Can you please share what was the issue and what is the fixed behaviour?*) ipsec - fixed "sa-src-address" deduction from "src-address" in tunnel mode;
/ip ipsec policy add dst-address=192.168.1.0/24 sa-dst-address=10.155.107.5 sa-src-address=0.0.0.0 src-address=10.155.107.6 tunnel=yes
1 src-address=10.155.107.6/32 src-port=any dst-address=192.168.1.0/24 dst-port=any protocol=all action=encrypt level=require ipsec-protocols=esp tunnel=yes sa-src-address=10.155.107.6 sa-dst-address=10.155.107.5 proposal=default ph2-count=0
Does not prevent users to connect and so far everything works well for us ...*) wireless - added option to disable PMKID for WPA2;
When adding (or importing) a new IPsec policy it automatically used src-address to calculate sa-src-address if it was not specified in tunnel mode, which is not correct. E.g.
Code: Select all/ip ipsec policy add dst-address=192.168.1.0/24 sa-dst-address=10.155.107.5 sa-src-address=0.0.0.0 src-address=10.155.107.6 tunnel=yes
Created:
Code: Select all1 src-address=10.155.107.6/32 src-port=any dst-address=192.168.1.0/24 dst-port=any protocol=all action=encrypt level=require ipsec-protocols=esp tunnel=yes sa-src-address=10.155.107.6 sa-dst-address=10.155.107.5 proposal=default ph2-count=0
What is the Capsman-part of this?*) wireless - added option to disable PMKID for WPA2;
*) wireless - added option to disable PMKID for WPA2;
*) wireless - fixed memory leak when performing wireless scan on ARM;
*) wireless - fixed packet processing after removing wireless interface from CAP settings;
*) wireless - updated "united-states" regulatory domain information;
Even in CAPsMAN! \o/*) wireless - added option to disable PMKID for WPA2;
upgraded almost 300MK devices today and so far without problem
...
or stupid..upgraded almost 300MK devices today and so far without problem
...
Wow, you are brave
@MonkeyDan
This wireless problem has been fixed. If your problem is listed there, it should be okCode: Select all*) wireless - added option to disable PMKID for WPA2; *) wireless - fixed memory leak when performing wireless scan on ARM; *) wireless - fixed packet processing after removing wireless interface from CAP settings; *) wireless - updated "united-states" regulatory domain information;
Just wanted to comment the same.... upgrading 300 devices on the day of release, wow!Wow, you are braveupgraded almost 300MK devices today and so far without problem
...
I once painted myself in the corner by upgrading 2 CCR1009 routers before noticing that further upgrades were impossible due to reduced diskspace, and now I am a little more careful because it took a lot of convincing on the forum before a fix package was released that would solve the problem without doing netinstall... and even that I installed too quickly on one of them. (getting rescued only by the fallback to second partition)or stupid..Wow, you are braveupgraded almost 300MK devices today and so far without problem
...
no, really, I have lack of bad experiences...
How did you upgrade 300 devices?upgraded almost 300MK devices today and so far without problem
already done on types :
crs 328 poe
crs 317
crs 125
crs 226
ccr 1009
ccr 1016
ccr 1032
metal 52SHPacn
metal 2SHPn
rb 2011
rb 3011
mAP lite
In late night I must also done upgrade on main-rt CCR1072, I supposing no problem also
I'm guessing Dude and/or CapsMan.How did you upgrade 300 devices?
Script / tool? Or manual 1 by 1 ?
Hi, I have linux (ubuntu server) with bash script ( which I made myself ) to upgrade all of it.
How did you upgrade 300 devices?
Script / tool? Or manual 1 by 1 ?
This really drove me crazy on my CRS328-24P-4S+ Thanks for fixing!*) crs3xx - fixed tagged packet forwarding without VLAN filtering (introduced in 6.42.6);
After upgrading to 6.42.7, my CRS328-24P-4S+ logs increased CPU temperature from SNMP readouts. The attached graph shows a significant change in CPU temperature after upgrading (at around 23:00). The switch's CPU load (also from SNMP) has not increased.
Were there any changes to the SNMP output for this unit, not mentioned in the changelog? It could look like the temperature of the board was previously listed as CPU temperature.
CRS328-24P-4S+-after-upgrade.png
After upgrading to 6.42.7, my CRS328-24P-4S+ logs increased CPU temperature from SNMP readouts. [...]
