v6.40.3 [current]

Fri Sep 01, 2017 1:34 pm

To upgrade, click "Check for updates" at /system package in your RouterOS configuration interface, or head to our download page: http://www.mikrotik.com/download

If you experience version related issues, then please send supout file from your router to support@mikrotik.com. File must be generated while router is not working as suspected or after crash.

*) dhcpv6-server - do not release address of static binding from pool after server removal;
*) export - fixed "/system routerboard" export (introduced in 6.40.1);
*) export - fixed export for PoE-OUT related settings;
*) ike1 - fixed initiator ID comparison to NAT-OA;
*) led - fixed "on" and "off" triggers when multiple LEDs are selected;
*) led - fixed RB711UA ether1 LED (introduced in v6.38rc16);
*) lte - do not show USB LTE modem under "/port" menu;
*) lte - fixed ethernet flap when LTE establishes connection;
*) lte - fixed SXT LTE graphs in QuickSet;
*) lte - improved reliability of USB LTE modems;
*) poe-out - fixed router reboot after "poe-out-status" changes;
*) rb1100ahx4 - fixed HW acceleration fragmented packet decryption when fragment is smaller than 64 bytes;
*) rb750gr3 - show warning and do not allow to use "protected-bootloader" feature if "factory-firmware" older than 3.34.4 version;
*) routerboard - added "mode-button" support for RB750Gr3 (CLI only);
*) ssh - do not execute command if it starts with "-" symbol;
*) traffic-flow - fixed reboots when IPv6 address has been set as target address without active IPv6 package;
*) userman - fixed "limitation" and "profile-limitation" update;
*) userman - fixed CoA packet processing after changes in "/tool user-manager router" configuration;
*) webfig - allow to open table entry even if table is not sorted by # (introduced in v6.40);
*) webfig - allow to unset "rate-limit" for DHCP leases;
*) winbox - added possibility to define "comment" for "/routing bgp network" entries;
*) winbox - do not show FAN related information under "/system health" menu for devices which does not have it;
*) winbox - do not show LCD menu for devices which does not have it;
*) winbox - fixed ARP table update after entry changes state to incomplete;
*) wireless - added "russia3" country settings;
*) wireless - added New Zealand regulatory domain information for P2P links;
*) wireless - updated China and New Zealand regulatory domain information;
*) www - fixed unresponsive Web services (introduced in v6.40);

bajodel · Fri Sep 01, 2017 3:42 pm

Updated some test boards on my desk (RB750Gr3, RB951G-2HnD, RBmAP2n), they are working well.
Thanks for this new 'current.point' release.

rsobczak · Fri Sep 01, 2017 3:47 pm

Hi all,
Is already fixed problem with OVPN which (in theory) should be fixed in 6.40.1? Is client getting correct mask provided by server?

Regards,
Radek

Fri Sep 01, 2017 4:02 pm

rsobczak - It was already fixed in 6.40.1 and we have not received complaints any more after that:
viewtopic.php?f=21&t=124247

rsobczak · Fri Sep 01, 2017 4:07 pm

I know that in according to changelog it should be fixed, but when I upgraded two routers RB2011UiAS and tried to use OVPN then all time client was getting /29 mask instead of /24 set up on server. When I downgrade version to 6.38.7 then problem gone. It had/has infuence on OSPF too, because network is "incorrectly" set on both ends of link. I think that this problem may still exists in the newest version...

Fri Sep 01, 2017 4:18 pm

Please send supout file from your router to support@mikrotik.com. Generate file on latest RouterOS version after router has received route with incorrect netmask.

rsobczak · Fri Sep 01, 2017 4:42 pm

Now I have "old" version on both routers (which is working correctly) and I would like to avoid going into troubles again... Already I have no router where I have installed ROS in version 6.40.0 or higher.

vadimbn · Fri Sep 01, 2017 7:20 pm

RB11000AHx4 permanently work with the SPI with this release. Is it normally?

liviu2004 · Fri Sep 01, 2017 7:23 pm

Immediately after update.
1. interface, warning ether6: bridge port received packet with own address as source address (64:d1:54:3c:be:89), probably loop.
There is nothing connected to ether6 (it is master port). Router is RB493G.
2. Connection with wAP ac fails frequently, had to abort connecting to it, I use CAPsMAN.
mac@cap1 disconnected, interface disabled
mac@cap2 disconnected, interface disabled

storp · Fri Sep 01, 2017 7:45 pm

Successfully upgraded two wAP AC, two cAP, one RB951Ui-2nD and one 1100Hx2. Noticed one small glitch, on the 1100Hx2 I have a bonding interface with two GbE ports to a HP Procurve switch (lacp trunk). After upgrade all looked normal but no traffic went through the bonded link. Disabling and thereafter enabling the bonding interface fixed it. Haven't tried another reboot to see if the problem persists but will do when I get the chance.

Update: I've done a reboot on the RB1100Hx2 and now also the bonding interface worked fine. Meaning zero problems so far

zibik · Fri Sep 01, 2017 9:42 pm

There is no security-profile for nv2 protocol in 6.40.x version. So, how to use radius for MAC aurhentication CPE ?

edit:
l find workaround: set proto "nv2 nstreme 802.11" and now you can select correct security-profile.

But what about question ?

Pozdrawiam
Z.K.

eddieb · Fri Sep 01, 2017 9:44 pm

upgraded RB2011UAS (gateway running CAPSMAN), 8x RB962 (CAP), CRS125, RB1100 without noticable problems.

ThunderXP · Sat Sep 02, 2017 9:55 am

Please help with Dude crashing

Technical Info:
modules:
400000-796000 dude.exe
75a90000-75b90000 USER32
76350000-76460000 kernel32
77980000-77b00000 ntdll

VladIgnatyev · Sat Sep 02, 2017 11:53 am

Hi all. I can not update the Mikrotik hAP lite firmware. Error: not enought disk space.

bratislav · Sat Sep 02, 2017 1:15 pm

Hi all. I can not update the Mikrotik hAP lite firmware. Error: not enought disk space.

