Read and compare the changelogs.Does this have all the fixes from v7.3 beta 37?
That issue was fixed with 7.3b37.Bricked another ac3 with wifiwave2
I am amazed that MikroTik promotes a version to "stable" with such an issue present, and without mentioning it in the release notes.That issue was fixed with 7.3b37.Bricked another ac3 with wifiwave2
It should be mentioned this version 7.2.2 is not to be used for devices using ww2 !
This is NOT stable ... my view.
Must all of us compare the changelogs? It's a tedious process. I have the same question. Please can someone advise if it's worth upgrading to 7.2.2 from 7.3beta37. Due to the ww2 issue it seems 7.3beta37 is more up to date than 7.2.2 as it contains more fixes. It my assumption correct ?Read and compare the changelogs.Does this have all the fixes from v7.3 beta 37?
bbbwuahhh...same here....coming from 7.2.1...WTF?That issue was fixed with 7.3b37.Bricked another ac3 with wifiwave2
It should be mentioned this version 7.2.2 is not to be used for devices using ww2 !
This is NOT stable ... my view.
You mean fire the users??!? We are the BETA testers. It's the only explanation ive heard that makes senseAnother great STABLE release from MT. Guys you should wake up and fire the person who is responsible for testing those stable builds. I have 2x hap AC3 and Im lucky I did not install this very stable biuld. The problem has been announced at 2:35 p.m. Im just curious how long MT will leave this build on their update servers and how many bricked wifiwave 2 devices they will need to react.
Netinstall using older version which worked before.Yeap, as I always read this forum before upgrade, this time I did not. My ac3 with ww2 is now bricked. Stable version, hahaha.
Read one post above ......jep...and the hap-ac3 is now fully bricked, as netinstall (v7.2.1) does not seem to work either.
According to this: https://help.mikrotik.com/docs/display/ ... andJumpers the LED should go off when entering netinstall mode, but it never does go off....it starts flashing green, but then returns to solid blue.
.....anyone care to elborate on how to unbrick this unit?
It may require several attempts....thats the problem...it never does appear in netinstall...
Thank you for your answer, but I was made it many times from years.Netinstall using older version which worked before.Yeap, as I always read this forum before upgrade, this time I did not. My ac3 with ww2 is now bricked. Stable version, hahaha.
Or go immediately to 7.3b37.
...thats the problem...it never does appear in netinstall...
Netinstall possible or is it completely bricked? If so, thats fuktYeap, as I always read this forum before upgrade, this time I did not. My ac3 with ww2 is now bricked. Stable version, hahaha.
Netinstall to 7.3.b37, configuration restored from backup file and it works with ww2 :-)Netinstall possible or is it completely bricked? If so, thats fuktYeap, as I always read this forum before upgrade, this time I did not. My ac3 with ww2 is now bricked. Stable version, hahaha.
EDIT: Looks like it is possible, just needs some perseverance
And the same bug was also introduced in 7.2.2 "stable", which was released later than 7.3beta37...What's new in 7.3beta37 (2022-Apr-25 15:29):
*) system - fixed RouterOS bootup when wifiwave2 package is installed (introduced in v7.3beta34);
...thats the problem...it never does appear in netinstall...
Having developed SW myself for many years, I really wonder how MT manages to produce such blunders again and again.
Sorry MikroTik, but something with your SW dev and test processes is more than broken.
Well, of course there is always some "freeze" moment where code is no longer being ported into a "stable" version, and it may be before the release of the last beta.And the same bug was also introduced in 7.2.2 "stable", which was released later than 7.3beta37...What's new in 7.3beta37 (2022-Apr-25 15:29):
*) system - fixed RouterOS bootup when wifiwave2 package is installed (introduced in v7.3beta34);
Bugs introduced in v7.3beta34 also appear in 7.2.2 "stable"?
I've never had NetInstall work on a Windows machine. Never had it fail on Linux....thats the problem...it never does appear in netinstall...
If we step back from the immediate issue (RoS v7 is public beta) we can see smart, helpful, and talented people following a broken process. This is a concern for management and I do wonder what is the incentive for them to change that process? What is happening makes them look juvenile and brand damaging.
In the normal world yes. In the world of Mikrotik labels like "stable" and "RC" and are just randomly attached to versions and do not mean much.But in case of such a blocking problem I would expect either a delay of the stable version, or a warning "do not upgrade to this version when you use wifiwave2".
tragedy:"bgp - added initial support for prefix limit"
Thank you for confirming, but I will stay on the 7.3beta.Upgrade on Chateau LTE12 no issues.
