroTik-Malthus] > /system/ntp/client/ print
         enabled: yes
            mode: unicast
         servers: nz.pool.ntp.org
      freq-drift: 415.449 PPM
          status: synchronized
   synced-server: nz.pool.ntp.org
  synced-stratum: 3
   system-offset: -1126.627 ms

That drift looks worrisome - that's, ~45% error?

No, drift printed is 415 PPM and a fraction. So it's not 45%. However, this drift is still pretty high, usual x64 hardware usually runs well below 100PPM (PC I'm using to write this has -1.676 PPM). The information drift is giving is how much device's clock (either HW clock or kernel clock depending on implementation, knowing MT devices don't have HW clock it's kernel clock in ROS) drifts with time due to imprecise base frequency. NTP daemon automatically adjusts clock ticking mechanism to accommodate drift, but it takes some time (days) to get the correct value. While the drift value is not anywhere near correct value, it adjusts time in larger steps (e.g. by 1 second). If device clock frequency is stable, things get under control eventually. If device clock frequency varies (e.g. due to large temperature changes), then things might remain out of (complete) control.

I'm not sure about recent implementations of NTP daemon (and I surely don't know which implementation is now in ROS v7), but a while ago NTP daemons died if drift was larger than some big number (could be 500, could be 4000). And in this case things never got under control.

no kvm on X86? why?

KVM directly likely not. But "coming soon" docker-compatible Container support is likely the offering in V7.

Now for Container on X86, dunno specifically – RN, /container isn't shipping for any platform, so hard to say what will happen in future for Containers on X86

Anyone know when RBLHGGR will be supported? Every-time I have tried to test it, unit fails to boot and have to reset it with v6 software.

It works for me...

no kvm on X86? why?

KVM and MetaRouter is removed, Container will come as a replacement for all platforms.

Thanks @holvoetn,

I was looking for it on the stable 6.x versions changelog, so its available only on ROS 7 ...

Probably has to do with the change in Linux kernel from ROS 6 to 7.
I doubt the old ROS 6 kernel had support for it or it was too difficult to squeeze it in.

OpenVPN does not have or need any kernel support at all! It is a userspace application that uses sockets.
Unfortunately RouterOS does not use the OpenVPN reference application but is a workalike version created by MikroTIk themselves.
This originally did not support UDP and a lot of other features the reference application does support, but apparently nobody wanted to touch it anymore and it got more and more behind the times.
As part of the v7 project it has been updated, but I think it still isn't the reference application and there still are problems and incompatibilities...
(on the other hand, the reference application does not support TCP and UDP at the same time, so that would likely also receive bad comments)

On a terminal console, seems the question mark ?, that displayed a help screen on the RouterOS v6, is no longer working as expected (printing the help there) on the RouterOS v7.1.
Is the help missing an expected behavior or that's a bug? I REALLY expect that's a bug :)

RouterOS v6.49.2 RouterOS v7.1

The help is available with F1 key now.

The help is available with F1 key now.

Awesome, it's really there, thanks!! It would be also great to have it back on the question mark key, just like it worked on the v6, however.

Awesome, it's really there, thanks!! It would be also great to have it back on the question mark key, just like it worked on the v6, however.

I think this was changed because the question mark sometimes had to be entered as a literal character and was not supposed to bring up the help at that time.
There are also other characters which have special meaning that is undesired at some time (e.g. Ctrl-V which means "hotmode toggle" where people expect it to be "paste") which have been overhauled in v7.

You will have to get used to it.

Guys, I strongly believe i'm facing some memory leak on the new RouterOS v7.1. I have two boxes, one RB2011 and one RB450G (not powerful devices, I know, but powerful enough for the applications they are being used), that were upgraded to V7.1. They are both used to serve wireless networks with DHCP, VLAN and basically firewall filtering/NAT. There's really nothing fancy on these boxes. Just plain routing/filtering/dhcp/nat. Both have an OpenVPN UDP instance for management traffic.

There's one difference among them, however. One has a hotspot enabled, the other one the hotspot is still disabled (will be enabled on a few days). On the one with the hotspot enabled, memory usage seems to keep raising with time. I can see flat lines (on memory usage) when the place is closed and I know there's no traffic on the hotspot-enabled wireless network. But as soon as I start getting traffic there, memory usage starts rising again. Hotspot uses RADIUS so, the difference among them is the use of hotspot features and also RADIUS client ones.

System 1 (RB2011, hotspot enabled - upgrade moment highlighted on the Weekly graphic, memory usage on v7.1 is higher, no problem. But usage keeps rising with time and that's not expected, maybe indicating a memory leak)

System 2 (RB450G, hotspot *NOT* enabled, was upgraded Friday, higher usage can be seen on the graph, but keeps basically flat since there)

Rebooting System1 drops its memory usage, as expected to see when memory leaks are happening

IPv6 prefix acquisition over PPPoE was working perfectly with v6.x with this:
Now with v7.1 it just shows "Status: searching..." whereas it used to always grab a prefix immediately. I'm not using any VLANs or anything like that, just a very simple setup. What has changed? I've tried all the prefix lengths that I can think of, 56, 60, 62, not sure if that's the problem? I never used to have to enter it before.

Please add:

- "Old style" routing filters add (like v6), not only syntax

- BFD (for OSPF and BGP)

- MPLS Fast-Path

- BGP session show number of prefix


Thanks MK.

[:Y:]

Agreed. It would be nice to have both.

Is CAPSMAN supposed to be in v7.1 tile? I installed 7.1 and there was no capsman so I downgraded back to 6.49rc2 immediately.

IPv6 prefix acquisition over PPPoE was working perfectly with v6.x with this:

Code: Select all

/ipv6 dhcp-client add add-default-route=yes comment="IPv6 myISP" interface=pppoe-out1 pool-name= IPv6-myISP-pool1 request=prefix use-peer-dns=no
/ipv6 address add comment="IPv6 myISP" from-pool=IPv6-myISP-pool1 interface=bridge

That was wrong all the time. This example is copied over and over but the "add-default-route=yes" really should not be there!
A default route is already added by PPPoE, and doing it again in the dhcp-client was ignored in v6, but causes it to fail in v7.

Now with v7.1 it just shows "Status: searching..."

That is a different problem, though. Not sure why that happens, for me it works OK.

Is CAPSMAN supposed to be in v7.1 tile? I installed 7.1 and there was no capsman so I downgraded back to 6.49rc2 immediately.

It is there and works just fine. Using it on CCR1009.

system/resource/print

uptime: 1h41m2s
version: 7.1 (testing)
build-time: Dec/01/2021 14:07:27
factory-software: 6.46.4
free-memory: 3805.2MiB
total-memory: 4032.0MiB
cpu-count: 4
cpu-load: 31%
free-hdd-space: 90.8MiB
total-hdd-space: 129.0MiB
architecture-name: arm64
board-name: CCR2004-1G-12S+2XS
platform: MikroTik


I'm using as PPPoE-Server we have 500 simultaneous login connected, and simply the Routerboard is rebooting from time to time like 20x a day.
when I downgrade to v6.49x it works perfectly

無題.png

I am also seeing the same problem.
In my environment, restarting CCR1009 did not improve the CPU usage.

Same here, I disabled ipsec and some routing filters that were migrated over, and everything is OK for now. Before that a reboot would help for a little, but then the IPSec connection would die and 8 of the 9 cores would be at 100%.

I've moved my IPSec connection to Wireguard and the problem hasn't come back. But clearly there is something wrong I've just been able to avoid it.

Same here on CCR1016-12G.
None of the IPSEC setups are working anymore (L2TP PPP, IPSEC Site2Site with Azure).
A reboot helps for a few minutes... Then again, all connections drop and no one can connect...

Is CAPSMAN supposed to be in v7.1 tile? I installed 7.1 and there was no capsman so I downgraded back to 6.49rc2 immediately.

It is there and works just fine. Using it on CCR1009.

That is interesting since I have a CCR1009-7G-1C. Did you do anything special to make it appear? With v6, the winbox list of sections begins Quick Set, CAPsMAN, Interfaces, etc. With v7 WireGuard is in the position where CAPsMAN would be and CAPsMAN is nowhere to be found.

Did you have a problem with 6.x software? I have no problem with my 7.x clients (not RB750Gr3) syncing using unicast on he client side.

I didn't see the issue on 6.x

That is interesting since I have a CCR1009-7G-1C. Did you do anything special to make it appear? With v6, the winbox list of sections begins Quick Set, CAPsMAN, Interfaces, etc. With v7 WireGuard is in the position where CAPsMAN would be and CAPsMAN is nowhere to be found.

Did you happen to make a custom webfig skin to hide the CAPsMAN option? In v7, Winbox suddenly supports webfig skins, so if you did make a webfig skin that hid the CAPsMAN option it would suddenly disappear in Winbox.

I am using some simple queues could those be the issue?

Yes, looks like queue are to be blamed. Same for me: IPv6 works for my internal networks, but fails for my guest network with queue.