For some reason, fan speed is only shown in "System Health" when the fans are running. So during the last screenshot the fans were running, but in the other two the fans were off.Why did 2 fans appear in the last print?
The irony!!upgraded almost 300MK devices today and so far without problem
...
Wow, you are brave
Linux user here. Please, please share the script!Hi, I have linux (ubuntu server) with bash script ( which I made myself ) to upgrade all of it.
Yeah... your avatar is pretty self-explaining.FYI, my nick name is based on a well known product from Czech Republic
Posted in wrong thread, should be in the Winbox thread.There are some GUI differences for ease of use. It is not v6.42.7 specific.
This would be really nice to know!Is there any indication of the attack surface of "vulnerabilities CVE-2018-1156, CVE-2018-1157, CVE-2018-1158, CVE-2018-1159"?
What kind of installations would be directly at risk and require quick updates?
(e.g. admin service ports open on internet, certain types of VPN service open on internet, any configuration with internet connection, etc)
Well after realizing I was more dangerous then helpful I decided to lay low for awhile.Anav, welcome back, been a while since I last saw a post from you.
FYI, my nick name is based on a well known product from Czech Republic
Updated seveal routers, amongst them some RB750Gr3, no such problem noticed.Upgraded RB3011. Ipsec tunnels broken
RB750Gr3,. Ipsec tunnels broken ,
Well, the problem I have is that I now need to update many routers because they may be vulnerable and taken over tomorrow, and thenI do not see problem to release more information later. Maybe it is about to give people some time to upgrade before releasing details.
Think logically,I can confirm that the security fixes were added to the notes after the 6.42.7 thread was already posted! Why was this?
No, this is because CVE entries were only made after the fixes were released.I can confirm that the security fixes were added to the notes after the 6.42.7 thread was already posted! Why was this?
Please send the supout.rif file from 6.42.7 to support@mikrotik.comUpgraded RB3011. Ipsec tunnels broken
RB750Gr3,. Ipsec tunnels broken ,
Reverted back to 42.6 , all ipsec tunnels back working well ,
I hope you understand that when you post a new version AND update the bugfix version at the same time, mentioning 4 security issues that have no description, the users get a bit nervous and assume that an update is urgently required.Well you guys did want a security blog and more awareness. Now you complain that the issues are not serious enough We will try to find a balance.
The issue is you need to provide a little more information please. Adding the CVSS score would be useful, along with information about mitigations (firewall rules!) that, in this case, are most likely already in place.Well you guys did want a security blog and more awareness. Now you complain that the issues are not serious enough We will try to find a balance.
I think the normally preferred procedure is to create CVE numbers and keep them nondescript until fixes are released.Tenable allows vendors to fix issues and release the fixed versions, before they advertise discovered issues.
We could not add non-existing CVE numbers into the changelog.
This is bad. I check for updates every day on this forum. The day this release was posted, I read the full changelog and there was nothing of concern to update all my routers immediately.I can confirm that the security fixes were added to the notes after the 6.42.7 thread was already posted! Why was this?
!) security - fixed vulnerabilities CVE-2018-1156, CVE-2018-1157, CVE-2018-1158, CVE-2018-1159;
!) security - fixed vulnerabilities in WebFig service where already authenticated user could cause service to crash
I confirm this.Webserver Crashed
Telnet service crashed.
In our network we have one RB 3011 that works as a Load Balance and in this routerboard we have one DHCP server on one interface and it stoped working as well after applying 6.42.7. It was installed on top of 6.40.8. Downgraded it to 6.40.8 and then updated it to 6.40.9 and everything is working fine again.Dhcp error still persists as offering lease withour success on CCR1009. After many errors the dhcp stops to answer for a while.
I can confirm this also.I confirm this.Webserver Crashed
Telnet service crashed.
Same here.... Ntp Server stopped aswell. Stayinger at 6.40.9 dir awhileIn our network we have one RB 3011 that works as a Load Balance and in this routerboard we have one DHCP server on one interface and it stoped working as well after applying 6.42.7. It was installed on top of 6.40.8. Downgraded it to 6.40.8 and then updated it to 6.40.9 and everything is working fine again.