/ip ipsec policy set 0 dst-address=2001:xxxx/128 src-address=2001:yyyy/128
/ip ipsec policy set 0 disabled=no dst-address=::/0 group=default proposal=default protocol=all src-address=::/0 template=yes
So those two ww2 related fixes were tested on 7.2.2 before release despite the fact 7.2.2 does not even boot with ww2 installed?*) leds - fixed wireless related LED behavior with WW2 package;
*) ww2 - fixed VLAN tag handling;
Reading this... Can we have something similar for DNS, please? At least if DoH is enabled the DNS subsystem can flood the log extensively.*) conntrack - limited full Connection Tracking warning to 1 message per minute;
+1Reading this... Can we have something similar for DNS, please? At least if DoH is enabled the DNS subsystem can flood the log extensively.*) conntrack - limited full Connection Tracking warning to 1 message per minute;
I finally managed to get it working for mine.So, I am the one who get bricked device (HAP ac3). Is it possible to bring it back to live? I've tried to download 7.3beta37 netinstall script and firmware, than connect lan1 port with Ethernet cable, on my PC I set 192.168.88.3 ip and default gateway (disabled systemd-networkd unit), started netinstall. Than, I hold reset button and plug the cable in device, it start flash pink, then blue, than flashing green, than constant green, and then blue again. I tried to hold 5, 10, 15, 20, 30 seconds, no result. The link shows as up on my Linux machine, but the netinstall still “Waiting for RouterBOARD...”.
I do not see why all is upset. Yes, its not good that software do breaks the router.I am really angry, as it took me literally 5-6 hours to setup and test...even with linux and dump switch in between did not work with netinstall
Put it in the signature part of your user account ;)1. Upgrade a router just minutes after a new release. Just stupid thing to do. Wait some weeks.
2. Did you do this on a production unit, again your fault. If you have a network of routers, always test on a same type of router with same config and see if all goes fine.
3. Last but not least, do READ this forum!!! It will save all of you for problems.
Not the first time I have written this, and I am not sure it will be the last.
Did you read this forum before upgrade?Argh the wiviwave2 package make my hAPAC3 bootloop forever...
Does not help if owner of Mikrotik Routers does not bother to read the forum.Put it in the signature part of your user account ;)
If the upgrade is making from routers bricks Mikrotik should remove it ASAP. No one expect from stable version this bricking firmware....Does not help if owner of Mikrotik Routers does not bother to read the forum.Put it in the signature part of your user account ;)
It didn’t brick any of mine. Everything is relative and, as has been said before, rushing to install a new release means you assume risks. Before EVER installing an update, I check out the forum. Once I determine that the issue doesn’t affect me (2xRB5009 and HeX) I update.If the upgrade is making from routers bricks Mikrotik should remove it ASAP. No one expect from stable version this bricking firmware....
I'm not saying the problems could not have been avoided and this particular one is a real goof-up (since the fix WAS available already in 7.3b37) but ... there is a small company in Redmond also known for releasing (sometimes CRITICAL) hotfixes each time they release a new major version.If the upgrade is making from routers bricks Mikrotik should remove it ASAP. No one expect from stable version this bricking firmware....
I am not upset that the update did break the router....my decision, my fault.I do not see why all is upset. Yes, its not good that software do breaks the router.I am really angry, as it took me literally 5-6 hours to setup and test...even with linux and dump switch in between did not work with netinstall
Its not silly. After working with Mikrotik for many years, I have seen this several times before.Its very sily to say that I have to read forum before upgrade to stable version...when is new firmware distributed by official upgrade from HAP AC3, the firmware must be checked for this device. How anyone could explain how Mikrotik is testing new stable version when it bricks all HAP AC3?????
tell me who is using HAP AC3 without WifiWave 2 ?????You also need WifiWave 2 packets for it briks.
I do not think all users have launched the update with automatic scripts than just upgrade the device when the update is just publishied... or not?No, you've called some users idiots. That's just bad.
In this case idiots = users that updated to the latest stable v7.
I agree to Jotne: please, if you don´t have the time for recovery or if you have a critical system (this definitly includes your home router) never ever upgrade immediately after a release.Its not silly. After working with Mikrotik for many years, I have seen this several times before.Its very sily to say that I have to read forum before upgrade to stable version...when is new firmware distributed by official upgrade from HAP AC3, the firmware must be checked for this device. How anyone could explain how Mikrotik is testing new stable version when it bricks all HAP AC3?????
MT are not the only one with this problem (Have seen several Cisco devices and other breaks as well).
But since RouterOS is a very complex software with very many function (lot more than most other vendors I have used), there will be situation where some combination has not been tested.
Using wifiwave2 requires ROS7, most hAP AC3 might still be on 6.
tell me who is using HAP AC3 without WifiWave 2 ?????
Updating via script or updating via a manual check makes no difference here.I do not think all users have launched the update with automatic scripts than just upgrade the device when the update is just publishied... or not?No, you've called some users idiots. That's just bad.
In this case idiots = users that updated to the latest stable v7.
The tftp transfer can be checked with Wireshark or other similar tool.For some reason netinstall v7 was always flaky, the device boots, does the tftp transfer but it doesn't show up in netinstall, you have to close netinstall after tftp transfer finishes and reopen netinstall, then the device shows up and you can hit install on it. Smells like a bug.
....I am saying this again.....If you have the craving for instant updates you will not be "screwed" only with MikroTik...
I think you have misunderstood this feature. Which is not surprising.Using the reset-button procedure to recover a perfectly fit for use backup firmware DID NOT WORK (it did with my wAP-ac, I tested in parallel, which still has v6 as backup firmware).
Why did you not put a remark in the changelog to notify new users trying the upgrade of this already known problem?Everyone - We are very sorry for the issue with ww2 package. We are working on a fix for this problem and will publish it as soon as possible.