IPv6 prefix acquisition over PPPoE was working perfectly with v6.x with this:

Code: Select all

/ipv6 dhcp-client add add-default-route=yes comment="IPv6 myISP" interface=pppoe-out1 pool-name= IPv6-myISP-pool1 request=prefix use-peer-dns=no
/ipv6 address add comment="IPv6 myISP" from-pool=IPv6-myISP-pool1 interface=bridge

That was wrong all the time. This example is copied over and over but the "add-default-route=yes" really should not be there!
A default route is already added by PPPoE, and doing it again in the dhcp-client was ignored in v6, but causes it to fail in v7.

Now with v7.1 it just shows "Status: searching..."

That is a different problem, though. Not sure why that happens, for me it works OK.

Yes, I read about the default route thing too. I tried disabling that, but it makes no difference. At any rate that shouldn't affect it obtaining a prefix, but rather the routing of packets over IPv6.

That is interesting since I have a CCR1009-7G-1C. Did you do anything special to make it appear? With v6, the winbox list of sections begins Quick Set, CAPsMAN, Interfaces, etc. With v7 WireGuard is in the position where CAPsMAN would be and CAPsMAN is nowhere to be found.

Did you happen to make a custom webfig skin to hide the CAPsMAN option? In v7, Winbox suddenly supports webfig skins, so if you did make a webfig skin that hid the CAPsMAN option it would suddenly disappear in Winbox.

If I did, it was an accident since I don't even know what a webfig skin is. Where would I look for one?

If I did, it was an accident since I don't even know what a webfig skin is. Where would I look for one?

There would be a json file in the skins folder under Files. Skins are used to allow you to simplify the MikroTik interface for certain roles, devices and user types. For instance we hide a lot of advanced options from home customers who do not need to see such options and would be confused by them.

If I did, it was an accident since I don't even know what a webfig skin is. Where would I look for one?

There would be a json file in the skins folder under Files. Skins are used to allow you to simplify the MikroTik interface for certain roles, devices and user types. For instance we hide a lot of advanced options from home customers who do not need to see such options and would be confused by them.

There is a default.json file dated 11/25/2017. There was an .npk file accidently placed in the skins folder which I have now deleted. Should I delete default.json and try v7 again?

There is a default.json file dated 11/25/2017. There was an .npk file accidently placed in the skins folder which I have now deleted. Should I delete default.json and try v7 again?

Yes, that is probably the cause of the problem. Delete that file and reboot the device.

I have same NTP issues as reported earlier in this thread.

Specify NTP servers by IP-address fail (Waiting...).
Specify DNS name seem to work.

NTP works fine on my RB4011 and my hAP ac2, with literal IP addresses.

I had issues with the IP addresses configured with an upgrade from 6.49 to 7.1 i deleted the NTP servers and re-added them and they synced up right away, guessing something associated with the config translation.

no kvm on X86? why?

KVM and MetaRouter is removed, Container will come as a replacement for all platforms.

thanks!

no kvm on X86? why?

KVM directly likely not. But "coming soon" docker-compatible Container support is likely the offering in V7.

Now for Container on X86, dunno specifically – RN, /container isn't shipping for any platform, so hard to say what will happen in future for Containers on X86

thanks!

we are still getting issue about packets forward in a very strange way after several hours of operation.
- 3 days on a router with less route and less traffic
- 36 hours in a router with heavy traffic and 4 full routes

when the router enter in strange beavior it is:
- traceroute icmp through the router stops at the router ip
- tracerout from the router works
- some route are not able to be checkd by ping
- the router is fully accessible
- disabling/enabling interface doesn't cure the issue
- the only way to recover the router is a reboot

all my experience point to some issue about packet forwarding of some specific protocoll like icmp but also udp is at leeast affected.

waiting an answer from Mikrotik since dec1!!!!!!!

We were able to repoduce in the lab the strange behavior on two routers back to back doing bandwidth test for 2 days.
We updated the ticket [SUP-67221] with the sup out generated.

Any new news about the fix?
regards
Ros

I think v7.1.1 and 7.2(testing) and 6.50(testing) are gonna be fun changelogs to read once they arrive.

Given that flood of issues (and there are certainly many, many more behind the scenes not seen in the forum, I guess it would be easiest to state a blank "various bugfixes" in the changelog of v7.1.1...

While I'm surprised that v7.1 was released as stable before the holidays, I do hope that the Mikrotik took care and obtained agreement of employees to manage such an extraordinary support/dev workload during the holidays before deciding to publish this release.

My best wishes to Mikrotik support!

Bridge filtering and bridge IP Firewall Filter still working abnormal. Chateau device. I mention it in SUP-66472.

I also observed bridge filters not working in 7.1, which worked fine in 6.49.1. One example: This blocked all multicast traffic including igmp requests. In 7.1 no frames are matched.
I've set up acl rules on the upstream interface (comware) now, which works.

I have problem with SSTP tunnel to Windows Server 2016 in ROS v7.1 on RB2011. In previous software (6.49.2) everything is ok. In v7.1 log showed error: ssl: decode error (6) and even debug mode no show more info.

In v6.49.2 SSTP work great. Has anyone else encountered such a problem?

does 7.1 stable support DAC cable in rb4011?

Despite the recommendations seems that yes, it does. I have connected rb4011 and ccr2004 by DAC, works well. Also tested with Mellanox ConnectX3, no issues so far.

I assume that SUP-issue numbers are sequentially ascending. I submit a ticket then and now, and there are huge number-gaps between. Judging by that assumption, they get a heavy load of support-tickets right now (or at least since 7.1 got into RC state). Pretty brave to release stable, when the whole user community finds so many issues.

1) User manager shows time without timezone offset.
2) frequency value in the wave2 wifi settings is disappeared after click to apply or ok buttons. It's actually saves in config, but not not displayed.

i "add batch users" 100 vouchers in user manager. i right click "generate voucher" and i only find the option to print one voucher.
how do you print multiple vouchers?

I switched to 7.1 from 6.48 and know i'm stumbled across some issues:
- RouterOS v7 BGP feature Confederation is green in the feature list, but how to configure it? - the BGP session with confederation are marked as invalid, and the peer router says Notification Error Message: (OPEN Message Error/Bad Peer AS.) or Notification Error Message: (UPDATE Message Error/Malformed AS_PATH.) if tried to play with AS and Sub AS
- how to view current BGP advertised routes for neighbors?

if anything is invalid maybe a message with what exactly causes that entry to be invalid would be great ...

There is a default.json file dated 11/25/2017. There was an .npk file accidently placed in the skins folder which I have now deleted. Should I delete default.json and try v7 again?

Yes, that is probably the cause of the problem. Delefe that file and reboot the device.

It was the cause. Thanks for the help.

I have 7.1 release installed on Audience AP and it seems that the Access List feature is either not working or I am not using it correctly.

I have whitelisted MAC addresses of devices which I want to allow to connect, but my Apple Watch still connected with random MAC address (the latest iOS update turned Private Address back on).

Pre-emptive clarification: Yes, I am aware that MAC whitelisting should not be used as a security measure -- I am only using it as an additional level of deterrence (AP is properly secured with long random password and strong encryption).

I have 7.1 release installed on Audience AP and it seems that the Access List feature is either not working or I am not using it correctly.

I have whitelisted MAC addresses of devices which I want to allow to connect, but my Apple Watch still connected with random MAC address (the latest iOS update turned Private Address back on).

So what are you saying? That Apple Watch can't connect? That would be expected if it uses random MAC addresses ... Or is it that Apple Watch can connect with random MAC address which means whitelisting doesn't really work?

Did you properly configure your wireless interfaces without "default authenticate" option?

I have access lists perfectly working on 7.1 with Chateau. But Private Address means random MAC. So you won't be successful whitelisting such a device....

Testing the CHR vhdx on-top of Windows 10 hype-v with 8 nics (gen1 vm).
* Can only see the first network interface
* ACPI reboot+shutdown doesn't work (reboot is stuck with the "Rebooting..." text"

Same as before, but now target scope needs to be greater than scope, while in RouterOS v6 you could set target scope to the same as scope.

Wrong, btw:

[admin@Oracle-CHR-7] > ip/route/print detail

Flags: D - dynamic; X - disabled, I - inactive, A - active; c - connect, s - static, r - rip, b - bgp, o - ospf, d - dhcp, v - vpn, m - modem, y - copy; H - hw-offloaded; + - ecmp

DAd dst-address=0.0.0.0/0 routing-table=main pref-src="" gateway=10.0.0.1 immediate-gw=10.0.0.1%ether1 distance=1 scope=30 target-scope=10 vrf-interface=ether1 suppress-hw-offload=no

0 As dst-address=8.8.4.4/32 routing-table=main pref-src="" gateway=10.0.0.1 immediate-gw=10.0.0.1%ether1 distance=1 scope=15 target-scope=10 suppress-hw-offload=no
cope=10 suppress-hw-offload=no

1 As dst-address=12.12.12.12/32 routing-table=main pref-src="" gateway=8.8.4.4 immediate-gw=10.0.0.1%ether1 distance=1 scope=30 target-scope=15 suppress-hw-offload=no

15 is definitely not greater than 15

BGP's "Remove Privarte AS" function does not work

IPv6 prefix acquisition over PPPoE was working perfectly with v6.x with this:

Code: Select all

/ipv6 dhcp-client add add-default-route=yes comment="IPv6 myISP" interface=pppoe-out1 pool-name= IPv6-myISP-pool1 request=prefix use-peer-dns=no
/ipv6 address add comment="IPv6 myISP" from-pool=IPv6-myISP-pool1 interface=bridge

Now with v7.1 it just shows "Status: searching..." whereas it used to always grab a prefix immediately. I'm not using any VLANs or anything like that, just a very simple setup. What has changed? I've tried all the prefix lengths that I can think of, 56, 60, 62, not sure if that's the problem? I never used to have to enter it before.