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 some problem has appeared on device
I've seen crashed www service on my ARM-based device as well. Already sent supout.rif to support a few days ago. I'm waiting for any response (other thant automated response text). During crash, that service consumes 100% of CPU which is not a huge problem on my hAP ac2 with 4 CPU cores.Please follow the steps mentioned in the change log to let MikroTik fix this issue asap:
I stand with pe1chlI hope you understand that when you post a new version AND update the bugfix version at the same time, mentioning 4 security issues that have no description, the users get a bit nervous and assume that an update is urgently required.Well you guys did want a security blog and more awareness. Now you complain that the issues are not serious enough We will try to find a balance.
I asked for the attack surface, you could have replied "issues in the webserver exploitable by authenticated users" and I would have known that it is not urgent for me to update, and I could have waited until more details became known.
But as the issue was completely unknown and it could well have been in ISAKMP, SSTP, PPP or whatever, I decided to work overtime to update everything.
So indeed it would be welcome when it was a little more clear what is going on.
Normis, don't let them annoy you. This is great progress! And I really love that we finally got CVE numbers!Well you guys did want a security blog and more awareness. Now you complain that the issues are not serious enough We will try to find a balance.
noproblemo@rushlife: Thank you for the scripts!
Finally, thats good to knowIt looks like the default routerboot boot mode has changed from "nand if fail then ethernet" to "try-ethernet-once-then-nand".
This wasn't mentioned in the release note. However, I think it is a good idea, I already set difficult-to-access devices that way.
Advantage: you can netinstall a device by powercycling it while connected to a computer with running netinstall, without having to fiddle with the button.
(which can be very tricky when the device is mounted on a tower)
how fix this please help me !RouterOS version 6.42.7 has been released in public "current" channel!
Before an upgrade:
1) Remember to make backup/export files before an upgrade and save them on another storage device;
2) Make sure the device will not lose power during upgrade process;
3) Device has enough free storage space for all RouterOS packages to be downloaded.
What's new in 6.42.7 (2018-Aug-17 09:48):
MAJOR CHANGES IN v6.42.7:
----------------------
!) security - fixed vulnerabilities CVE-2018-1156, CVE-2018-1157, CVE-2018-1158, CVE-2018-1159;
----------------------
*) bridge - improved bridge port state changing process;
*) crs326/crs328 - fixed untagged packet forwarding through tagged ports when pvid=1;
*) crs3xx - added command that forces fan detection on fan-equipped devices;
*) crs3xx - fixed port disable on CRS326 and CRS328 devices;
*) crs3xx - fixed tagged packet forwarding without VLAN filtering (introduced in 6.42.6);
*) crs3xx - fixed VLAN filtering when there is no tagged interface specified;
*) dhcpv4-relay - fixed false invalid flag presence;
*) dhcpv6-client - allow to set "default-route-distance";
*) dhcpv6 - improved reliability on IPv6 DHCP services;
*) dhcpv6-server - properly update interface for dynamic DHCPv6 servers;
*) ethernet - improved large packet handling on ARM devices with wireless;
*) ethernet - removed obsolete slave flag from "/interface vlan" menu;
*) ipsec - fixed "sa-src-address" deduction from "src-address" in tunnel mode;
*) ipsec - improved invalid policy handling when a valid policy is uninstalled;
*) ldp - properly load LDP configuration;
*) led - fixed default LED configuration for RBLHGG-5acD-XL devices;
*) lte - added signal readings under "/interface lte scan" for 3G and GSM modes;
*) lte - fixed memory leak on USB disconnect;
*) lte - fixed SMS send feature when not in LTE network;
*) package - do not allow to install out of bundle package if it already exists within bundle;
*) ppp - fixed interface enabling after a while if none of them where active;
*) sfp - hide "sfp-wavelength" parameter for RJ45 transceivers;
*) tr069-client - fixed unresponsive tr069 service when blackhole route is present;
*) upgrade - fixed RouterOS upgrade process from RouterOS v5;
*) userman - fixed compatibility with PayPal TLS 1.2;
*) vrrp - fixed VRRP packet processing on VirtualBox and VMWare hypervisors;
*) w60g - added distance measurement feature;
*) w60g - fixed random disconnects;
*) w60g - general stability and performance improvements;
*) w60g - improved MCS rate detection process;
*) w60g - improved MTU change handling;
*) w60g - properly close connection with station on disconnect;
*) w60g - stop doing distance measurements after first successful measurement;
*) winbox - added "secondary-channel" setting to wireless interface if 80 MHz mode is selected;
*) winbox - fixed "sfp-connector-type" value presence under "Interface/Ethernet";
*) winbox - fixed warning presence for "IP/IPsec/Peers" menu;
*) winbox - properly display all flags for bridge host entries;
*) winbox - show "System/RouterBOARD/Mode Button" on devices that has such feature;
*) wireless - added option to disable PMKID for WPA2;
*) wireless - fixed memory leak when performing wireless scan on ARM;
*) wireless - fixed packet processing after removing wireless interface from CAP settings;
*) wireless - updated "united-states" regulatory domain information;
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 some problem has appeared on device
Please keep this forum topic strictly related to this concrete RouterOS release.