It worked fine for me ... although I uploaded the npk file to router manually ...
As far as I know files for upgrades are stored temporarily in RAM so there must be something hogging your memory ...
You could use

system resource print

to check how much free memory is left ... and maybe reboot the router and try again ...

VladIgnatyev · Sat Sep 02, 2017 3:47 pm

Hi all. I can not update the Mikrotik hAP lite firmware. Error: not enought disk space.
It worked fine for me ... although I uploaded the npk file to router manually ...
As far as I know files for upgrades are stored temporarily in RAM so there must be something hogging your memory ...
You could use
Code: Select all
system resource print 
to check how much free memory is left ... and maybe reboot the router and try again ...

Thank you! I rebooted the router. Free memory 7.4 MB. I updated the firmware, all ok.

doctorrock · Sat Sep 02, 2017 9:46 pm

Hi, after upgrading, I can't connect my OVPN tunnels any more.

It says "TLS failed" in log messages. From 6.40.2 (client) to 6.40.3 (server) and from 6.40.3 (client) to 6.40.3 (server) , impossible.

I'm OVPN server with CCR1009-7G-1C-1S+ and got different ROS clients , one being a 2011UAS-2HnD. All worked fine in 6.40.2

DMITRYB · Sat Sep 02, 2017 10:02 pm

Immediately after update.
2. Connection with wAP ac fails frequently, had to abort connecting to it, I use CAPsMAN.

Having the same problem after update. I have one hAP AC as capsman manager and 2 wAP ac as capsman clients. Reset does not help. Also noticed that if i enable only one interface (wlan2) as managed and wlan1 and eth1 as discover - works fine. If I make both wlan as managed - i've got troubles not even with hAp ac but with laptops wifi modules - they begin to hang and reboot.

Any suggestions how to make everything work as it was with v6.40.2?

pe1chl · Mon Sep 04, 2017 11:03 am

I have upgraded our CCR1009 from 6.39.2 to 6.40.3 but the (infrequent) problem issuing DHCP leases to Apple devices has remained:

Sep 4 09:27:32 N0212 dhcp-wifi offering lease 192.168.247.53 for 1C:1A:C0:05:F2:16 without success
Sep 4 09:28:06 N0212 dhcp-wifi offering lease 192.168.247.53 for 1C:1A:C0:05:F2:16 without success

(note 1C:1A:C0 is registered to Apple. DHCP lease time is 1 week)

gollandci · Mon Sep 04, 2017 11:45 am

ssh - do not execute command if it starts with "-" symbol;

What does that mean?

sindudas · Mon Sep 04, 2017 2:37 pm

I use a custom default configuration script (installed with NetInstall + "Configure Script" parameter), but since v6.40.x, it stop uploading the Script.

Tried a simple script:

/system package
disable hotspot
disable mpls
disable routing
enable ipv6
enable dhcp
enable ppp

But it does nothing, anything scheduled for enable or disable.

Maybe "Configure Script" function on NetInstall is broken?.
Using the same version of NetInstall that the package installed.

Edit: Tried adding :delay 15s at top of the script and it works.
Whats the problema trying to disable or enable packages at bootup? Or if trying to Add a new PPPoE-Client? is the a better/correct way to solve it different that ":delay 15s" ?

bratislav · Mon Sep 04, 2017 2:54 pm

Thank you! I rebooted the router. Free memory 7.4 MB. I updated the firmware, all ok.

You are welcome ... on the other hand this make me wonder ...
Seems that hAP lite is very tight with RAM memory and that new upgrades can barely fit so problems like yours should be expected in the future ...
Maybe Mikrotik should consider removing advanced routing, hotspot and MPLS packages from combined routeros-smips npk, after all hAP lite is targeted for home use and I doubt anyone using it will ever really need BGP, OSPF ... and if there is someone who really want to push any of those in such a small thingy like hAP lite one could always install them additionally and as needed ...

Mon Sep 04, 2017 4:16 pm

We will see what we can do about this in future, but you do not have to use bundle package if you do not want to do so. You can freely install only those packages which you need and use router with them.

doctorrock · Mon Sep 04, 2017 9:44 pm

Hi.

What about OVPN stopping to work in 6.40.3 ?

I made some investigations, and have my OVPN work back now. OVPN clients on PC log that certificate expired, which obviously is not true (worked in 6.40.2 , as a reminder. And certificates dates are OK, obviously).
To make them work again, I had to set use_crl and crl_download to off , they were both checked before (and worked as-is in 6.40.2)

Also , I noticed what seems to be a bug in OVPN server implementation 6.40.3
Sometimes, relatively randomly , the server stops answering new connections. It seems to still listen on OVPN server configured port, but it answers RST to any SYN it sees.
To fix the problem, I have to turn OVPN server off (and thus loose any OVPN client), and turn it back on. Then, any client will now reconnect with no problem (classic TCP dialog and OVPN).

Never seen such a scenario before 6.40.3

Any ideas ?

xbipin · Mon Sep 04, 2017 10:48 pm

I just upgraded to this and after like 24 hours or so something crashed, wifi signal shows on all devices but none of them can connect, log entry also didn't show anything at all what happened but rebooting it fixed it

niamul · Tue Sep 05, 2017 6:44 am

is it me or is something wrong with ipv6?
2404:ac80::b/64 becomes 2404:ac80::/64.
is that normal?
just upgraded to 6.40.3. should i downgrade?

Tue Sep 05, 2017 8:39 am

xbipin, DMITRYB, liviu2004 - We are sorry for any inconvenience caused. At the moment we do not see any version related issues with wireless and this concrete RouterOS version in our lab. Please generate supout file on your router which has this issue and send it to support@mikrotik.com. Please remember to generate file while problem is actual on device and provide brief problem description so we can try to reproduce the same problem on our lab. Mention from which RouterOS version you did upgrade router to 6.40.3.

Note that version does not have any wireless (only updated country info) related changes so problem might not be directly related to the wireless.

pe1chl · Tue Sep 05, 2017 11:06 am

is it me or is something wrong with ipv6?
2404:ac80::b/64 becomes 2404:ac80::/64.
is that normal?
just upgraded to 6.40.3. should i downgrade?