Agreed, if you use the Mikrotik Pro Android or iOS App, it just states new version available and offers to update now, no changelog, no warning that you could potential end up having to netinstall to bring your device back to active service.Guys thank you for your advices but it would be much easier if Mikrotik did their job. Here you can see, they not even tried to install this STABLE build on their newest router and just roll out the firmaware, so we can imagine how they test old devices. They just release stable version without any testing and wait until sili user install it on their HW. But the worst is that the firmware is still available thorough their update. The HAP AC3 is stil showing the update to 7.2.2.
Could MikroTik please work on a version of NetInstall that runs native on Mac OSX ? 🥺Please, make sure you are using v7.x netinstall, that there is no other active network interfaces on your devices, and IP address configured on Ethernet card and Netinstall are from the same subnet.
Are you able to reinstall any other device?
There's no mention on the wiki of it not being Production Ready: https://help.mikrotik.com/docs/display/ROS/WifiWave2Is the WiFi2 package considered production ready? I was under the impression that it were in the development stage. No, I don't have WiFi2 devices, so I'm not keeping up with this.
Netinstall (on all systems) has problems with systems that have more than one LAN interface. I could not get it to work on my Linux system either.This is exactly what i am doing for last 3 hours. Please, check this post viewtopic.php?t=185628#p930821Unfortunately, to recover the device, you need to reinstall the device using Netinstall utility.
https://help.mikrotik.com/docs/display/ROS/Netinstall
Yep, working fine on Audience.Has already any hero tried to install 7.2.3 on HAP AC3 with wifiwave 2?
...ugh...OK, thank you very much for clarifying things (why the hell is the numbering scheme of OS and firmware the same).I think you have misunderstood this feature. Which is not surprising.Using the reset-button procedure to recover a perfectly fit for use backup firmware DID NOT WORK (it did with my wAP-ac, I tested in parallel, which still has v6 as backup firmware).
"firmware" of a MikroTik router is not the RouterOS. It is just the bootloader (BIOS) of the hardware.
The "reset button procedure" switches you back to the backup bootloader, which handles the problem of "power lost during netinstall" etc.
This has nothing to do with botched RouterOS upgrades, as both the backup firmware and current firmware will still boot the same RouterOS.
THIS is exactly what did NOT work for me...only - and I tried several combinations of OS and with/without dump switch in the middle - netinstall v6.49.6 worked reliably/reproducable (with a switch in middle).Please, make sure you are using v7.x netinstall, that there is no other active network interfaces on your devices, and IP address configured on Ethernet card and Netinstall are from the same subnet.
see this: viewtopic.php?t=185628#p930699 ...I was only successful with a dump switch in the middle (and using v6,49.6 of netinstall)This is exactly what i am doing for last 3 hours. Please, check this post viewtopic.php?t=185628#p930821Unfortunately, to recover the device, you need to reinstall the device using Netinstall utility.
https://help.mikrotik.com/docs/display/ROS/Netinstall
Please send support output file from your router with the configuration, that was upgraded to 7.2.3.Just give a try to update 750G r3 mmips from 7.2.1 to 7.2.3 (create backop before the upgrade)...after I cannot access my router from several LAN connectors, cannot even ping the router form some lan, while the PC gets IP from the router DHCP on that lan. So tried
* restore saved configuration before the upgrade - not help
* reset router to defaults on 7.2.3 and resore saved config - not help
* then downgraded do 7.2.1 (config was probably corrupted at all by that..no accees to router from any LAN connector)
* so again reset to defaults on 7.2.1 and restore saved config - everything works like before!
so for me - keep away from v7.2.3 as well
Sorry, I am back on 7.2.1 in working state now, router is in production. I cannot afford to do all steps again now. What I can do is to send you current working config on 7.2.1 you you can try it by yourself?Please send support output file from your router with the configuration, that was upgraded to 7.2.3.Just give a try to update 750G r3 mmips from 7.2.1 to 7.2.3 (create backop before the upgrade)...after I cannot access my router from several LAN connectors, cannot even ping the router form some lan, while the PC gets IP from the router DHCP on that lan. So tried
* restore saved configuration before the upgrade - not help
* reset router to defaults on 7.2.3 and resore saved config - not help
* then downgraded do 7.2.1 (config was probably corrupted at all by that..no accees to router from any LAN connector)
* so again reset to defaults on 7.2.1 and restore saved config - everything works like before!
so for me - keep away from v7.2.3 as well :-(
It isn't but it is crucial for success.Having more than one interface shouldn't be a problem on Windows, you can disable the ones you don't need before attempting a netinstall.
It's not rocket science, that part.
And it is ridiculous. When you are trying to revive your bricked router, the last thing you want is modifying the config of your still working PC.It isn't but it is crucial for success.Having more than one interface shouldn't be a problem on Windows, you can disable the ones you don't need before attempting a netinstall.
It's not rocket science, that part.
I hear you and while I fully agree, it is what it currently is until they change it.And it is ridiculous. When you are trying to revive your bricked router, the last thing you want is modifying the config of your still working PC.
(it is already bad enough that you need to change the IP address configuration instead of having netinstall add and remove the required static address)
Netinstall should get an interface selection list where you see the interfaces present in your system and select the one you are using.