If you router has a storage issue (not enough space due to RouterOS, not by other files stored on the device), use package from this link:
https://www.mikrotik.com/download/share/fix_space.npk
- upload package to your router;
- run /system reboot
Other affected installations will be fixed automatically, if there is enough space left for an upgrade by this fix:
"package - free up used storage space consumed by old RouterOS upgrades"
1. Please don't put questions not related to a specific release into the release topichow fix this please help me !
That's the reason:Hello,
I have a big problem for me. I wanted to update RouterOS to v 6.42.7 but without success. Now I have disable Wireless module and my AP's are broken because my CAPsMAN is off.
If I want to enable a Wireless module nothing will hapend after restart
Can you help me?
Thank's
Update.png
Try removing the duplicate package.*) package - do not allow to install out of bundle package if it already exists within bundle;
Looks like netinstall is the only solution:Yes, I tried it as first. But after restart is still Wireless package on it
Try it with a very old netinstall, which has been released before EOL declared on MIPSLE branch, about at version 5.26.I am trying to reset a 532a router board back to factory so I can access it. They recomended using netinstall so I am. However every time I try I am getting an error. bind tftp general failed: Please advise what I can do to resolve this error
Thank you @xvo. I used Netinstall and now is my RB1100AH readyLooks like netinstall is the only solution:Yes, I tried it as first. But after restart is still Wireless package on it
viewtopic.php?f=1&t=133352&p=655190
I have the same problem, noticed on 6.42.6 version. Seems to be only Winbox problem, in console and WebFig it's correct.I noticed that interface "last link up/down times" are in the future.
I can confirm this also.I confirm this.Webserver Crashed
Telnet service crashed.
Thanks Uldis I wait because we find the same problem when I've upgraded from 6.42.3 to 6.42.7,We have found out that sometimes the R11e-LTE modems fail to get an IP address in 3G mode using RouterOS v6.42.7 - we will fix this problem in the next RC and then push that change to the Current release.
That's great news! It happens seldom, but I saw this myself several times. As I had no clue how to reproduce I did not yet mail the support.Thanks Uldis I wait because we find the same problem when I've upgraded from 6.42.3 to 6.42.7,We have found out that sometimes the R11e-LTE modems fail to get an IP address in 3G mode using RouterOS v6.42.7 - we will fix this problem in the next RC and then push that change to the Current release.
1. Please don't put questions not related to a specific release into the release topic, make new topic![time=1535139686 user_id=118648]
how fix this please help me !
2. You are killing your router by Sector Writes, maybe logging on disc? Check your config, stop it!
I found same issue on ccr1016-12G, very intermittent..We have found out that sometimes the R11e-LTE modems fail to get an IP address in 3G mode using RouterOS v6.42.7 - we will fix this problem in the next RC and then push that change to the Current release.
Not for me .My BGP blackholes in both ipv4 and ipv6 are sending out withdrawals and ends are receiving withdrawals.6.42.7 BGP problem
we have 6.42.7 updated this morning.
I'm noticing that our BGP blackhole peer with our upstream is not sending withdrawal from the network list like it used to.
anyone else have issues with this?
Thank you for information, but I not understand why this not tested before a release published!?We are aware of a specific IPsec issue caused by the new improved policy handling in 6.42.7. A fix is already available in the latest release candidate version and will be included in the next current version as well. Very sorry for inconvenience.
Yes, get ERROR: wrong username or password.Have you tried to use WinBox with MAC connection?
No saves password, only type.Just type in again the password, don't use the saves password.
Just curious. Which version of Winbox do you use?Upgrade v6.40.x to v6.42.7 and can't login into system (via ssh, winbox). But CHR work done. Why unmanaged?