I don't see such an issue. But you provide too little detail to comment.

pe1chl · Tue Sep 05, 2017 11:08 am

We will see what we can do about this in future, but you do not have to use bundle package if you do not want to do so. You can freely install only those packages which you need and use router with them.

Please plan some method to phase-out bundle package.
(i.e. install separate packages from the factory, in some new version replace bundle package with the separate packages from the bundle)

niamul · Tue Sep 05, 2017 11:40 am

is it me or is something wrong with ipv6?
2404:ac80::b/64 becomes 2404:ac80::/64.
is that normal?
just upgraded to 6.40.3. should i downgrade?
I don't see such an issue. But you provide too little detail to comment.

what information would you like?
ipv6 address add address=2404:ac80::b/64 interface=ether4
when i do that from terminal, the "b" at the end disappears.
tried the same on other routers running older versions and they work fine

edit :: never mind. tried it again after a second reboot and it is working now

bratislav · Tue Sep 05, 2017 12:33 pm

Please plan some method to phase-out bundle package.
(i.e. install separate packages from the factory, in some new version replace bundle package with the separate packages from the bundle)

Why should Mikrotik do that? Packages are already available separately if you prefer it that way, but bundles are actually much more convenient for factory installations and for most people IMHO ...

CompuSoft · Tue Sep 05, 2017 2:38 pm

Ive got an issue with upgrading from 6.38.5 to 6.40.2&3
After upgrade is complete then the server profile/hotspot profile is missing.
If I try to add the profile manual again, then im not allowed to do so.
RouterOs is running architecture: tile
Board Name: CCR1036-12G-4S

Wed Sep 06, 2017 8:13 am

CompuSoft - Sorry for any inconvenience caused. This issue is already fixed in 6.41rc version and we will include fix in upcoming 6.40.x versions. Most likely you can not add configuration again due to this fix:
*) hotspot - require "dns-name" to contain "." symbol under Hotspot Server Profile configuration;

pe1chl · Wed Sep 06, 2017 11:17 am

Please plan some method to phase-out bundle package.
(i.e. install separate packages from the factory, in some new version replace bundle package with the separate packages from the bundle)
Why should Mikrotik do that? Packages are already available separately if you prefer it that way, but bundles are actually much more convenient for factory installations and for most people IMHO ...

Because the new 16-MB flash models are quite tight on storage and with the separate package method you can remove packages that you
do not use (e.g. mpls, hotspot) entirely instead of merely disable them.
It is not so easy to migrate from bundle package to separate packages, if I am right. It would require a netinstall, doesn't it?
I would be in favour of an (optional) mechanism to migrate a router from bundle to separate packages.

Wed Sep 06, 2017 11:59 am

You can install RouterOS version X with bundle package, upload separate packages from another RouterOS version Y and then simply either reboot (if packages are from newer version) or run "/system package downgrade" (if packages are from older RouterOS version).

Bundle package is used in most cases. We can not provide two types of default installation so we have to use one which is the most popular and in this case it is bundle package.

pe1chl · Wed Sep 06, 2017 12:13 pm

You can install RouterOS version X with bundle package, upload separate packages from another RouterOS version Y and then simply either reboot (if packages are from newer version) or run "/system package downgrade" (if packages are from older RouterOS version).

Oh does that work? That makes it simpler, I can make the change at a new version. It was my understanding that it would have to be reloaded using netinstall.

Bundle package is used in most cases. We can not provide two types of default installation so we have to use one which is the most popular and in this case it is bundle package.

I can understand that it was changed in the days when you had to ftp a new version to the router and then it saved some effort, but now with the built-in updater that fetches the required files automatically, who cares if they are bundled?

Wed Sep 06, 2017 12:18 pm

You can install RouterOS version X with bundle package, upload separate packages from another RouterOS version Y and then simply either reboot (if packages are from newer version) or run "/system package downgrade" (if packages are from older RouterOS version).

Is it possible to use eg. ancient wireless packages with current version of ROS ? Old CAPSMAN ? Just asking.

pe1chl · Wed Sep 06, 2017 12:35 pm

Is it possible to use eg. ancient wireless packages with current version of ROS ? Old CAPSMAN ? Just asking.

All packages must be the same version...

Wed Sep 06, 2017 1:02 pm

Please try to keep this topic related to 6.40.3. If you want to discuss package installation process, then you can create new forum topic for such purposes.

Yes, you can upgrade from bundle to separate packages without Netinstall. Yes, all packages installed on router must be from the same RouterOS version.

freemannnn · Wed Sep 06, 2017 2:37 pm

You can install RouterOS version X with bundle package, upload separate packages from another RouterOS version Y and then simply either reboot (if packages are from newer version) or run "/system package downgrade" (if packages are from older RouterOS version).

wow.... good to know that. thanx. i just test that and of course it works.

jrecabeitia · Thu Sep 07, 2017 2:12 pm

I just installed a PowerBox with version 6.40.3 and I notice that it stops responding by IP and does not show mac. There is no way to gain access to i

null31 · Fri Sep 08, 2017 6:44 am

Seems that hAP lite is very tight with RAM memory and that new upgrades can barely fit so problems like yours should be expected in the future ...
Maybe Mikrotik should consider removing advanced routing, hotspot and MPLS packages from combined routeros-smips npk, after all hAP lite is targeted for home use and I doubt anyone using it will ever really need BGP, OSPF ... and if there is someone who really want to push any of those in such a small thingy like hAP lite one could always install them additionally and as needed ...

In hAP lite, the downloaded packages are saved to flash memory and not to RAM...
Today I made a upgrade to rc, have 8MB free for the flash and this is tight to upgrade if have more package installed than that is listed on screenshot.

Screenshot: https://i.imgur.com/iwyGae5.png

Fri Sep 08, 2017 4:10 pm

This problem is not related to topic, but here is an answer. We ensure that hAP lite can be upgraded if there are no files in memory and you use only bundle package.

If you want to use other packages, then you can use separate packages.

Question for you - why do you need, for example, routing package on this device? hAP lite should be used for home or office usage. There is plenty of space for all necessary packages.

pe1chl · Fri Sep 08, 2017 4:14 pm

There is plenty of space for all necessary packages.