LOL, yeah, it's like doing brain surgery.[...] the last thing you want is modifying the config of your still working PC [...]
No, I cannot atm.@Hominidae: Can you try what I wrote here? viewtopic.php?p=890726#p890726 (the 2nd half of that post)
Follow the netinstall procedure, and when the device is supposed to show up but it doesn't (after the initial tftp transfer which you can check with wireshark) close netinstall and start it again.
This way it "worked" for me.
It comes down to having a stable IP including interface on the PC...or rather find the point in time when this is the case.What all recent OS (Win + Linux) will do is, that when the ETH link is not physically up and running, it will simply disable that logical IP connection and netinstall will not listen to that connection/port, hence.
I noticed the port on the PC flap, when the hap-ac3 boots...with a switch in the middle, the PC port will stay up, regardless of the state of the link to the MT device.
Yes, but there you just were lucky because your ethernet appears in the interface list before the wifi. As I explained above, with two ethernet it does not work on Linux and you have to plug the cable in "the right port" and know or experiment which one that is. In the old days probably the one that shows up as eth0, but in the current mess caused by mr. Poettering you cannot know which one is the first from looking at the silly "predictable names".I hear you and while I fully agree, it is what it currently is until they change it.
You need to disable all network interfaces expect the one being needed, on Windows.
On Linux not so. I have an old laptop with ethernet and wifi. Last weeks I had "the pleasure" of using it for recovery a couple of times using netinstall.
Didn't have to disable the wifi interface there. It just worked
I remember some time in the past I have succeeded in netinstall only after putting a small dumb desktop switch between the laptop and the device.It comes down to having a stable IP including interface on the PC...or rather find the point in time when this is the case.
Either by examining traffik or simply force a stable connection (and IP for netinstall to lock-in) with a switch in the middle.
https://help.mikrotik.com/docs/display/ROS/NetinstallYeah and regarding recovery, it would be really nice if MT could produce some detailed and working "best practices" for "Netinstall" that is really needed when things fricks up badly...
https://help.mikrotik.com/docs/display/ROS/NetinstallYeah and regarding recovery, it would be really nice if MT could produce some detailed and working "best practices" for "Netinstall" that is really needed when things fricks up badly...
https://help.mikrotik.com/docs/pages/vi ... Id=1409054
There is always para H. - viewtopic.php?t=182373Well, there is "basic" instructions and there is "the realty" with all the weird issues and not to speak of the different platforms. That's why there is room for improvements as in "best practice" (ie specification by example) like for example the setting for "NetBIOS" that should be a part of the standard use as Rextended mentioned
The instructions are not that bad, but you need to have some experience or else at least follow them to the letter and semicolon.
Well, there is "basic" instructions and there is "the realty" with all the weird issues and not to speak of the different platforms.
Out of the window goes your plan to use the RB941-2nD as the preferred routing platform for your enterprise!Just noticed - MPLS subsystem is removed from RB941-2nD
As I explained above, with two ethernet it does not work on Linux and you have to plug the cable in "the right port" and know or experiment which one that is. In the old days probably the one that shows up as eth0, but in the current mess caused by mr. Poettering you cannot know which one is the first from looking at the silly "predictable names".
There is always para H. - viewtopic.php?t=182373
Strange omission though seeing as MPLS is still available on 7.3beta for the rb941Out of the window goes your plan to use the RB941-2nD as the preferred routing platform for your enterprise!Just noticed - MPLS subsystem is removed from RB941-2nD
Well, that little low-power box is still useful for IoT (sorry, obviously a new term for you).Out of the window goes your plan to use the RB941-2nD as the preferred routing platform for your enterprise!Just noticed - MPLS subsystem is removed from RB941-2nD
Yes, looks like it is considered "production ready". Well, at least we have RoS 7.2.3 now.There's no mention on the wiki of it not being Production Ready: https://help.mikrotik.com/docs/display/ROS/WifiWave2
And it's being produced as a Stable package, so if it was not production ready, I'd assume it would only be available for the testing branch?
has anyone reported this? I ran into the same problem, connection working for few minutes, then reboot.Hello .... The problem of rebooting the LHG XL 52 ac has not been resolved since version 7.2 The device starts auto restart minutes after the update .. The device works satisfactorily on version 7.1.5 and below
I've just updated my hAP lite to 7.3beta37 - nope, menus for winbox and webfig indeed exist, just they are not functional, and MPLS is not available via terminal. I somehow like checking new releases on the weakest hardware I have as it's easiest to spot potential problems ;) Maybe it is an omission indeed. And, if intended - an official note should be made.Strange omission though seeing as MPLS is still available on 7.3beta for the rb941
Try this viewtopic.php?p=929964#p929782Upgrading from 7.1.3 to 7.2.3 does not fix the RB5009UG+S+IN port speed mismatch issues with the 2.5Gb port.
I am currious if my issue viewtopic.php?p=930880#p930866 could be related ? (but not using any config in BGP)7.2.2 & 7.2.3 are unable to select the correct ipv4 local address in BGP!!!!
if you leave it automaticly choosed the bgpv4 dosn't do anything and refuses the connection, you have to manuallly write it to let it works!!!
Opened a ticket about it: SUP-81263
regards
Ros
My main reason for going to v7 was wireguard und udp-openvpn!on my main home router I am still using vers 6.48 long term firmware.