Another major issue on my hAP-ac-lite: the entire router seems to hang, all of my active internet connections stop working, and I can't even login to Webfig and have to unplug the power to the router. This has happened multiple times since the v7.1 upgrade, so I just downgraded it to 6.49.2.

oh my dear. `/export` hangs on RB941-2nD.

But Private Address means random MAC. So you won't be successful whitelisting such a device....

Fortunately for most systems, "random MAC" only means a one-time chosen random MAC for each SSID in the connect list on the device.
So next time your device comes back to your network, it has the same MAC as before.
(this was often not true in the first release of the feature, causing massive pool exhaustion problems in networks with multiple access points on a single router+DHCP server)

oh my dear. `/export` hangs on RB941-2nD.

In this case, be patient. It will finish after 5 or 10 minutes (or even more).
Then look at the result, it will have failure messages in it that indicate where it is hanging.

Yeah, I was patient. Waited for 30min or even longer but it did not finish and I went to bed instead. But yes, there were error-messages in the export-file like "#error exporting /mpls/interface".

oh my dear. `/export` hangs on RB941-2nD.

Try

Another potential bug, when upgrading from V7.1rc7, all roads lead to v7.1 [.0].

Is upgrading to V7 a one-way road? Since I don't have any V6 downgrade options anymore...e.g. selecting "long-term", "stable", and "testing" all seem to want to use v7.1). Selecting "development" reports "file not found".

Or, and maybe this is documented, but can you still use upload the V6 packages to file system to downgrade it – or is netinstall required to downgrade?

i went to 7.1 750gr3 and it was a horror. i have a few scripts and all of them at certain line failed. the ppoe was selected right in the route but i have to remove the route and add it again for it not be shown in red

i put the 6.49.2 file in the root of router and told it to downgrade and it works.

Hope this helps

hangs too. I remember that from the early 7.1 beta versions when the export hung on every platform.

oh my dear. `/export` hangs on RB941-2nD.

It's not hanged, it just takes several hours to complete, outputting stuff that's not relevant to a smips device.

RB931 has the same issue.

It is trying to gather config that is simply just not there. After a timeout it continues. Trying to dump config from features/services that aren't available on smips - good job.

You can try if there is any difference after "reset to defaults" or even "format and netinstall".
But of course a new issue with v7 is that you are forced to install functional blocks that you are never going to use (like mpls), and be faced with any bugs caused by them...

does OSPF works over WireGuard ? I just get hello packets and nothing more...

if i not create te default route directly connected example:dst.address:0.0.0.0/0 geteway:192.168.0.1 (eth1) wich is my modem/router and only leave these rotutes for example: dst.address:1.1.1.1-gateway:192.168.1.1 scope:11 target scope:10
dst.address:0.0.0/0-gateway:1.1.1.1 scope:12 target scope:11. the internet not working, i have to add the route directly connecte in order to make it work, it normal?

Not normal. You have different gateways (192.168.0.1 VS 192.168.1.1) - is it a typo? All routes are in 'main' routing table, right?

Can you provide 'ip/route/print detail' when the problem occurs?

does OSPF works over WireGuard ? I just get hello packets and nothing more...

Yes, works for me. The configuration for interface-template looks like this:
The peers that initiate the wireguard connection (having endpoint-address set) need a static neighbor to initiate the connection:
Update addresses and names to match your configuration...

Note that OSPFv3 is currently broken as no link local addresses are assigned to wireguard interfaces.

Why only Zerotier on ARM architecture?

Does ROS 7 incerase cpu-count on CRS326-24G-2S?

ON ROS6:
ON ROS7.1

I am currently experiencing an issue with BGP and VRFs on a CRS309.
I added a BGP peer and set the routing-table and vrf to the desired VRF:

After session establishment, I check it and the routing-table is correctly set to public, but there is no mention of the vrf attribute:

But when I check the routes learned via BGP, the gateway doesn't mention the public table and the route is marked as inactive, because the gateway is not reachable on the default routing table/VRF.

BGP is installing the routes without specifying the routing table for the gateway. Is this a bug or am I missing a configuration?

Does ROS 7 incerase cpu-count on CRS326-24G-2S?

ON ROS6:

Code: Select all

                   uptime: 5w1d19h43m58s
                  version: 6.49 (stable)
               build-time: Oct/06/2021 11:55:34
         factory-software: 6.44
...
                cpu-count: 1
...

ON ROS7.1

Code: Select all

                   uptime: 1h18m12s
                  version: 7.1 (testing)
               build-time: Dec/01/2021 14:07:27
         factory-software: 6.38.2
...
                cpu-count: 2
...

Factory SW version also changed ??

You can try if there is any difference after "reset to defaults" or even "format and netinstall".
But of course a new issue with v7 is that you are forced to install functional blocks that you are never going to use (like mpls), and be faced with any bugs caused by them...

No difference, seen same behaviour since 7.0beta4.

Does ROS 7 incerase cpu-count on CRS326-24G-2S?

It does on CRS328. The hardware always had a cpu with two cores, but RouterOS 6.x used just one.
I guess it's the same for CRS326.

Factory SW version also changed ??

I guess he compares two devices of same type. 😉

is it possible 7.1 is showing logical cores instead of physical cores?

edit: searching for that exact cpu model shows it as a dual core. this may be something unique to that exact model and how it worked on v6?

Hi
I have some question. Could someone answer and help. I have a Chateau LTE12.
1) Is it possible to downgrade to version 6.49.2 from version 7.1 ? It was upgraded to version 7.0.3 earlier .
2) I have configured dot1x. The ISE server is behind the ipsec tunnel. I do not see any traffic on the firewall in the direction of ISE, the computer does not authenticate. Firmware 7.03 did not have dot1x in the menu. Does dot1x work correctly in 7.1?

is it possible 7.1 is showing logical cores instead of physical cores?

edit: searching for that exact cpu model shows it as a dual core. this may be something unique to that exact model and how it worked on v6?

There are the same models of CRS326. On v6 CPU is seen as single core. According to manual this router has single core: https://mikrotik.com/product/CRS326-24G-2SplusRM but according to 98DX3236 chip manual there is dual core http://www.datasheet.hk/view_download.p ... 598419.pdf probably v6 didn't quite support this chip

Did the MLPPP implementation somehow lose its packet-reordering capabilities in ROS 7? On ROS 6 i could use MLPPP over three connections on a bond-interface and it would take care of packet reordering. Since using ROS 7 i am getting a large amount (25%+) of out of order packets.

RouterOS 7.1 does not work properly with boards equipped with 32MB of RAM (e.g. RB411)

Code: Select all

00:01:06 system,error,critical router was rebooted without proper shutdown
00:01:07 system,error,critical kernel failure in previous boot
00:01:07 system,error,critical out of memory condition was detected
00:01:25 interface,info wireguard1 link up
00:01:44 wireguard,debug wireguard1: Interface created
00:01:55 ssh,error Corrupt host's key, regenerating it! Reboot required!

Working fine for me on an old SXT G-2HnD which only has 32Mb of RAM:

Code: Select all

]admin@gxxxxxxxxxxx > /system/resource/print
                   uptime: 1d19h1m11s
                  version: 7.1 (testing)
               build-time: Dec/01/2021 14:07:27
              free-memory: 5.6MiB
             total-memory: 32.0MiB
                      cpu: MIPS 24Kc V7.4
                cpu-count: 1
            cpu-frequency: 400MHz
                 cpu-load: 39%
           free-hdd-space: 109.9MiB
          total-hdd-space: 128.0MiB
  write-sect-since-reboot: 6063
         write-sect-total: 2891337
               bad-blocks: 0%
        architecture-name: mipsbe
               board-name: SXT G-2HnD
                 platform: MikroTik

The problem is repeatable. Below instruction howto crash all services on ROS7.1:
1. On RB411 set wireguard to other Mikrotik router
2. On another Mikrotik (in my case RB211iL) set wireguard to RB411 3. On another Mikrotik (in my case RB2011iL) run Tools->Bandwidth test, and type IP 192.168.99.3 (via wireguard). Change several times between UDP and TCP mode.
4. After several seconds all services ROS (ftp, ssh, Winbox) on RB411 will hang (only icmp works from kernel, but unstable).

this is on v6.49 but is the same on 7.1:
see the image, if i delete the route on "red circle" internet not work, its a config with balance pcc.

So, you're deleting the only default route in 'main' table. After that (as you didn't tell what "internet is not working" means) router's DNS can't contact external servers, for example.

why did they remove routing_table parameter from PING/TraceRoute ?

could it be readded? We currently do not use VRF but actively use Route Tables.