That is why I wonder why you use a bundle package. The "mpls" and also "routing", "hotspot" and even "ppp"
packages do not serve a useful purpose in a device mainly used for "wireless access point". When using
separate packages these could be omitted, but in bundle they can only be disabled.

null31 · Fri Sep 08, 2017 6:40 pm

Question for you - why do you need, for example, routing package on this device?

I'm using for dynamic routing with mixed brands for learning purpose. The hap lite is used to do tests with CHR and as access point in trips.
And sorry for off topic.

When using separate packages these could be omitted, but in bundle they can only be disabled.

Yeah, to use separate packages you need to format with netinstall, not everybody knows this trick to reduce the storage use.
Then could be installed separated from factory instead bundle.

pe1chl · Fri Sep 08, 2017 7:25 pm

Of course an improvement to the package management would be possible: the package list should include
all packages available from the server, and offer some "download" button. So, you can add packages simply
by clicking (or using commandline) in the /system packages menu, not by downloading some zip file, unpacking
it, and uploading a subset of those files to the router.
When it is possible to do this when doing "download&update", it should also be feasible to offer it as install option.

Miracle · Sun Sep 10, 2017 6:29 pm

I can not force hotspot user using https after upgrade to 6.40.3.
Both http and https working althought I try disable www service.

JohnTRIVOLTA · Sun Sep 10, 2017 8:38 pm

RB3011 stop sending logs to email when upgrade the board to 6.40.3 "error handling file" .After that i revert the board to 6.39.2 and problem is gone !

Mon Sep 11, 2017 9:42 am

JohnTRIVOLTA - This problem is already fixed in 6.41rc version. However, I recommend to fix the script. E-mail tool does not have a full file name in your script so if there will be two files with similar names, then you will still get an error. For example, now you specify file=test, but actual file name is test.backup. If you will have a file test1.backup and test2.backup, then script will fail anyway. Use full name on e-mail command file=test.backup.

JohnTRIVOLTA · Mon Sep 11, 2017 11:07 am

JohnTRIVOLTA - This problem is already fixed in 6.41rc version. However, I recommend to fix the script. E-mail tool does not have a full file name in your script so if there will be two files with similar names, then you will still get an error. For example, now you specify file=test, but actual file name is test.backup. If you will have a file test1.backup and test2.backup, then script will fail anyway. Use full name on e-mail command file=test.backup.

Thanks Strods, now I understand !

anuser · Mon Sep 11, 2017 11:46 am

Looking at the wireless interfaces of my CAPSMAN enabled "WAP AC" devices: "WMM Enabled" is not enabled?

Reason for asking: I have users complaining about bad battery time when they are connected to our Mikrotik access points, e.g. After 1/2 day the battery on an Galaxy S6 is down when he´s connceted with Mikrotik. While he´s at home, the battery runs for almost 2 days while it is connected to that local wlan access point.

mbx · Tue Sep 12, 2017 11:50 am

Installation and operations appear to run fine after upgrade to 6.40.3.

However, on one CRS125 the system complained about damages ssh host files and required another reboot to generate new host keys.

Furthermore I have noticed that on our devices a new routerboard firmware was available, stating version 3.41.
Is there a changelog available? The wiki included only changed up to 3.40 at https://wiki.mikrotik.com/wiki/RouterBOOT_changelog

Regards,

MBX

Tue Sep 12, 2017 1:18 pm

Looking at the wireless interfaces of my CAPSMAN enabled "WAP AC" devices: "WMM Enabled" is not enabled?

Reason for asking: I have users complaining about bad battery time when they are connected to our Mikrotik access points, e.g. After 1/2 day the battery on an Galaxy S6 is down when he´s connceted with Mikrotik. While he´s at home, the battery runs for almost 2 days while it is connected to that local wlan access point.

WMM is always on if you enable N/AC wireless modes.

vict0r · Tue Sep 12, 2017 3:15 pm

In addition to other problems with 6.40.3. Tested on RB951-2n and RB750r2.
- no DHCP client on ether1 in default config, while should be.
- software and hardware reset to defaul config has different results for LED's on HEX Lite: empty LED table after soft config reset

Tue Sep 12, 2017 4:51 pm

Did DHCP client problem appeared after downgrade from 6.41rc?
We did install 6.40.3 on hEX lite router, did reset configuration and ran "/system led print". Whole configuration was still there.

bajodel · Wed Sep 13, 2017 1:25 am

On RB3011/6.40.3 the sfp interface (DAC cable to CRS125 in my current test) doesn't correctly detect the down status, the CRS125/6.40.3 instead correctly detects if cable unplugged, other device is rebooted (temporary down), etc..
It was the same on 6.40.2 and I'm quite sure also 6.39.2 was affected. Not tested current 6.41rc because I'm not ready for new bridge style jet.

ErfanDL · Wed Sep 13, 2017 10:55 am

SNTP Client not working after update to 6.40.3 on RB2011UiAS-2hND !!!

vict0r · Wed Sep 13, 2017 5:20 pm

Did DHCP client problem appeared after downgrade from 6.41rc?
We did install 6.40.3 on hEX lite router, did reset configuration and ran "/system led print". Whole configuration was still there.

No downgrade. Clean router after netinstall 6.40.3 has same problem. Routerboard BIOS version has been updated, but without positive result.
About HEX Lite, problem appears when doing reset configuration via Winbox.

I did several updates and downgrades - 6.40.3 has described problem.

anuser · Wed Sep 13, 2017 6:34 pm

WMM is always on if you enable N/AC wireless modes.

- On a CAPSMAN enabled WAP AC:
/interface wireless
print advanced
gives:
"wmm-support=diabled"

- When I disable CAPSMAN for wlan0 and wlan1 interfaces on the access point and do:
/interface wireless
set 1 wmm-support=enabled
set 0 wmm-support=enabled

- After that enable CAPSMAN for wlan0 and wlan 1 interfaces and do:
/interface wireless
print advanced
gives: "wmm-support=enabled"

So do I have to enable wmm-support by hand before activating CAPSMAN on the access points?