I cannot imagine anyone in a business using anything less than long term, which to means, proven over time to be stable.
I do it because my family are the most important clients of all.
I had random reboots with RB2011 and 7.2.x, you're ok?RB1100 and RB2011 upgraded from 7.2.1 without problems
Do you use CAPsMAN? If so, the fix for that problem will be available soon (I'm just testing the pre-release version and it's already fixed there).I had random reboots with RB2011 and 7.2.x, you're ok?RB1100 and RB2011 upgraded from 7.2.1 without problems
It seems like the same issue like mine viewtopic.php?p=931069#p930866 , did you try to connect WinBox via MAC address?My router is RB5009UG+S+IN, when I upgrade to v7.2.3 from v7.2.1, my router is working but I can't open login page of the router in the browser, the winbox can't connect to the router.
How can I fix this problem, can I downgrade to v 7.2.1 from usb?
Many thanks !
As there is so much off-topic, so, here is my 5c...This is know by most people who works in the industry and semi-professional private users BUT not the regular end user for home usage which is a total different matter. Also, a regular home user would never cope with Netinstall.
Yeah but that is only your point of view. Because you apparently are interested in wireguard (only).from my point of view
if MT decide do backport WireGuard to V6 branch, then everything will be OK
if you be kind to read whole postYeah but that is only your point of view. Because you apparently are interested in wireguard (only).
jan/02/1970 00:00:33 system,error,critical error while running customized default
configuration script: bad command name wireless (line 979 column 25)
jan/02/1970 00:00:33 system,error,critical
[admin@MikroTik] > log pri
00:00:20 system,info crossfig will upgrade version 6 configuration
00:00:21 system,info router rebooted
00:00:28 interface,info ether1 link up (speed 1G, full duplex)
00:00:33 system,error,critical error while running customized default configuration script: bad command name wireless (line 979 column 25)
00:00:33 system,error,critical
979 :while ([/interface wireless find default-name="wlan3"] = "") do={
------------------------^ 25 unexpected reless, expected fiwave2
...but this is a bug, that had been introduced and reported for 7.2.1 alreadyConfirm, installed wifiwave2 on Audience, no other problems than the line in the log
[...]
This happen because get-custom-defconf is not maded for wifiwave2 only,
and have internal references to /interface wireless that do not exist on wifiwave2 only devices.
The problem happen also if is added later the wifiwave2 extra package.
At home, I run everything v7. No problems so far, but I am not doing fancy stuff as @ work, like PIM-Routing.My main reason for going to v7 was wireguard und udp-openvpn!
Hope everyone learned a lesson?
@ work, we run everything v6, except 1 device with v7 for WireGuard.
Bad. And now I have bricked x86-64 box, that cant boot. Netinstall does not work on this PC... Very badI tried to upgrade x86-64 with ROS 7.1.2 to 7.2.3 and got cycle reboot after ROS booted... Cant stop this cyclic reboot...
No, is not the same bug, previous bug is (if you write it correctly)...but this is a bug, that had been introduced and reported for 7.2.1 already
Edit: see viewtopic.php?p=929862#p929862
vrf - fixed VRF leaking;
sudo ip route add 255.255.255.255 dev eth1
Netinstall may also work on a PC with multiple interfaces, but it is tricky. Netinstall broadcasts the initial handshake message to 255.255.255.255, so the packets are usually sent to the interface where the default gateway is located. To use Netinstall on a different interface, you need to (temporarily) redirect global broadcasts there. Here is an example for Linux that allows using Netinstall on eth1 without disabling other interfaces:
Code: Select allsudo ip route add 255.255.255.255 dev eth1
We have plans to rework the Netinstall protocol, allowing users to specify the network interface or/and IP subnet of the client IP. For example, if the client IP is 192.168.1.2/24, broadcasts should be made to 192.168.1.255 instead of 255.255.255.255. Unfortunately, some legacy compatibility issues make it harder to implement than it sounds.
That part probably works, but it looks like it is going wrong in the next phase, where the netinstall program has to answer an explicit request from the router.Netinstall may also work on a PC with multiple interfaces, but it is tricky. Netinstall broadcasts the initial handshake message to 255.255.255.255, so the packets are usually sent to the interface where the default gateway is located. To use Netinstall on a different interface, you need to (temporarily) redirect global broadcasts there.
Oh great, thanksDo you use CAPsMAN? If so, the fix for that problem will be available soon (I'm just testing the pre-release version and it's already fixed there).
I had random reboots with RB2011 and 7.2.x, you're ok?
I have just had this issue upgrading from 6.48.6 to 7.2.3 however my experience is that the correct local address is set for eBGP but defaults to the connection Router ID for iBGP connections.7.2.2 & 7.2.3 are unable to select the correct ipv4 local address in BGP!!!!
if you leave it automaticly choosed the bgpv4 dosn't do anything and refuses the connection, you have to manuallly write it to let it works!!!
Opened a ticket about it: SUP-81263
regards
Ros
[...]
Netinstall broadcasts the initial handshake message to 255.255.255.255, so the packets are usually sent to the interface where the default gateway is located.
[...]