I would appreciate if it can be re-added in ROS v7. It was very useful for debugging.

Thanks a lot.

Upgraded my RB751U, boot time increased to 4-5 minutes! I noticed that after FW upgrade. Not complaining, it's a lab unit.

Hi
I have some question. Could someone answer and help. I have a Chateau LTE12.
1) Is it possible to downgrade to version 6.49.2 from version 7.1 ? It was upgraded to version 7.0.3 earlier .
2) I have configured dot1x. The ISE server is behind the ipsec tunnel. I do not see any traffic on the firewall in the direction of ISE, the computer does not authenticate. Firmware 7.03 did not have dot1x in the menu. Does dot1x work correctly in 7.1?

The Chateau is a v7 device only. Says so right on the product page. https://mikrotik.com/product/chateau_lte12

why did they remove routing_table parameter from PING/TraceRoute ?

could it be readded? We currently do not use VRF but actively use Route Tables.

I would appreciate if it can be re-added in ROS v7. It was very useful for debugging.

You can set the source address. When you have ip route rules to select a route table based on the source address (or maybe DSCP) you can still force ping to use the desired table.

I am very impressed by the achievements of MikroTik with version 7.1 stable; in a short time it seems to me unthinkable to switch from Linux v3 to v5, but nevertheless I ask you to pay attention to the MBIM protocol. I think Telit is a good device. As I wrote in viewtopic.php?p=881764&hilit=LM960A18#p881764, not everything could be fixed.
Point one: transaction error is a thing of the past, but instead errors [lte, info lte1 mbim: error: >>> E service: conn_ext, command: bs info, error: 2
lte, info lte1 mbim: error: >>> E service: conn_ext, command: bs info, error: 7
lte, info lte1 mbim: error: >>> E service: conn_ext, command: bs info, error: 12]
What does this mean in terms of MicroTik? Please describe the entire list.
Point two: I managed to overcome it with the help of NetInstall.
The rest of the items remained unchanged.
Please, help.

so i have to leave for example if i have 2 isp's 1 route "main" for every one? sorry for my english

You may leave both for failover reasons. The main point is you need to have an active default route in 'main' table

Hello everyone,

first of all, congratulations on releasing the v7.1.
I was wondering if someone can help me out with this problem?
I have Wireguard client running on the HAP AC Lite, latest ROS, latest firmware.
The problem is that the tunnel never comes back after the power outage, in order to restablish the tunnel again, I have to reset the wireguard interface.
I have some netwatch script to reset the interface if the ping goes down, but it doesn't seem to work.
Normal reboot seems to be fine?
Additionally, I'm using FQDN in the endpoint address.

Thank you!

I ask you to pay attention to the MBIM protocol. I think Telit is a good device. As I wrote in viewtopic.php?p=881764&hilit=LM960A18#p881764, not everything could be fixed.

True... With their current miniPCIe-based devices, the Telit 960 is works pretty well and we use them. So I know this problem very well. The lack of RSRP & RSRQ is actually even more noticeable since with only RSSI, hard to separate interference from pointing/position. It also results in the Mikrotik mobile app graphs producing odd results since the RSSI scale is not same as RSRP/RSPQ.

AFAIK Mikrotik's v7 LTE MBIM support is just providing what the MBIM data the modem provides. Now MBIM does support RSRP/RSPQ directly via MBIM, but it's very unclear if it's Mikrotik uses it or Telit is just not sending it right...

While MBIM doesn't offer the the LTE-A/5G CA data directly, MBIM allows AT commands (e.g. how "at-chat" works). So ROS could, and should, issue the "AT#CAINFO" and use that info to report DIRECTLY into RouterOS existing variables for RSRP, CA, etc., thus useable with Mikrotik's UI.

I use scripts to read/record all the Telit data, but since I can't report to the Mikrotik app, it poses issues for us since we'd like customer to use the app with the Telit modem installed.

So +1 from me here.

Background how MBIM works vs ECM for LTE interfaces: For modems in ECM modem, Mikrotik issues AT commands to make the modem connect. So it makes sense to read the signal using the same AT channel. The problem is these AT command vary and change over time, so subtle change in a AT command between modems, things could stop working. Which is why MBIM was created – to abstract away the AT commands into a protocol. MBIM protocol makes LTE connections are more like VPN interfaces (as example, not in practice): you provide some auth data/connection info and the protocol hides most differences between modem. Since MBIM offers signal data as part of the MBIM protocol, Mikrotik uses that information, not AT commands. So what you see in ROS should be what the Telit is reporting via MBIM. The question is there a bug in Mikrotik's reading of the data OR is Telit not sending it (or sending in not according to the MBIM protocol)...hard to say. BUT no RSRQ and RSRP appear in ROS, least make verifying/tuning LTE much more difficult.