IanEKB · Thu Sep 14, 2017 2:50 pm

Hello. I had had a problem with updating 6.40 -> 6.40.3 with transmission voip traffic (tcp 5060 port?). I couldn't be connected to voip provider (time out).
I use FreePBX 13. At connection directly there are no problems. Ports are forwarded and before everything worked (mode: netmap).

I have 2 ISP - the problem was on both.

I haven't found the reason.

Fix: donwnrgade to 6.40, restore backup config, switching ISP (on 6.40.3 hasn't helped).

Sorry for my english.

vytuz · Fri Sep 15, 2017 12:58 pm

Hello, I load custom scripts with netinstall and worked normally until 6.40.x. From version 6.40.x after reset, flashed scripts do not load. In some models script do not load after resetting with no default configuration, and reseting normally again. In some models (even same model and same script) after normal reset (button reset or reset configuration) router starts with no configuration ar all. What was changed with script load after reset? Basic script sample (951Ui):

/interface bridge
add admin-mac=[/interface ethernet get 1 mac-address] auto-mac=no name=bridge-local
/interface wireless
set [ find default-name=wlan1 ] band=2ghz-b/g/n channel-width=20/40mhz-Ce disabled=no distance=indoors \
    frequency=auto mode=ap-bridge rx-chains=0,1 tx-chains=0,1 adaptive-noise-immunity=ap-and-client-mode \
     wireless-protocol=802.11 wmm-support=enabled
/interface ethernet
set [ find default-name=ether3 ] master-port=ether2
set [ find default-name=ether4 ] master-port=ether2
set [ find default-name=ether5 ] master-port=ether2
/ip pool
add name=dhcp ranges=192.168.88.50-192.168.88.250
/ip dhcp-server
add address-pool=dhcp disabled=no interface=bridge-local name=default lease-time=1d
/interface bridge port
add bridge=bridge-local interface=ether2
add bridge=bridge-local interface=wlan1
/ip address
add address=192.168.88.1/24 comment="default configuration" interface=ether2 network=192.168.88.0
/ip dhcp-client
add comment="default configuration" dhcp-options=hostname,clientid disabled=no interface=ether1
/ip dhcp-server network
add address=192.168.88.0/24 comment="default configuration" gateway=192.168.88.1
/ip dns static
add address=192.168.88.1 name=router
/ip firewall filter
add action=fasttrack-connection chain=forward connection-state=established,related	
add chain=input in-interface=ether1 protocol=udp dst-port=53 action=drop
add chain=input in-interface=ether1 protocol=tcp dst-port=53 action=drop
add chain=forward protocol=udp dst-port=53 out-interface=!ether1 action=drop
add chain=forward protocol=tcp dst-port=53 out-interface=!ether1 action=drop
add chain=input comment="default configuration" connection-state=established,related
add chain=input comment="default configuration" protocol=icmp
add action=drop chain=input comment="default configuration" in-interface=ether1
add chain=forward comment="default configuration" connection-state=established,related
add action=drop chain=forward comment="default configuration" connection-state=invalid
add action=drop chain=forward comment="default configuration" \
    connection-nat-state=!dstnat connection-state=new in-interface=ether1
/ip firewall nat
add action=masquerade chain=srcnat comment="default configuration" out-interface=ether1
/tool bandwidth-server
set enabled=no

What should be changed to work after reset?

mrz · Fri Sep 15, 2017 1:00 pm

Add delay at the top of your script or loop check if interface exists.

bajodel · Mon Sep 18, 2017 1:59 pm

On CRS326-24G-2S+ / ros 6.40.3 and RouterBOOT booter 3.41, the memory test fail (tested on 2 different board); I guess the test is broken otherwise I've two defecting board memory.

## choice "t" (do memory testing)

Error in address=0x016FFEEC, W=0x016FFEEC R=0x01009C5C X=0x006F62B0
Error in address=0x016FFEF0, W=0x016FFEF0 R=0x01002C04 X=0x006FD2F4
Error in address=0x016FFEF8, W=0x016FFEF8 R=0x20000000 X=0x216FFEF8

(see details:)

BootROM 1.41
Booting from SPI flash
BootROM: Image checksum verification PASSED

RouterBOOT booter 3.41

CRS326-24G-2S+

CPU frequency: 800 MHz
  Memory size: 512 MiB
 Storage size:  16 MiB

..[CUT]..

your choice: i - board info

Board Info:

        Board type: CRS326-24G-2S+
     Serial number: 763C075929FE
  Firmware version: 3.41
     CPU frequency: 800 MHz
       Memory size: 512 MiB
      Storage size: 16 MiB
        Build time: 2017-08-09 09:58:46
  eth1 MAC address: 64:D1:54:C6:C4:F0
 
..[CUT]..

your choice: t - do memory testing

This may take some time, ESC aborts testing

Fill memory with addresses:
  Filling memory...
   Error in address=0x016FFEEC, W=0x016FFEEC R=0x01009C5C X=0x006F62B0
   Error in address=0x016FFEF0, W=0x016FFEF0 R=0x01002C04 X=0x006FD2F4
   Error in address=0x016FFEF8, W=0x016FFEF8 R=0x20000000 X=0x216FFEF8

tangram · Mon Sep 18, 2017 2:44 pm

Hi,

I have a big problem updating from 6.39.2 to 6.40.3. I use ovpn to connect locations and after upgrade in don't see correct local/remote addresses and routes are not installed:
Because of this dynamic routing fails and network breaks down.
Addresses look like this:
6 D 10.91.21.254/24 10.91.21.0 OVPN
Instead of:
6 D 10.91.21.254/32 10.250.1.10 OVPN

Routing table:
8 ADC 10.91.21.0/24 10.91.21.254 OVPN 0
Instead of:
10 ADC 10.250.1.10/32 10.91.21.254 OVPNBCA 0

Where 10.250.1.10 is the server and 10.91.21.254 is the client.
I returned to 6.39.2, please fix this - 6.40.2 and 6.40.1 are the same.

skylark · Mon Sep 18, 2017 3:43 pm

On CRS326-24G-2S+ / ros 6.40.3 and RouterBOOT booter 3.41, the memory test fail (tested on 2 different board); I guess the test is broken otherwise I've two defecting board memory.