[...]Code: Select allsudo ip route add 255.255.255.255 dev eth1
C:\WINDOWS\system32>route print -4 =========================================================================== Interfacce list 8...30 85 a9 98 1f 0b ......Realtek PCIe GBE Family Controller 6...30 85 a9 98 24 51 ......Intel(R) 82579V Gigabit Network Connection 1...........................Software Loopback Interface 1 =========================================================================== IPv4 Route Table =========================================================================== Active routes: Address Mask Gateway Interface Metric [...] 255.255.255.255 255.255.255.255 On-link 127.0.0.1 331 255.255.255.255 255.255.255.255 On-link 192.168.0.195 281 255.255.255.255 255.255.255.255 On-link 192.168.88.3 281 =========================================================================== Permanent routes: Nothing C:\WINDOWS\system32>route DELETE 255.255.255.255 MASK 255.255.255.255 IF 8 OK C:\WINDOWS\system32>route print -4 [...] 255.255.255.255 255.255.255.255 On-link 127.0.0.1 331 255.255.255.255 255.255.255.255 On-link 192.168.88.3 281 [...]Deleted route is recreated again on interface off/on cycle or reboot Windows.
...ah, my bad...yes, you are right....thanks for the clarification.No, is not the same bug, previous bug is (if you write it correctly)
no such item
this time is
bad command name wireless (line 979 column 25)
Excellent solution !Yes, but... depends on who you explain it to... :roll:
Using cmd is not permanent and prevent mess... just reboot the PC for fix errors...
I agree on this, there should be no reason code errors like the ones we're seeing should ever make it into a stable tree, regular users should be safe to assume that a 'stable' release wont lunch their router to the point where they need to factory netinstall them back to a working state, if it means we don't see as frequent updates then I'm all for that as long as what gets released as 'stable' is tested thoroughly, otherwise the term stable is meaningless.Hope everyone learned a lesson?
Well, tell that to the regular home user that never visits the forum. I suggest that MT introduce a new "pre-release" channel which acts kind of a production test to capture serious flaws before moving it to "stable".
I don't expect any action from MikroTik in that direction. Look, they repeatedly refused to take down a clearly broken "stable" versions from the download server and even to include warnings in release notes that "normal users" see when they open their "check for upgrades" window.I agree on this, there should be no reason code errors like the ones we're seeing should ever make it into a stable tree, regular users should be safe to assume that a 'stable' release wont lunch their router to the point where they need to factory netinstall them back to a working state
Well, tell that to the regular home user that never visits the forum. I suggest that MT introduce a new "pre-release" channel which acts kind of a production test to capture serious flaws before moving it to "stable".
Off topic, but I was very surprised when I did see that 7.1.5 thread was closed in favor for 7.2.1.Version 7.1.6 would be the next in long-term branch...
But this is off topic, we should stop here.
We have plans to rework the Netinstall protocol, allowing users to specify the network interface or/and IP subnet of the client IP. For example, if the client IP is 192.168.1.2/24, broadcasts should be made to 192.168.1.255 instead of 255.255.255.255.
Unfortunately, some legacy compatibility issues make it harder to implement than it sounds.
Also, I cannot ping anymore my router. My configs worked just fine for 2 years,.... until this updateI upgraded from 7.2.0 to 7.2.3, and now I cannot access via IP, only Mac address. Tested from winbox for Windows and from android, same results
7.2.1From what version?
And what device ?7.2.1From what version?
RB2011UiAS-2HnDAnd what device ?
7.2.1
What do they look like now? Again these "lookup in table abc" replaced by "lookup in table *0x4000" or similar?after upgrade to 7.2.3 Route rules not working.
local VPN can't see my local networks. my local VPN address is 192.168.22.0/32 and local network is 10.5.51.0/24. before upgrade to 7.2.3 everything working without any issue.What do they look like now? Again these "lookup in table abc" replaced by "lookup in table *0x4000" or similar?after upgrade to 7.2.3 Route rules not working.
That has happened to me before on upgrades, twice in fact. It seems they cannot get that under control.
can you explain about this ("lookup in table *0x4000"), i have same problem, for me i have mangle rule, usually i override with route rule for some ip, for some reason it go to mangle rule route first and ignore the route rule, i decided to downgrade to 7.2.1 and all rule work againWhat do they look like now? Again these "lookup in table abc" replaced by "lookup in table *0x4000" or similar?after upgrade to 7.2.3 Route rules not working.
That has happened to me before on upgrades, twice in fact. It seems they cannot get that under control.
The bare minimum of a stable release should be that it boots without bricking the device, i don't mind if there's some other bug that means I roll back, no stable release should ever crash a device to the point where manual intervention to physically recover the device is required. that is what the 'testing' or 'development' tree is for.I don't install any new version after first others having a go at it. If more first keep an eye on the comments for a time before taking the step and upgrade then a lot of nasty expierences would be avoided.
Atleast Mikrotik did not release just before the weekend and that is already a possitive.
There is no need for that in v7, ROS automatically continues in the chain when you do not reject or accept.does anyone know what is the "passthrough" action to use in routing filters in v7?