any ticket opened about x86 platform are answered by Arturs in the same very stupid way..... "maybe some drevers are misisng."
But it is not the case!!!!!
# DEVICE VENDOR NAME IRQ
0 00:00.0 Intel Corporation Sky Lake-E DMI3 Registers (rev: 7) 0
1 00:04.0 Intel Corporation Sky Lake-E CBDMA Registers (rev: 7) 11
2 00:04.1 Intel Corporation Sky Lake-E CBDMA Registers (rev: 7) 10
3 00:04.2 Intel Corporation Sky Lake-E CBDMA Registers (rev: 7) 11
4 00:04.3 Intel Corporation Sky Lake-E CBDMA Registers (rev: 7) 11
5 00:04.4 Intel Corporation Sky Lake-E CBDMA Registers (rev: 7) 11
6 00:04.5 Intel Corporation Sky Lake-E CBDMA Registers (rev: 7) 10
7 00:04.6 Intel Corporation Sky Lake-E CBDMA Registers (rev: 7) 11
8 00:04.7 Intel Corporation Sky Lake-E CBDMA Registers (rev: 7) 11
9 00:05.0 Intel Corporation Sky Lake-E MM/Vt-d Configuration Registers (rev: 7) 0
10 00:05.2 Intel Corporation Sky Lake-E RAS (rev: 7) 0
11 00:05.4 Intel Corporation Sky Lake-E IOAPIC (rev: 7) 0
12 00:08.0 Intel Corporation Sky Lake-E Ubox Registers (rev: 7) 0
13 00:08.1 Intel Corporation Sky Lake-E Ubox Registers (rev: 7) 0
14 00:08.2 Intel Corporation Sky Lake-E Ubox Registers (rev: 7) 0
15 00:11.0 Intel Corporation C620 Series Chipset Family MROM 0 (rev: 4) 0
16 00:14.0 Intel Corporation C620 Series Chipset Family USB 3.0 xHCI Controller (rev: 4) 11
17 00:16.0 Intel Corporation C620 Series Chipset Family MEI Controller #1 (rev: 4) 11
18 00:16.4 Intel Corporation C620 Series Chipset Family MEI Controller #3 (rev: 4) 11
19 00:17.0 Intel Corporation C620 Series Chipset Family SATA Controller [AHCI mode] (rev: 4) 11
20 00:1c.0 Intel Corporation C620 Series Chipset Family PCI Express Root Port #1 (rev: 244) 11
21 00:1c.2 Intel Corporation C620 Series Chipset Family PCI Express Root Port #3 (rev: 244) 11
22 00:1c.5 Intel Corporation C620 Series Chipset Family PCI Express Root Port #6 (rev: 244) 10
23 00:1f.0 Intel Corporation C627 Series Chipset LPC/eSPI Controller (rev: 4) 0
24 00:1f.2 Intel Corporation C620 Series Chipset Family Power Management Controller (rev: 4) 0
25 00:1f.4 Intel Corporation C620 Series Chipset Family SMBus (rev: 4) 0
26 00:1f.5 Intel Corporation C620 Series Chipset Family SPI Controller (rev: 4) 0
27 02:00.0 ASPEED Technology, Inc. AST1150 PCI-to-PCI Bridge (rev: 3) 0
28 03:00.0 ASPEED Technology, Inc. ASPEED Graphics Family (rev: 48) 11
29 17:00.0 Intel Corporation Sky Lake-E PCI Express Root Port A (rev: 7) 11
30 17:02.0 Intel Corporation Sky Lake-E PCI Express Root Port C (rev: 7) 11
31 17:05.0 Intel Corporation Sky Lake-E VT-d (rev: 7) 0
32 17:05.2 Intel Corporation Sky Lake-E RAS Configuration Registers (rev: 7) 0
33 17:05.4 Intel Corporation Sky Lake-E IOxAPIC Configuration Registers (rev: 7) 0
34 17:08.0 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
35 17:08.1 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
36 17:08.2 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
37 17:08.3 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
38 17:08.4 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
39 17:08.5 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
40 17:08.6 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
41 17:08.7 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
42 17:09.0 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
43 17:09.1 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
44 17:09.2 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
45 17:09.3 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
46 17:09.4 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
47 17:09.5 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
48 17:09.6 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
49 17:09.7 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
50 17:0a.0 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
51 17:0a.1 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
52 17:0a.2 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
53 17:0a.3 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
54 17:0a.4 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
55 17:0a.5 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
56 17:0a.6 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
57 17:0a.7 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
58 17:0b.0 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
59 17:0b.1 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
60 17:0b.2 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
61 17:0b.3 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
62 17:0e.0 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
63 17:0e.1 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
64 17:0e.2 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
65 17:0e.3 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
66 17:0e.4 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
67 17:0e.5 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
68 17:0e.6 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
69 17:0e.7 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
70 17:0f.0 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
71 17:0f.1 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
72 17:0f.2 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
73 17:0f.3 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
74 17:0f.4 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
75 17:0f.5 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
76 17:0f.6 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
77 17:0f.7 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
78 17:10.0 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
79 17:10.1 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
80 17:10.2 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
81 17:10.3 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
82 17:10.4 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
83 17:10.5 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
84 17:10.6 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
85 17:10.7 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
86 17:11.0 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
87 17:11.1 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
88 17:11.2 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
89 17:11.3 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
90 17:1d.0 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
91 17:1d.1 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
92 17:1d.2 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
93 17:1d.3 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
94 17:1e.0 Intel Corporation Sky Lake-E PCU Registers (rev: 7) 0
95 17:1e.1 Intel Corporation Sky Lake-E PCU Registers (rev: 7) 0
96 17:1e.2 Intel Corporation Sky Lake-E PCU Registers (rev: 7) 0
97 17:1e.3 Intel Corporation Sky Lake-E PCU Registers (rev: 7) 0
98 17:1e.4 Intel Corporation Sky Lake-E PCU Registers (rev: 7) 0
99 17:1e.5 Intel Corporation Sky Lake-E PCU Registers (rev: 7) 0
100 17:1e.6 Intel Corporation Sky Lake-E PCU Registers (rev: 7) 0
101 18:00.0 Intel Corporation Ethernet Controller X710 for 10GbE SFP+ (rev: 2) 11
102 18:00.1 Intel Corporation Ethernet Controller X710 for 10GbE SFP+ (rev: 2) 11
103 18:00.2 Intel Corporation Ethernet Controller X710 for 10GbE SFP+ (rev: 2) 11
104 18:00.3 Intel Corporation Ethernet Controller X710 for 10GbE SFP+ (rev: 2) 11
105 19:00.0 Intel Corporation Ethernet Controller X710 for 10GbE SFP+ (rev: 2) 11
106 19:00.1 Intel Corporation Ethernet Controller X710 for 10GbE SFP+ (rev: 2) 11
107 19:00.2 Intel Corporation Ethernet Controller X710 for 10GbE SFP+ (rev: 2) 11
108 19:00.3 Intel Corporation Ethernet Controller X710 for 10GbE SFP+ (rev: 2) 11
109 3a:00.0 Intel Corporation Sky Lake-E PCI Express Root Port A (rev: 7) 11
110 3a:05.0 Intel Corporation Sky Lake-E VT-d (rev: 7) 0
111 3a:05.2 Intel Corporation Sky Lake-E RAS Configuration Registers (rev: 7) 0
112 3a:05.4 Intel Corporation Sky Lake-E IOxAPIC Configuration Registers (rev: 7) 0
113 3a:08.0 Intel Corporation Sky Lake-E Integrated Memory Controller (rev: 7) 0
114 3a:09.0 Intel Corporation Sky Lake-E Integrated Memory Controller (rev: 7) 0
115 3a:0a.0 Intel Corporation Sky Lake-E Integrated Memory Controller (rev: 7) 0
116 3a:0a.1 Intel Corporation Sky Lake-E Integrated Memory Controller (rev: 7) 0
117 3a:0a.2 Intel Corporation Sky Lake-E Integrated Memory Controller (rev: 7) 0
118 3a:0a.3 Intel Corporation Sky Lake-E Integrated Memory Controller (rev: 7) 0
119 3a:0a.4 Intel Corporation Sky Lake-E Integrated Memory Controller (rev: 7) 0
120 3a:0a.5 Intel Corporation Sky Lake-E LM Channel 1 (rev: 7) 0
121 3a:0a.6 Intel Corporation Sky Lake-E LMS Channel 1 (rev: 7) 0
122 3a:0a.7 Intel Corporation Sky Lake-E LMDP Channel 1 (rev: 7) 0
123 3a:0b.0 Intel Corporation Sky Lake-E DECS Channel 2 (rev: 7) 0
124 3a:0b.1 Intel Corporation Sky Lake-E LM Channel 2 (rev: 7) 0
125 3a:0b.2 Intel Corporation Sky Lake-E LMS Channel 2 (rev: 7) 0
126 3a:0b.3 Intel Corporation Sky Lake-E LMDP Channel 2 (rev: 7) 0
127 3a:0c.0 Intel Corporation Sky Lake-E Integrated Memory Controller (rev: 7) 0
128 3a:0c.1 Intel Corporation Sky Lake-E Integrated Memory Controller (rev: 7) 0
129 3a:0c.2 Intel Corporation Sky Lake-E Integrated Memory Controller (rev: 7) 0
130 3a:0c.3 Intel Corporation Sky Lake-E Integrated Memory Controller (rev: 7) 0
131 3a:0c.4 Intel Corporation Sky Lake-E Integrated Memory Controller (rev: 7) 0
132 3a:0c.5 Intel Corporation Sky Lake-E LM Channel 1 (rev: 7) 0
133 3a:0c.6 Intel Corporation Sky Lake-E LMS Channel 1 (rev: 7) 0
134 3a:0c.7 Intel Corporation Sky Lake-E LMDP Channel 1 (rev: 7) 0
135 3a:0d.0 Intel Corporation Sky Lake-E DECS Channel 2 (rev: 7) 0
136 3a:0d.1 Intel Corporation Sky Lake-E LM Channel 2 (rev: 7) 0
137 3a:0d.2 Intel Corporation Sky Lake-E LMS Channel 2 (rev: 7) 0
138 3a:0d.3 Intel Corporation Sky Lake-E LMDP Channel 2 (rev: 7) 0
139 3b:00.0 Mellanox Technologies MT27800 Family [ConnectX-5] (rev: 0) 11
140 3b:00.1 Mellanox Technologies MT27800 Family [ConnectX-5] (rev: 0) 10
141 5d:00.0 Intel Corporation Sky Lake-E PCI Express Root Port A (rev: 7) 11
142 5d:05.0 Intel Corporation Sky Lake-E VT-d (rev: 7) 0
143 5d:05.2 Intel Corporation Sky Lake-E RAS Configuration Registers (rev: 7) 0
144 5d:05.4 Intel Corporation Sky Lake-E IOxAPIC Configuration Registers (rev: 7) 0
145 5d:0e.0 Intel Corporation Sky Lake-E KTI 0 (rev: 7) 0
146 5d:0e.1 Intel Corporation Sky Lake-E UPI Registers (rev: 7) 0
147 5d:0f.0 Intel Corporation Sky Lake-E KTI 0 (rev: 7) 0
148 5d:0f.1 Intel Corporation Sky Lake-E UPI Registers (rev: 7) 0
149 5d:10.0 Intel Corporation Sky Lake-E KTI 0 (rev: 7) 0
150 5d:10.1 Intel Corporation Sky Lake-E UPI Registers (rev: 7) 0
151 5d:12.0 Intel Corporation Sky Lake-E M3KTI Registers (rev: 7) 0
152 5d:12.1 Intel Corporation Sky Lake-E M3KTI Registers (rev: 7) 0
153 5d:12.2 Intel Corporation Sky Lake-E M3KTI Registers (rev: 7) 0