## choice "t" (do memory testing)

Error in address=0x016FFEEC, W=0x016FFEEC R=0x01009C5C X=0x006F62B0
Error in address=0x016FFEF0, W=0x016FFEF0 R=0x01002C04 X=0x006FD2F4
Error in address=0x016FFEF8, W=0x016FFEF8 R=0x20000000 X=0x216FFEF8

(see details:)
Code: Select all
BootROM 1.41
Booting from SPI flash
BootROM: Image checksum verification PASSED

RouterBOOT booter 3.41

CRS326-24G-2S+

CPU frequency: 800 MHz
  Memory size: 512 MiB
 Storage size:  16 MiB

..[CUT]..

your choice: i - board info

Board Info:

        Board type: CRS326-24G-2S+
     Serial number: 763C075929FE
  Firmware version: 3.41
     CPU frequency: 800 MHz
       Memory size: 512 MiB
      Storage size: 16 MiB
        Build time: 2017-08-09 09:58:46
  eth1 MAC address: 64:D1:54:C6:C4:F0
 
..[CUT]..

your choice: t - do memory testing

This may take some time, ESC aborts testing

Fill memory with addresses:
  Filling memory...
   Error in address=0x016FFEEC, W=0x016FFEEC R=0x01009C5C X=0x006F62B0
   Error in address=0x016FFEF0, W=0x016FFEF0 R=0x01002C04 X=0x006FD2F4
   Error in address=0x016FFEF8, W=0x016FFEF8 R=0x20000000 X=0x216FFEF8

Hi,

We found this test do not work correctly and most likely device memory works without errors. It will be fixed in next versions.

bajodel · Mon Sep 18, 2017 4:02 pm

On CRS326-24G-2S+ / ros 6.40.3 and RouterBOOT booter 3.41, the memory test fail (tested on 2 different board); I guess the test is broken otherwise I've two defecting board memory.
..[CUT]..
Hi, we found this test do not work correctly and most likely device memory works without errors. It will be fixed in next versions.

Ok, one of these CRS326 will stay on my test lab so I'll be able to test/stress it and we'll see.

balanila · Mon Sep 18, 2017 5:04 pm

Hi
I have CRS125-24G-1S
After update to 6.40.3 seems it ok, but then I selected from winbox "Switch -> Settings" then device is rebooted. After reboot I see in logs: "router rebooted because some critical program crashed"
and after 10-15 second vpn tunnels disconnected with message: "...-system is going to shutdown" and router rebooted in cycle again and again.

If I rebooted manualy from system - reboot - router rebooted once and working properly.
If I selected Switch -> Settings... router rebooted with message "router rebooted because some critical program crashed"

Tue Sep 19, 2017 4:22 pm

A wireless CPE still consideres only the strongest 2 APs with the same SSID for connection. This changed after 6.35.4. So if I block a CPE with access-list to connect to 2 APs it does not connect to the third. Not mentioned in the change log so I guess this is a bug.

Tue Sep 19, 2017 4:25 pm

ste, enabled wireless,debug logging and check the logs. Wireless station will always try to connect to the strongest signal AP unless you have some specific Connect-list rules. Also note that we have station-roaming setting in newer versions.

Tue Sep 19, 2017 4:46 pm

ste, enabled wireless,debug logging and check the logs. Wireless station will always try to connect to the strongest signal AP unless you have some specific Connect-list rules. Also note that we have station-roaming setting in newer versions.

Yes. I see it tries to connect to the strongest 2 APs but did not succeed as I blocked it with access-lists at the AP. It does not consider the 3. AP. I have situations where I want to route CPEs to APs which do not have the strongest signal. This worked up to 6.35.4. So downgraded the cpe and it reconnects.

vytuz · Tue Sep 19, 2017 5:03 pm

Still hoping You will solve Mikrotik 2.4G Band and Qualcomm chip based phones connection issue. Mikrotik transmission rate connects maximum 54Mbps, G-standard. Rx rate connects normally, as fast as phone supports. Mikrotik connects to other brand cards, samsung phones i.e. normally, but not qualcomm chip phones. These phones connects to other routers at full speed. Even then other routers uses same chips. Mikrotik wifi chip configuration need to be tuned. Tested on QCA9531, AR9344.

ksuuk · Wed Sep 20, 2017 8:55 am

JohnTRIVOLTA - This problem is already fixed in 6.41rc version. However, I recommend to fix the script. E-mail tool does not have a full file name in your script so if there will be two files with similar names, then you will still get an error. For example, now you specify file=test, but actual file name is test.backup. If you will have a file test1.backup and test2.backup, then script will fail anyway. Use full name on e-mail command file=test.backup.

Hi, sorry about hijacking this thread, I have similar issue, so I removed from backup script global enviroment and used full file name:

# Old script:
:global mikrotikexport export_backup
/export file=$mikrotikexport
:delay 10s
/tool e-mail send to="user@domain.com" subject=([/system identity get name] . \
" Export Backup " . [/system clock get date]) file=$mikrotikexport
:delay 10s
/file remove $mikrotikexport

# New script:
/export file=mikrotikexport.rsc
:delay 10s
/tool e-mail send to="user@domain.com" subject=([/system identity get name] . \
" Export Backup " . [/system clock get date]) file=mikrotikexport.rsc
:delay 10s
/file remove mikrotikexport.rsc

Tested the new script in gui - run and in cli - /system script run Backup-script and it worked, but it does not work when sheduled, giving the same error "Error sending e-mail <Backup sep/19/2017>: error handling file" All policy's are allowed except romon and dude. Do I still need add global enviroment to the full name or something else?

How to fix this?

Edit: Strange, today new config worked.

fanMikroTik · Wed Sep 20, 2017 3:41 pm

Still hoping You will solve Mikrotik 2.4G Band and Qualcomm chip based phones connection issue. Mikrotik transmission rate connects maximum 54Mbps, G-standard. Rx rate connects normally, as fast as phone supports. Mikrotik connects to other brand cards, samsung phones i.e. normally, but not qualcomm chip phones. These phones connects to other routers at full speed. Even then other routers uses same chips. Mikrotik wifi chip configuration need to be tuned. Tested on QCA9531, AR9344.