Agreed 100%!!!!!The bare minimum of a stable release should be that it boots without bricking the device, i don't mind if there's some other bug that means I roll back, no stable release should ever crash a device to the point where manual intervention to physically recover the device is required. that is what the 'testing' or 'development' tree is for.I don't install any new version after first others having a go at it. If more first keep an eye on the comments for a time before taking the step and upgrade then a lot of nasty expierences would be avoided.
Atleast Mikrotik did not release just before the weekend and that is already a possitive.
There should be a semblance of stability in the 'stable' branch.
Reading the forum to see if a stable update kills your device isn't good enough, If Microsoft bricked 25% of machines with every patch Tuesday people would be screaming, they wouldn't be saying "But did you read the forums?"
For the most part we aren't installing RouterOS on home built hardware (x86 aside) we're installing it on devices Mikrotik have designed, built and pertain to support, if the software doesn't run reliably on their own hardware in the stable branch then whats the point of upgrading at all? How do we know what will or wont kill our devices?
Kind Regards,
Jim.
I am on the same boat as you are. I am sticking for now on version 7.1.5 and work around the IPSec Identities issue with a script for my RB5009. Any other remote devices will remain on v6.x until I see a real stable version. 7.2 was a disaster for me and I got stability with 7.1.5I agree on this, there should be no reason code errors like the ones we're seeing should ever make it into a stable tree, regular users should be safe to assume that a 'stable' release wont lunch their router to the point where they need to factory netinstall them back to a working state, if it means we don't see as frequent updates then I'm all for that as long as what gets released as 'stable' is tested thoroughly, otherwise the term stable is meaningless.
Well, tell that to the regular home user that never visits the forum. I suggest that MT introduce a new "pre-release" channel which acts kind of a production test to capture serious flaws before moving it to "stable".
I started using Mikrotik with the move from 5.x to 6.x (My very first RB for my home router was a RB2011 back in 2013) and I really dont recall the level of issues I have seen with 7.x tree.
I still maintain that 7.1.5 is the best version currently available, 7.2.x has not been an enjoyable venture, hanging back until things are sorted out.
Jim.
And even more important: in the case where things go wrong and errors go into stable releases that brick devices, the IMMEDIATE action should be to hold back such releases while the matter is investigated and rectified.The bare minimum of a stable release should be that it boots without bricking the device, i don't mind if there's some other bug that means I roll back, no stable release should ever crash a device to the point where manual intervention to physically recover the device is required. that is what the 'testing' or 'development' tree is for.
/export show-sensitive
That did it! Thanks. :DUse:
Code: Select all/export show-sensitive
hi,Upgraded the router with memory leak in 7.1.5 as shown here: viewtopic.php?p=922854#p922854
After one day running 7.2.3 it seems to not have memory problem with same config (just upgrade).
This is a generic problem with RouterOS v7: displayed data in winbox not updated correctly over time.using winbox 64-bit v3.35, watching PPP active sessions still shows random connected sessions that are not really there (disconnecting winbox and connecting back doesn't show these phantom sessions)
In / System logging You can alter the default "info" logging rule so it's topics are info and !ovpn
We also have an OVPN with a NAS and the message log is filling me up, so many that it is impossible to see other messages easily
Excuse me.
Can anybody face an issue with latest netinstall-7.2.3 and netinstall-6.48.6 ?
Two of my x86-boxes gone to infinite bootloop after upgrading to 7.2.3. And now netinstall take my boxes to "NBP is too big" error after getting an IP-address from netinstall application.
So I can`t recover my two x86 boxes in any form now...
Any suggestions on this problem?
Yes, this is BUG... you must chose "auto" frequency.After upgrade to 7.2.3 getting "cpu not running on default frequency" on RBD52G-5HacD2HnD (hap ac2).
Bug?
Default frequency is 716MHz. Not "auto" like 488MHz...So this is non logical bug about warning.This is not a bug, is wanted. New default is auto and if changed can be a problem.
I roll back from 7.2.3 to 7.2.0 and now I can login again via local IP. This is definitely a bug.I upgraded from 7.2.0 to 7.2.3, and now I cannot access via IP, only Mac address. Tested from winbox for Windows and from android, same results
How to l3-hw on such big (3.2M routes) routing table?how about l3-hw. have you try and facing some issue?
This is only known for frequent reader. Known issues are forum tribal know-how distributed among thousands of forum entries and hidden behind a poor forum search.Come on, this has already been discussed 20 times before!
YES, the default was changed. YES, devices that were installed from defaults before that change now display a warning. YES, that warning is needlessly alarming.
This is ALL already known.
It works fine for me, RB4011 as L2TP/IPsec client to a v6 router. But I am running 7.2 on the 4011. Did it work for you in that case?L2TP/IPsec as client over PPPoE is still unstable on my RB4011. WinBox session on remote router (6.49.5) over that tunnel is namely reconnecting every minute or so while working stable over SSTP tunnel. But it is much better than on 7.0 when router just rebooted few seconds after start of that L2TP tunnel.
Known issues are forum tribal know-how distributed among thousands of forum entries and hidden behind a poor forum search ...It is no wonder the same issues and bugs are reported and asked about again and again ...
If we are at it, proper release notes would make thighs easier too.If we had a public bug tracker / issues list currently known it would be so much easier.