154 5d:12.4 Intel Corporation Sky Lake-E M3KTI Registers (rev: 7) 0
155 5d:12.5 Intel Corporation Sky Lake-E M3KTI Registers (rev: 7) 0
156 5d:15.0 Intel Corporation Sky Lake-E M2PCI Registers (rev: 7) 0
157 5d:15.1 Intel Corporation Sky Lake-E DDRIO Registers (rev: 7) 0
158 5d:16.0 Intel Corporation Sky Lake-E M2PCI Registers (rev: 7) 0
159 5d:16.1 Intel Corporation Sky Lake-E DDRIO Registers (rev: 7) 0
160 5d:16.4 Intel Corporation Sky Lake-E M2PCI Registers (rev: 7) 0
161 5d:16.5 Intel Corporation Sky Lake-E DDRIO Registers (rev: 7) 0
162 5d:17.0 Intel Corporation Sky Lake-E M2PCI Registers (rev: 7) 0
163 5d:17.1 Intel Corporation Sky Lake-E DDRIO Registers (rev: 7) 0
164 5e:00.0 Intel Corporation unknown device (rev: 4) 11
165 5f:00.0 Intel Corporation unknown device (rev: 4) 11
166 5f:01.0 Intel Corporation unknown device (rev: 4) 10
167 5f:02.0 Intel Corporation unknown device (rev: 4) 11
168 5f:03.0 Intel Corporation unknown device (rev: 4) 11
169 60:00.0 Intel Corporation C62x Chipset QuickAssist Technology (rev: 4) 11
170 61:00.0 Intel Corporation C62x Chipset QuickAssist Technology (rev: 4) 10
171 62:00.0 Intel Corporation C62x Chipset QuickAssist Technology (rev: 4) 11
172 63:00.0 Intel Corporation Ethernet Connection X722 for 10GbE SFP+ (rev: 4) 11
173 63:00.1 Intel Corporation Ethernet Connection X722 for 10GbE SFP+ (rev: 4) 11
174 80:04.0 Intel Corporation Sky Lake-E CBDMA Registers (rev: 7) 11
175 80:04.1 Intel Corporation Sky Lake-E CBDMA Registers (rev: 7) 10
176 80:04.2 Intel Corporation Sky Lake-E CBDMA Registers (rev: 7) 11
177 80:04.3 Intel Corporation Sky Lake-E CBDMA Registers (rev: 7) 11
178 80:04.4 Intel Corporation Sky Lake-E CBDMA Registers (rev: 7) 11
179 80:04.5 Intel Corporation Sky Lake-E CBDMA Registers (rev: 7) 10
180 80:04.6 Intel Corporation Sky Lake-E CBDMA Registers (rev: 7) 11
181 80:04.7 Intel Corporation Sky Lake-E CBDMA Registers (rev: 7) 11
182 80:05.0 Intel Corporation Sky Lake-E MM/Vt-d Configuration Registers (rev: 7) 0
183 80:05.2 Intel Corporation Sky Lake-E RAS (rev: 7) 0
184 80:05.4 Intel Corporation Sky Lake-E IOAPIC (rev: 7) 0
185 80:08.0 Intel Corporation Sky Lake-E Ubox Registers (rev: 7) 0
186 80:08.1 Intel Corporation Sky Lake-E Ubox Registers (rev: 7) 0
187 80:08.2 Intel Corporation Sky Lake-E Ubox Registers (rev: 7) 0
188 85:00.0 Intel Corporation Sky Lake-E PCI Express Root Port A (rev: 7) 11
189 85:05.0 Intel Corporation Sky Lake-E VT-d (rev: 7) 0
190 85:05.2 Intel Corporation Sky Lake-E RAS Configuration Registers (rev: 7) 0
191 85:05.4 Intel Corporation Sky Lake-E IOxAPIC Configuration Registers (rev: 7) 0
192 85:08.0 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
193 85:08.1 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
194 85:08.2 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
195 85:08.3 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
196 85:08.4 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
197 85:08.5 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
198 85:08.6 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
199 85:08.7 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
200 85:09.0 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
201 85:09.1 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
202 85:09.2 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
203 85:09.3 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
204 85:09.4 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
205 85:09.5 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
206 85:09.6 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
207 85:09.7 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
208 85:0a.0 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
209 85:0a.1 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
210 85:0a.2 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
211 85:0a.3 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
212 85:0a.4 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
213 85:0a.5 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
214 85:0a.6 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
215 85:0a.7 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
216 85:0b.0 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
217 85:0b.1 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
218 85:0b.2 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
219 85:0b.3 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
220 85:0e.0 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
221 85:0e.1 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
222 85:0e.2 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
223 85:0e.3 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
224 85:0e.4 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
225 85:0e.5 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
226 85:0e.6 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
227 85:0e.7 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
228 85:0f.0 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
229 85:0f.1 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
230 85:0f.2 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
231 85:0f.3 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
232 85:0f.4 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
233 85:0f.5 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
234 85:0f.6 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
235 85:0f.7 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
236 85:10.0 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
237 85:10.1 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
238 85:10.2 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
239 85:10.3 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
240 85:10.4 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
241 85:10.5 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
242 85:10.6 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
243 85:10.7 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
244 85:11.0 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
245 85:11.1 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
246 85:11.2 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
247 85:11.3 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
248 85:1d.0 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
249 85:1d.1 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
250 85:1d.2 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
251 85:1d.3 Intel Corporation Sky Lake-E CHA Registers (rev: 7) 0
252 85:1e.0 Intel Corporation Sky Lake-E PCU Registers (rev: 7) 0
253 85:1e.1 Intel Corporation Sky Lake-E PCU Registers (rev: 7) 0
254 85:1e.2 Intel Corporation Sky Lake-E PCU Registers (rev: 7) 0
255 85:1e.3 Intel Corporation Sky Lake-E PCU Registers (rev: 7) 0
256 85:1e.4 Intel Corporation Sky Lake-E PCU Registers (rev: 7) 0
257 85:1e.5 Intel Corporation Sky Lake-E PCU Registers (rev: 7) 0
258 85:1e.6 Intel Corporation Sky Lake-E PCU Registers (rev: 7) 0
259 86:00.0 Mellanox Technologies MT27800 Family [ConnectX-5] (rev: 0) 11
260 86:00.1 Mellanox Technologies MT27800 Family [ConnectX-5] (rev: 0) 10
261 ae:00.0 Intel Corporation Sky Lake-E PCI Express Root Port A (rev: 7) 11
262 ae:05.0 Intel Corporation Sky Lake-E VT-d (rev: 7) 0
263 ae:05.2 Intel Corporation Sky Lake-E RAS Configuration Registers (rev: 7) 0
264 ae:05.4 Intel Corporation Sky Lake-E IOxAPIC Configuration Registers (rev: 7) 0
265 ae:08.0 Intel Corporation Sky Lake-E Integrated Memory Controller (rev: 7) 0
266 ae:09.0 Intel Corporation Sky Lake-E Integrated Memory Controller (rev: 7) 0
267 ae:0a.0 Intel Corporation Sky Lake-E Integrated Memory Controller (rev: 7) 0
268 ae:0a.1 Intel Corporation Sky Lake-E Integrated Memory Controller (rev: 7) 0
269 ae:0a.2 Intel Corporation Sky Lake-E Integrated Memory Controller (rev: 7) 0
270 ae:0a.3 Intel Corporation Sky Lake-E Integrated Memory Controller (rev: 7) 0
271 ae:0a.4 Intel Corporation Sky Lake-E Integrated Memory Controller (rev: 7) 0
272 ae:0a.5 Intel Corporation Sky Lake-E LM Channel 1 (rev: 7) 0
273 ae:0a.6 Intel Corporation Sky Lake-E LMS Channel 1 (rev: 7) 0
274 ae:0a.7 Intel Corporation Sky Lake-E LMDP Channel 1 (rev: 7) 0
275 ae:0b.0 Intel Corporation Sky Lake-E DECS Channel 2 (rev: 7) 0
276 ae:0b.1 Intel Corporation Sky Lake-E LM Channel 2 (rev: 7) 0
277 ae:0b.2 Intel Corporation Sky Lake-E LMS Channel 2 (rev: 7) 0
278 ae:0b.3 Intel Corporation Sky Lake-E LMDP Channel 2 (rev: 7) 0
279 ae:0c.0 Intel Corporation Sky Lake-E Integrated Memory Controller (rev: 7) 0
280 ae:0c.1 Intel Corporation Sky Lake-E Integrated Memory Controller (rev: 7) 0
281 ae:0c.2 Intel Corporation Sky Lake-E Integrated Memory Controller (rev: 7) 0
282 ae:0c.3 Intel Corporation Sky Lake-E Integrated Memory Controller (rev: 7) 0
283 ae:0c.4 Intel Corporation Sky Lake-E Integrated Memory Controller (rev: 7) 0
284 ae:0c.5 Intel Corporation Sky Lake-E LM Channel 1 (rev: 7) 0
285 ae:0c.6 Intel Corporation Sky Lake-E LMS Channel 1 (rev: 7) 0
286 ae:0c.7 Intel Corporation Sky Lake-E LMDP Channel 1 (rev: 7) 0
287 ae:0d.0 Intel Corporation Sky Lake-E DECS Channel 2 (rev: 7) 0
288 ae:0d.1 Intel Corporation Sky Lake-E LM Channel 2 (rev: 7) 0
289 ae:0d.2 Intel Corporation Sky Lake-E LMS Channel 2 (rev: 7) 0
290 ae:0d.3 Intel Corporation Sky Lake-E LMDP Channel 2 (rev: 7) 0
291 af:00.0 Mellanox Technologies MT27800 Family [ConnectX-5] (rev: 0) 11
292 af:00.1 Mellanox Technologies MT27800 Family [ConnectX-5] (rev: 0) 10
293 d7:05.0 Intel Corporation Sky Lake-E VT-d (rev: 7) 0
294 d7:05.2 Intel Corporation Sky Lake-E RAS Configuration Registers (rev: 7) 0
295 d7:05.4 Intel Corporation Sky Lake-E IOxAPIC Configuration Registers (rev: 7) 0
296 d7:0e.0 Intel Corporation Sky Lake-E KTI 0 (rev: 7) 0
297 d7:0e.1 Intel Corporation Sky Lake-E UPI Registers (rev: 7) 0
298 d7:0f.0 Intel Corporation Sky Lake-E KTI 0 (rev: 7) 0
299 d7:0f.1 Intel Corporation Sky Lake-E UPI Registers (rev: 7) 0
300 d7:10.0 Intel Corporation Sky Lake-E KTI 0 (rev: 7) 0
301 d7:10.1 Intel Corporation Sky Lake-E UPI Registers (rev: 7) 0
302 d7:12.0 Intel Corporation Sky Lake-E M3KTI Registers (rev: 7) 0
303 d7:12.1 Intel Corporation Sky Lake-E M3KTI Registers (rev: 7) 0
304 d7:12.2 Intel Corporation Sky Lake-E M3KTI Registers (rev: 7) 0
305 d7:12.4 Intel Corporation Sky Lake-E M3KTI Registers (rev: 7) 0
306 d7:12.5 Intel Corporation Sky Lake-E M3KTI Registers (rev: 7) 0
307 d7:15.0 Intel Corporation Sky Lake-E M2PCI Registers (rev: 7) 0
308 d7:15.1 Intel Corporation Sky Lake-E DDRIO Registers (rev: 7) 0
309 d7:16.0 Intel Corporation Sky Lake-E M2PCI Registers (rev: 7) 0
310 d7:16.1 Intel Corporation Sky Lake-E DDRIO Registers (rev: 7) 0
311 d7:16.4 Intel Corporation Sky Lake-E M2PCI Registers (rev: 7) 0
312 d7:16.5 Intel Corporation Sky Lake-E DDRIO Registers (rev: 7) 0
313 d7:17.0 Intel Corporation Sky Lake-E M2PCI Registers (rev: 7) 0
314 d7:17.1 Intel Corporation Sky Lake-E DDRIO Registers (rev: 7) 0