I have the same issue rb951ui-2hnd + 4x 802.11n smartphone

Thu Sep 21, 2017 6:57 pm

A wireless CPE still consideres only the strongest 2 APs with the same SSID for connection. This changed after 6.35.4. So if I block a CPE with access-list to connect to 2 APs it does not connect to the third. Not mentioned in the change log so I guess this is a bug.

Somewhat related to your post ...
You might want to read this Mikrotik forum topic:
viewtopic.php?f=7&t=124884&e=1&view=unread#unread

This is a discussion about AP access-lists
Currently, this new feature/option is only in RC at this time at the CLI interface (not winbox).

Quick info about this topic - how to better control multiple clients connecting to multiple APs (same SSID).
Gives you something to think about & if you have any ideas on this topic, you are welcome to post your thoughts there also ...

North Idaho Tom Jones

petterg · Mon Sep 25, 2017 5:32 pm

When running reset-configuration from system menu in winbox all files are removed. Is this a bug? If this is not a bug, is there a chance to go back to the old practice where you could do a backup, reset config, and keep the backupfiles remained in the unit. (Also it seems pointless that the dialog has the option "Do not backup" in the reset dialog if files are deleted anyway.)

This happens on hAP ac and wAP ac (not tested 6.40.3 in other units)

mrz · Mon Sep 25, 2017 5:41 pm

Files are not removed if they are stored in flash. wap/hap ac by default stores files in ram disk, so on every reboot they are lost.

howdey57 · Mon Sep 25, 2017 11:56 pm

Hi All,

I installed 6.40.3 over the weekend and seem to have a problem.

I use passthrough firewall rules to track Bytes used by IP addresses in an Address List.

Before the update I was using about 2 GB per day, but since the upgrade that has fallen to approx 10MB. I know I am using more!

Has anyone noticed something similar.

Charles

ziegenberg · Tue Sep 26, 2017 12:31 am

Hi!

I installed 6.40.3 over the weekend and seem to have a problem.
I use passthrough firewall rules to track Bytes used by IP addresses in an Address List.
Before the update I was using about 2 GB per day, but since the upgrade that has fallen to approx 10MB. I know I am using more!
Has anyone noticed something similar.

Could it be, that you accidentally use of fasttrack?

howdey57 · Tue Sep 26, 2017 8:40 am

Thanks for the response.

Maybe I have. I will investigate.

howdey57 · Tue Sep 26, 2017 2:21 pm

I know. I originally put it as a question that i could find the answer to myself, then thought better of it but couldn't delete it. I then changed it to the anodyne response above!

My bad.

-end-

MarkoB · Tue Sep 26, 2017 11:50 pm

Still hoping You will solve Mikrotik 2.4G Band and Qualcomm chip based phones connection issue. Mikrotik transmission rate connects maximum 54Mbps, G-standard. Rx rate connects normally, as fast as phone supports. Mikrotik connects to other brand cards, samsung phones i.e. normally, but not qualcomm chip phones. These phones connects to other routers at full speed. Even then other routers uses same chips. Mikrotik wifi chip configuration need to be tuned. Tested on QCA9531, AR9344.

I have this problem too with MediaTek Helio X10 MT6795, TX rate max 54, RX 72 or 150 with 40mhz

Wed Sep 27, 2017 12:57 am

Still hoping You will solve Mikrotik 2.4G Band and Qualcomm chip based phones connection issue. Mikrotik transmission rate connects maximum 54Mbps, G-standard. Rx rate connects normally, as fast as phone supports. Mikrotik connects to other brand cards, samsung phones i.e. normally, but not qualcomm chip phones. These phones connects to other routers at full speed. Even then other routers uses same chips. Mikrotik wifi chip configuration need to be tuned. Tested on QCA9531, AR9344.
I have this problem too with MediaTek Helio X10 MT6795, TX rate max 54, RX 72 or 150 with 40mhz

I want to get a better understanding of the problem you are experiencing.

Question - Is the problem (to you) that you can not connect up a client device (that uses a Qualcomm chip) to a Mikrotik AP ?

If then is the issue, does it appear the problem is the wireless connection -or- the device does not get an IP address after the connection ?

Some things to try re the wireless:

Channel Width: 20/40MHz Ce
Band: 2GHz-B-G-N (or even better if you do not have old devices 2GHz G-N)
Wireless Protocol: 802.11
Frequency: Is the frequency used dark in the drop-down list of frequencies ?
Data Rates: default
Advanced: Distance: indoors
Advanced: Hw Retries: 15
Advanced: Preamble Mode: both
Advanced: Disconnect Timeout: 00:00:15
Advanced: On Fail Retry Time: 1.00

DHCP server: Authorative: after 2s delay

Also - try changing your security profile: to WPA-PSK -and-also- WPA2-PSK (use same key for both).

AP access-list (nothing in the access-list)

If anything here works/helps - please post the results.

North Idaho Tom Jones

eine · Wed Sep 27, 2017 1:27 pm

Changelog:
[...]
lte - improved reliability of USB LTE modems;

May I ask about more details?

Fri Sep 29, 2017 4:23 pm

Brand new device RB1100AHx4 Dude version... strange for me:

RB1100Dx4.PNG

BTW: Board name is RB1100AHx4 Dude Edition but in "Neighbours" it is detected as RB1100Dx4

Fri Sep 29, 2017 4:28 pm

Marketing name - RB1100AHx4 Dude Edition
Product code - RB1100Dx4

Same as:
Marketing name - hAP ac
Product code - RB962UiGS-5HacT2HnT

Fri Sep 29, 2017 4:36 pm

Thank you for answer ... naming is not a problem however why couldn't you name this board RB1100Dx4 for consistency .... it is rethorical question

Unknown devices are strange for me.

Fri Sep 29, 2017 4:59 pm

For the starters - this is 6.40.3 version topic. None of these questions are related to the RouterOS version so please create new topic for each individual topic, write in other appropriate topics or contact support.

Same reason why hAP ac is not RB962UiGS-5HacT2HnT - we have selected such marketing name for this particular 1100 series device.