Upgrade your winbox to latestHello,
The same thing happens to Kayu. Since I upgraded to v6.42.7 I can not access the routers, it indicates an incorrect username or password ...
I think the problem is for the admin user ... in some routers I only had this user configured and now I can not enter
Mikrotik any solution?
Thanks in advance
Have you cleared the cache?I have updated the wimbox to the latest version and it still fails
Thanks
He's got problems even to log in using ssh so it doesn't seem to be a Winbox-related issue.
@kayu, @XeviMB, it takes time for the ssh to generate a new key on the very first login attempt, and maybe a new one is generated after upgrade, so if you have tried to log in using ssh only once, try again now after some time has elapsed. But it is just a wild guess.
I have an RB750Gr3 in the field running 6.42.7 and ports 1 and 3 keep going up and down.I have port flapping issues on hap ac². Anyone else?
You have made a too-big version upgrade, you should have installed some intermediate versions.I´ve tried remote upgrade into controlled environment (via package menu) on Ccr1009-8g-1s-1s+ from V6.28 to V6.42.7 and lost connectivity so something went bad.
The screen upgrade showed image dowloaded and it was rebooting after that something more.
I´ll going to check on-site. Please any ideas?
Thank you for your fast replay! I´ve checked RN and the community and see nothing related minimum version that is because I proceed. The best I found is here below.You have made a too-big version upgrade, you should have installed some intermediate versions.I´ve tried remote upgrade into controlled environment (via package menu) on Ccr1009-8g-1s-1s+ from V6.28 to V6.42.7 and lost connectivity so something went bad.
The screen upgrade showed image dowloaded and it was rebooting after that something more.
I´ll going to check on-site. Please any ideas?
Tks for your support here!By now the only thing you likely can do is netinstall and reconfigure.
You COULD have repartitioned the device into 2 partitions and copy the old version so you could revert back
But that is too late now.
If it's any help, I have succeeded doing a couple of netinstalls using Windows running in a virtual machine on a Mac.Netconfig is limitted due it´s mandatory Windows PC and we use only Macbooks so another issue here.
Yes, make sense thank you for your contribution. The vendor should take care about this and create easily paths to do things. In this case I need to install Virtual environment + install entire OS just to be compliance to a limited netconfig..it doesn´t make any sense nowadays. Someone from the vendor (engineeer and product managers) must be reviewing that.If it's any help, I have succeeded doing a couple of netinstalls using Windows running in a virtual machine on a Mac.Netconfig is limitted due it´s mandatory Windows PC and we use only Macbooks so another issue here.
So basically you could take a trial of VMware, install W10 in it as a trial, recover your router and then get rid of the hypervisor and windows without spending a dime.
As I already wrote, you could have had that! But on MikroTik it is a user-decision to yield half of the fash capacity to have version rollback.Many others vendors roll-back automatically to previous version and at least give a console warning message please upgrade to x before that. It´s basic!
Thanks for the tip and the link. It was crystal clear on your previous post even it's manually and limited to specific systems so I'm going to start and use when it's possible. From my point of view I highly recommend the vendor implement for all systems by default in order to avoid upgrades issues, downtimes, field hours and the most import eliminate the SPoF. It never should be a user decision in my opinion. Nowadays even 8 ports basic switchs support 2 images (primary-secondary) automatically managed. I really appreciate your time and tips! Tomorrow I'll be on-site fixing the issue.As I already wrote, you could have had that! But on MikroTik it is a user-decision to yield half of the fash capacity to have version rollback.Many others vendors roll-back automatically to previous version and at least give a console warning message please upgrade to x before that. It´s basic!
See this: https://wiki.mikrotik.com/wiki/Manual:Partitions
In short: click on partition, select repartition, set number of partitions to 2.
The router will reboot and now instead of 128MB you will have 64MB of flash.
Go back to partition, select a partition and then "copy to", select the other partition.
Now you do your upgrade. When the router fails to boot, it will try boot from the other partition and you are back
in your old version, which you can copy again to try something else.
Apparently most users do not know that, and it is becoming less interesting as all new low-end routers do not have enough flash space to do this.
However, for your CCR it still was and is an option. I use it on the two CCRs I manage and it already has saved me.
I know my options so you don't need waste time to tell the obvious.@rjj
If you does not like it the way MikroTik do it, you can always go for another vendor like Cisco.
But it would be at a much higher cost.