Very disappointed by your low level of support!!!!
What a shame!

any ticket opened about x86 platform are answered by Arturs in the same very stupid way..... "maybe some drevers are misisng."
But it is not the case!!!!!
Very disappointed by your low level of support!!!!
What a shame!

they think that they provide FOSS software for free as in price.. what a shame!

almost all of our competitor WISPs gone for UI. we consider migrating to Ubiquiti, too.

Probably it was meant that there might be a problem with the driver, we do not write those drivers.

we do not write those drivers.

never late. hire some low-level programmers if necessary.

just like cloudflare.. to optimize OSS and fix one-in-a-million bugs

MikroTik does not make or sell x86 systems. Our developers are working on MikroTik made hardware support. If the used kernel supports your PC, it works. If not, not much we can do about it. CHR is a good option to make sure everything works. These guys just tested 100Gbit on CHR: https://stubarea51.net/2021/12/13/mikro ... g-barrier/

we consider migrating to Ubiquiti, too.

And good luck installing their OS on x86 barebone.

MikroTik does not make or sell x86 systems. Our developers are working on MikroTik made hardware support. If the used kernel supports your PC, it works. If not, not much we can do about it. CHR is a good option to make sure everything works. These guys just tested 100Gbit on CHR: https://stubarea51.net/2021/12/13/mikro ... g-barrier/

we don't have issue on running 100gbps cards on our x86.
We have issue on your software after some days due to icmp and udp issue not related to any speific interface: routers is not able to forward icmp and udp is malformed.
we are able to reproduce the issue on our lab
This is not an issue hardware related, but just software. I think related to the amount of traffic handled by the router.

Please look at the supouts attached to the ticket[SUP-67221].

I get your point, we will check it, but there are no issues with same amount of traffic on CHR or MikroTik hardware.

Hello everyone,

first of all, congratulations on releasing the v7.1.
I was wondering if someone can help me out with this problem?
I have Wireguard client running on the HAP AC Lite, latest ROS, latest firmware.
The problem is that the tunnel never comes back after the power outage, in order to restablish the tunnel again, I have to reset the wireguard interface.
I have some netwatch script to reset the interface if the ping goes down, but it doesn't seem to work.
Normal reboot seems to be fine?
Additionally, I'm using FQDN in the endpoint address.

Thank you!

Disable / enable peer.
No need to reset ITF.
Issue which is already pending for quite some time but not always.

can anybody check following issue - viewtopic.php?t=181202 ? just want to be sure, that dynamic mesh is still working between ROS 7 devices before upgrade.

thanks.

I get your point, we will check it, but there are no issues with same amount of traffic on CHR or MikroTik hardware.

100gbps rela traffic on CHR or Mikrotik Hardware?????????????? Don't think so.

Please look into it, you will find the issue. We contributed in last 18 months reporting a lot of issue on v7.
My idea is that the issue is related to default config for each route to use Hardware Offload, for the static I can chenge it but not for connected.

regards

Just noticed that pre 7.1 /routing pim is gone.

Guess there's not "automagic upgrade path" with the new /routing/pimsm/ ?

My idea is that the issue is related to default config for each route to use Hardware Offload, for the static I can chenge it but not for connected.

There is no hardware offload support for x86. Those settings simply doesn't do anything.

@Amm0: Thanks a lot for your feedback.

To receive data about LTE or 5G base stations, ROS v7 seems
to support Microsoft MBIM extension (MBIMEx v1),
at least request MBIM_CID_BASE_STATIONS_INFO, where for example
the MBIM_LTE_SERVING_CELL_INFO structure contains data about
CellID, EARFCN, RSRP, RSRQ, etc.

Moreover, the RSRQ parameter is described as an integer
(The range is -20 to -3, in units of 1dBm).
ROS shows this from -2.0 to -0.3; probably the issue on the ROS side.
RSRP shows ok.
RSSI differs from actual by ~ 30%
CA Band: line in WinBox is present, but it is empty.

Whether ROS v7 supports MBIMEx v2 or v3 I don't know.
Perhaps CA, SINR, CQI will appear in the next extensions :)

And of course, I still want to know what is behind the messages in the logs:
lte, info lte1 mbim: error: >>> E service: conn_ext, command: bs info, error: 12

IP firewall service-port (ALG) for QUIC protocol (HTTP3) is necessary.

Before I disable (reject in firewall) QUIC , whatsapp, messenger and many applications experience packet-loss and unacceptable jitter.

After I reject QUIC, connections fallback to TCP (HTTP2) and all connections are stable.

So I would ask you mikrotik guys to write QUIC ALG to keep connections alive. for example we could set different TIMEOUT for UDP 443 and UDP 80 and also something like keepalive mechanism may be implemented.

It is also possible to extract SNI information from QUIC protocol but I guess ROS will need to watch connections from the beginning. so The ALG could also do this job which is one of most-wanted features...

100gbps rela traffic on CHR or Mikrotik Hardware?????????????? Don't think so.

You don't think, but it's true. CHR is faster than x86 bare hardware. You have to start living in the now, not in the past.

https://stubarea51.net/2021/12/13/mikro ... g-barrier/

100gbps rela traffic on CHR or Mikrotik Hardware?????????????? Don't think so.

You don't think, but it's true. CHR is faster than x86 bare hardware. You have to start living in the now, not in the past.

https://stubarea51.net/2021/12/13/mikro ... g-barrier/

I live were found confortable. But if you publish x86 v7 release you have to support them because we bought several L6 licences!!!

Moving to the faster and more modern CHR is free, email us, we will issue you license for CHR.

Moving to the faster and more modern CHR is free, email us, we will issue you license for CHR.

fix our issue first. We don't have the time to test a completely new platforma after two years of v7 waiting.
then with calm we will go to investigate latest chr

It works for us here, there is nothing we can fix, most likely this is due to hardware incompatbilitiy with RouterOS kernel. We did not make the hardware, so there is no warranty. If you want warranty, please use products made by MikroTik.

It works for us here, there is nothing we can fix, most likely this is due to hardware incompatbilitiy with RouterOS kernel. We did not make the hardware, so there is no warranty. If you want warranty, please use products made by MikroTik.

may you looke at the supout before to say anything?
Working here is not an answer, application confioguration could araise bug you don't see in you local application!

Give a look at supout!!!
please!!!

We did look, before your ticket was answered. Like always.

We did look, before your ticket was answered. Like always.

into the ticket I reported the availability of the routers to investigate directly over them because I was able to replicate the issue in lab.
will you not go to investigate?

Looks like some memory is not released after removing static routes.
I was adding and deleting about 30k static routes on CHR.
After just 3-4 cycles CHR died with:

I am trying to set up a hotspot with RouterOS 7.1...
The device has a valid certificate, that is configured in hotspot profile. Looks like port 443 is accessible for clients, but the server does not speak https.
Anybody else with a problem like this?

Making my own Thread

And of course, I still want to know what is behind the messages in the logs:
lte, info lte1 mbim: error: >>> E service: conn_ext, command: bs info, error: 12

@Milecus, I'll watch out that one on our Telit 960s. One thought on the error: do you have the GPS package installed, or have GPS searching enabled on the Telit? Google suggests "bs info" might be related to A-GPS.

Moreover, the RSRQ parameter is described as an integer (The range is -20 to -3, in units of 1dBm).
ROS shows this from -2.0 to -0.3; probably the issue on the ROS side.
RSRP shows ok.
RSSI differs from actual by ~ 30%
CA Band: line in WinBox is present, but it is empty.

Yup pretty close to what I see, it's RSRP and RSRQ that don't seem to show up MBIM mode & RSSI is similar skewed wrong.

While I can run a script in a loop, to read all this stuff better. The bad signal data isn't exactly very end-user friendly for our customers if they had to use MT app/webfig.

Didn't think about this until reading your post: the exact signal reporting behavor may also vary by specific carrier firmware – 4 different ones inside. I'll try some SIMs & AT#FIRMWARE to see what effect that has on this. Been mainly using Verizon in US with them, but have AT&T and T-Mobile we can try. Verizon is generally weird... so maybe at least get same RSRP/RSRQ results as you with Generic or AT&T profile.