I recently discussed it (again) but such topics are usually commented with "that would be extra work, would you like us to spend work on that instead of on fixing problems?".If we are at it, proper release notes would make thighs easier too.
"fixed an issue with xy" is less than helpful to decide if it is worth to take the risk of an update.
No, I am using PPPoE as well. Over a VLAN, even. There are bugs in that. But it does not seem to affect my L2TP/IPsec tunnels.My tunnel is over PPPoE tunel required by internet provider and I suspect problem lies here. This seems to be the only difference from your case.
Pre-shared key needs to use same structure as Private/Public key, as far as I know.Can't set wireguard preshared-key, I got "invalid preshared key", a bug?
Thank you.Pre-shared key needs to use same structure as Private/Public key, as far as I know.Can't set wireguard preshared-key, I got "invalid preshared key", a bug?
I could break your config files by messing with the backup array of some important packages... 😈And with every arch Linux update I get I truly know for sure: it may break - but my damn config files keep.
/routing table
add disabled=no fib name=via95
/ip route
add disabled=no distance=50 dst-address=0.0.0.0/0 gateway=192.168.95.1 \
pref-src=0.0.0.0 routing-table=via95 scope=30 suppress-hw-offload=no\
target-scope=14
/routing rule
add action=lookup disabled=no dst-address=192.168.0.0/16 table=main
add action=lookup disabled=no dst-address=0.0.0.0/0 table=via95
/ip firewall mangle
add action=mark-routing chain=prerouting new-routing-mark=via95 passthrough=yes
/routing table
add disabled=no fib name=via95
#RLvia95
add disabled=no fib name=RLvia95
/ip route
# RLvia95
add disabled=no distance=50 dst-address=0.0.0.0/0 gateway=192.168.95.1 \
pref-src=0.0.0.0 routing-table=RLvia95 scope=30 suppress-hw-offload=no\
target-scope=14
/routing rule
add action=lookup disabled=no dst-address=192.168.0.0/16 table=main
#RLvia95
add action=lookup disabled=no dst-address=0.0.0.0/0 table=RLvia95
/ip firewall mangle
add action=mark-routing chain=prerouting new-routing-mark=via95 passthrough=yes
/ip route
# RLvia95
add disabled=no distance=50 dst-address=0.0.0.0/0 gateway=192.168.95.1 \
pref-src=0.0.0.0 routing-table=RLvia95 scope=30 suppress-hw-offload=no\
target-scope=14
furthermore ...suddenly after 4 day uptime, device rebooted with kernel failure
and, WG Private key was lost !!!
it generated a new key
/interface wireguard
add listen-port=8000 mtu=1420 name="WG: MGMN"
/interface wireguard
add listen-port=8000 mtu=1420 name="WG: MGMN" private-key="qAo+KmUIxdAML9d/2GA3nS6/43HgHP3p3lFsnHWTlUw="
Its a default value I never set it.can you explain what is pref-src=0.0.0.0 what thats for?
The ping to what?I noticed that the ping has more than doubled (12ms to 28ms).
In my main server room, my router is on 7.1.x, and last week I reboot it (power cycle) and all my queue configs goes blank. I restore an old backup to fix it, but a reboot should not remove settings.I truly want to understand what is wrong with a config system that loses settings, changes settings by magic wand or corrupts settings. I really don't get it. My Linux machine has config files that were not touched since 2018. And these files did not magically change. Still rocking. And with every arch Linux update I get I truly know for sure: it may break - but my damn config files keep.
I guess there will not be, as it really is a configuration issue. I advise to set MTU and MRU to 1400 in the L2TP server settings, then you don't have to set it for each client.Setting L2TP client's MTU and MRU to 1400 actually make WinBox session stable, thanks for the tip. Not let we wait for official patch of that problem..
That is correct, the RB4011 does not have settable CPU frequency. Older versions showed that, but that was an error that has been corrected.Good afternoon!
Just now I noticed, that the display of the CPU frequency is "lost" after update.
Sorry my mistake. Used speedtest and before the upgrade i has ping times to the test server in the range of 12ms and not it has more than doubled.The ping to what?I noticed that the ping has more than doubled (12ms to 28ms).
Yeah.. right.That is correct, the RB4011 does not have settable CPU frequency. Older versions showed that, but that was an error that has been corrected.Good afternoon!
Just now I noticed, that the display of the CPU frequency is "lost" after update.
Some more stuff.Routing Rules and Marking (maybe a fix)
When you use different routing tables and rules or marks, you will need to copy the "connected" routes for each local interface into those other tables.It looks like if you have a routing table, and routing marks on packets for that routing table.
The packets will go via that routing table, even if the destination IP is one of the addresses on the router.
Upgraded from 7.1.x to 7.2.3 on my RB4011 and it seems that it started crashing once in a while which was not an issue with 7.1.x.
Smells a bit like my earlier issues here: viewtopic.php?t=179071
Meanwhile I can tell that my periodic crashes are back exactly the same as in the other post.The key problem I see in that thread is that you're using on-device logging, so you can't see what led up to the reboot. All you know is the fact of the reboot, not the why of it.
I recommend setting up remote logging so you can get last-dying-gasp type messages from the router. Maybe that won't show anything useful, but it'd be good to know that for a fact rather than guess.
need to find the correct topics to send