As for unknown devices - they have been unknown since the release of the device. We are still working on improvements for this and other routers.

fanMikroTik · Sat Sep 30, 2017 11:24 am

Still hoping You will solve Mikrotik 2.4G Band and Qualcomm chip based phones connection issue. Mikrotik transmission rate connects maximum 54Mbps, G-standard. Rx rate connects normally, as fast as phone supports. Mikrotik connects to other brand cards, samsung phones i.e. normally, but not qualcomm chip phones. These phones connects to other routers at full speed. Even then other routers uses same chips. Mikrotik wifi chip configuration need to be tuned. Tested on QCA9531, AR9344.
I have this problem too with MediaTek Helio X10 MT6795, TX rate max 54, RX 72 or 150 with 40mhz
I want to get a better understanding of the problem you are experiencing.

Question - Is the problem (to you) that you can not connect up a client device (that uses a Qualcomm chip) to a Mikrotik AP ?

If then is the issue, does it appear the problem is the wireless connection -or- the device does not get an IP address after the connection ?

Some things to try re the wireless:

Channel Width: 20/40MHz Ce
Band: 2GHz-B-G-N (or even better if you do not have old devices 2GHz G-N)
Wireless Protocol: 802.11
Frequency: Is the frequency used dark in the drop-down list of frequencies ?
Data Rates: default
Advanced: Distance: indoors
Advanced: Hw Retries: 15
Advanced: Preamble Mode: both
Advanced: Disconnect Timeout: 00:00:15
Advanced: On Fail Retry Time: 1.00

DHCP server: Authorative: after 2s delay

Also - try changing your security profile: to WPA-PSK -and-also- WPA2-PSK (use same key for both).

AP access-list (nothing in the access-list)

If anything here works/helps - please post the results.

North Idaho Tom Jones

I have this problem too and it is very easy to understand...

All android smartphones with 802 N that I have allready connected to my rb951ui-2hnd with great signal use only RX Rate: 54 Mbps but TX Rate 72 Mbps and more... Why???

All my notebooks are without problems (TX and RX Rate: 300Mbps-40MHz/2S/SGI)

There is no way to change it - All android smartphones with 802 N use only RX Rate 54 Mbps and I don't know why???

duronus · Mon Oct 02, 2017 10:06 am

After update to 6.40.3 some device disappeared from the network and ARP is seen IP : 00:00:00:00:00
maybe I missed something?

После обновления на 6.40.3 некоторые устройства пропали из сети и в ARP видны как IP:00:00:00:00
Может это не глюк, а это я чего то не знаю?

Chupaka · Mon Oct 02, 2017 11:05 am

After update to 6.40.3 some device disappeared from the network and ARP is seen IP : 00:00:00:00:00
maybe I missed something?

После обновления на 6.40.3 некоторые устройства пропали из сети и в ARP видны как IP:00:00:00:00
Может это не глюк, а это я чего то не знаю?

00:00:00:00:00:00 are just incomplete entries, there was no ARP reply received from them (no letter 'C' in Flags in WinBox)

travisms · Mon Oct 02, 2017 4:33 pm

I have encountered an issue with the RB912 (Basebox 5) and Sierra Wireless MC8705 where after the lte connection has connected and got a DHCP address, the lte interface will disappear and reappear and then not reconnect. Log reflects as system, info item added twice. I downgraded to 6.38.7 and am no longer experiencing the issue.

Kraken2k · Mon Oct 02, 2017 6:11 pm

EoIP and IPsec tunnels - saga continues

at least I hope I'm not making any basic mistake. If so please tell me:

I encounter problems using the combination of above (IPsec site-to-site tunnel, using policies and EoIP tunnel inside this one). When EoIP tunnel (using IPs of LAN interfaces in both sites as Local and Remote address on both sides) is created before the IPsec tunnel, then the encapsulated traffic of EoIP tunnel is not processed through IPsec Policies and sent to WAN interface (and there blocked by firewall output chain of course, because the source IP is private address).

The problem is, that besides rebooting the router there is no way how to fix this state - I tried disable/enable EoIP tunnel, IPsec Policy, IPsec peer, flush Installed SAs... nothing helped. IPsec policies are not applied just for the EoIP encapsulated traffic (I see blocked packets of protocol 47 to wan interface), everything else (like ping from router to other site LAN, local traffic routed to the IPsec tunnel...) works normally.

It's easy to reproduce: just try to boot with IPsec disabled and EoIP enabled. After you enable the IPsec peer and Policies, encapsulated EoIP traffic will ignore IPsec Policies until next reboot, but all other traffic will flow through IPsec tunnel normally.

Routers are RB1100AHX2 and RB2011UiAS-2HnD, both with RoS v6.40.3 and latest routerboard firmware updates.

Mon Oct 02, 2017 7:34 pm

When EoIP tunnel (using IPs of LAN interfaces in both sites as Local and Remote address on both sides) is created before the IPsec tunnel, then the encapsulated traffic of EoIP tunnel is not processed through IPsec Policies and sent to WAN interface (and there blocked by firewall output chain of course, because the source IP is private address).

Your EoIP traffic should normally be catched by IPsec policy and be blocked until phase 2 comes up. If you are behind NAT and have to use some "magic", try simply blocking the unencrypted EoIP traffic using firewall.

null31 · Mon Oct 02, 2017 9:15 pm

There is no way to change it - All android smartphones with 802 N use only RX Rate 54 Mbps and I don't know why???

Android energy saving, to me is the same, except to 802.11ac which is around of 97~433 Mbps.

fanMikroTik · Tue Oct 03, 2017 10:25 am

These (Qualcomm chip based) phones connects to other routers at full speed. Even then other routers uses same chips.
Mikrotik TX Rate connects maximum 54Mbps, G-standard. Rx rate connects normally, as fast as phone supports.
Mikrotik wifi chip configuration need to be tuned.
Mikrotik connects to other brand cards, samsung phones i.e. normally, but not qualcomm chip phones.
Please MikroTik have a look at it

Tue Oct 03, 2017 2:43 pm

Version 6.40.4 has been released:
viewtopic.php?f=21&t=126307

Who is online