We did look, before your ticket was answered. Like always.

into the ticket I reported the availability of the routers to investigate directly over them because I was able to replicate the issue in lab.
will you not go to investigate?

You're kidding right?

Personally running V7 X86 ROS seems like a unnecessary risk - you really want to the the first person trying a specific chipset that may not have been test before with ROS? V7 still has issues on well-known hardware. Plus someone "looking at it" is likely only get a bug into a queue to get fixed... and since it may be chipset MT may not be familiar, likely not an easy one. Just saying this doesn't seem like a good path.

Moving to the faster and more modern CHR is free, email us, we will issue you license for CHR.

Leave the hardware abstraction to a hypervisor. And VMWare is generally free and simply to setup, you follow the instructions on CHR.

It's quite handy since you can do "snapshot" of RouterOS to rollback or forward upgrade pretty easily. Or, bring up a 2nd virtual router to try a new version. I have 0 complaints about CHR, although those are staying at 6.x for a bit longer myself.

It was maybe answered somewhere in this thread but I cannot find it: hAP ac2 units with 256MB RAM - can they benefit from ROS 7 in term of "wifiwave 2" package ?

Thanks !

It was maybe answered somewhere in this thread but I cannot find it: hAP ac2 units with 256MB RAM - can they benefit from ROS 7 in term of "wifiwave 2" package ?

Thanks !

No. Anything with 16MB storage cannot get wifiwave2 package. Bloated wifiwave2 package is here to stay unfortunately.

No. Anything with 16MB storage cannot get wifiwave2 package. Bloated wifiwave2 package is here to stay unfortunately.

... and underpowered (16MB flash) devices as well...

Looks like API does not support adding multiple items in single PUT: Only adding 1 by 1 works with shitty speed

The MikroTik website and forum were not available a day ago. Repeated post.

@Amm0

> Google suggests "bs info" might be related to A-GPS.

I am not using GPS package. It looks like this error refers to connecting to BS services.
FE: (error #12 - packet service detached) There is no time to dig into MBIM raw packets yet.
Since MBIM is a fairly unified protocol, it would be nice to have a description of
the errors add to MikroTik Wiki.

> Yup pretty close to what I see, it's RSRP and RSRQ that don't seem to show up MBIM mode & RSSI is similar skewed wrong.

It looks like Telit has added support for Microsoft MBIM extensions for the LM960A18 from firmware v6 and up.
You can check this with the AT#GETFW or AT#GETFWEXT commands. I now have firmware v8 (32.00.118);
"Generic" profile:

HOST FIRMWARE : 32.00.008_2
SLOT STATUS CARRIER VERSION TMCFG CNV
1 Activated Generic 32.00.118 1027 32101006
2 Verizon 32.00.128 2022 32101006
3 ATT 32.00.148 4024 32101006
4 TMUS 32.00.158 5007 empty

But issues with showing RSSI & RSRQ remain :(

Hi, I am new in mikrotik

Just wanna ask "dumb" question:

Is it really worth to upgrade to 7.1 from 6.49 2?

I have hAP lite and only use it to manage a home network with 3 APs in two bridges and approx 13-20 users.

I have no problems with the current version, but just curious.

If you don't need the new features and everything still works for you, no need to upgrade.
You can but it's not needed.

Stay on v6 if you do not need any of the v7 features.

Hi, I am new in mikrotik

Just wanna ask "dumb" question:

Is it really worth to upgrade to 7.1 from 6.49 2?

I have hAP lite and only use it to manage a home network with 3 APs in two bridges and approx 13-20 users.

I have no problems with the current version, but just curious.

i have an hap lite too...
just for testing i try to update to 7 stable setup openvpn udp and wireguard.
if i connect works but random cpu runs 100% and i must restart (manual poweroff / poweron)
so for now, i downgrade to 6...

i wait 7 in long term... :)

Why isn't the hap ac2 supported for wifiwave2? Mine has 256mb of RAM and ipq4019. Is it because there is only 16mb flash space?

hAP ac2 is not supported, read here about wifiwave2 requirements and features: https://help.mikrotik.com/docs/display/ROS/WifiWave2

I am sure reggger wants to read more about the features....

Maybe he will be relieved, because there is not much to gain yet

Where are our posts from the weekend???
Forum was down yesterday. Assuming you have guys restored from backup and they are lost?
Just guessing, log4j cve issue?

Any plans for spectral scan/history on ac chips in ROSv7?

Just guessing, log4j cve issue?

RouterOS does not use Java

Where are our posts from the weekend???
Forum was down yesterday. Assuming you have guys restored from backup and they are lost?
Just guessing, log4j cve issue?

Power failure.
My guess as well, most likely restore from incomplete backup.

Where are our posts from the weekend???
Forum was down yesterday. Assuming you have guys restored from backup and they are lost?
Just guessing, log4j cve issue?

Power failure.
My guess as well, most likely restore from incomplete backup.

Forum is functional as RouterOs...
In normal bussines is program reporting, if the server is down, immediately people get sms what's happend.
But in this company not..In Monday morning in work.. "Hmm server is down, but how long? Nevermind.."

MikroTik, please make some tender on very good people. Hire some dumb people, who only take money and waste your time..
If you embrace good people, it will take less time to make good firmwares and less money for it..
2 good people is better, than 10 bad..Give that 2 people more money offer...Where is the problem? Where is problem take some people from another country, if your country aren't that clever people..

Winap, please don't spread these lies.

Power failure.
My guess as well, most likely restore from incomplete backup.

Forum is functional as RouterOs...
In normal bussines is program reporting, if the server is down, immediately people get sms what's happend.
But in this company not..In Monday morning in work.. "Hmm server is down, but how long? Nevermind.."

MikroTik, please make some tender on very good people. Hire some dumb people, who only take money and waste your time..
If you embrace good people, it will take less time to make good firmwares and less money for it..
2 good people is better, than 10 bad..Give that 2 people more money offer...Where is the problem? Where is problem take some people from another country, if your country aren't that clever people..

If you dont like the forum then leave. Such comments are a waste of your energy and life. Probably have better things to do........

Winap, please don't spread these lies.

I thing, there is almost every month, when server is down..
Sometimes twice time. What is problem with servers and your firmware?
I buy some hardware, with some function..but in newer versions does not work ..Is it also my problem?
Make good FW and people will be happy..Is it like lottery..

If you dont like the forum then leave. Such comments are a waste of your energy and life. Probably have better things to do........

I like forum, but I don't like bad firmware...Forum is good to know, where is problem with every new firmware..so this is necessarily.
Critics is also necessarily..if people are happy on every buggy FW, is it ok?

Upgraded my home router CCR1009 from 6.49.2 to 7.1: OSPF stopped. Started to check - ospf-in rules imported but... looks like default "allow all" in 6.x changed to "deny all" and incoming routes were filtered. Ok. Not big problem. IPSec - works but after clicking on IPSec config tabs had to reboot due to channel stuck and packets didn't go inside/outside channel. I.e. if you something change in IPSec - please reboot to apply changes %)

But.. mangle not works at all! Adding simple connection marking rule like: breaks everything - all connections constantly are disconnecting ones per minute... "stable version". Reboot not helps.
PS: even if add change mss rule breaks all with same disconnects one per minute. Looks like mangle chain is broken fully in 7.1. I can't replace it with "routes rule" (like some people on forum propose to do) due to multi-wan balancing (input and output) and need mark packets basing on different criteria.

@winap go make your own thread for wallowing and crying 'woe is me'.....This thread is for the v7.1 release.
I'd be glad to hear about any specific problems you're having with v7.1.....many others here have mentioned problems or concerns they have with the new release and MT staff has been responding.

as an aside: my home router has been running the 7.1 stable (RB3011 with really basic config) since release and I have had no problems.

deadkat:
Ok so we are going to attack here? The problems, who has Mikrotik is a basic problem in OS. Please realize, so many people don't have time check this forum and control, which firmware is good and what is bad..brick HW, random restart in new OS, non functional some FSP optic port in new FW..
Every human have a opinion, so don't take it or make fun of that..

Marry christmas deadkat :)

winap, your post adds no value and makes it harder to find actual reports and things mikrotik can fix

normis:
Ok sorry...
Marry christmas also to you.

@winap go make your own thread for wallowing and crying 'woe is me'.....This thread is for the v7.1 release.
I'd be glad to hear about any specific problems you're having with v7.1.....many others here have mentioned problems or concerns they have with the new release and MT staff has been responding.

as an aside: my home router has been running the 7.1 stable (RB3011 with really basic config) since release and I have had no problems.

My RB3011 home router also has been working without issue (Aside from OpenVPN UDP transport dropping out and locking up - switching back to the original TCP method supported in 6.x stabilises the issue)

Still running 7.1 on the RB3011, 2 x hAPac2 running as CAPs managed with CAPsMAN on the RB3011 are also running 7.1 without a problem.

New version v7.1.1 has been released:
viewtopic.php?t=181472

New version v7.2rc1 has been released:
viewtopic.php?t=181474