6.9 released!

Fri Jan 31, 2014 4:03 pm

What's new in 6.9 (2014-Jan-31 11:18):

*) lcd - added option to change the color-scheme;
*) updated bootloader firmware;
*) ppp: fixed RADIUS accounting;
*) ppp: fixed IPV6-Prefix assigning;
*) ppp: fixed dial-on-demand;

What's new in 6.8 (2014-Jan-29 15:52):

*) bridge - default protocol-mode changed to RSTP for new bridges,
fixed bridge mac address changing when port (with lowest mac address) goes down
*) userman - improve startup time;
*) sstp client - support server name verification from certificate;
*) wireless - improved 11n and nv2 stability;
*) dhcp client - support interfaces in bridge;
*) dhcp - parse decimal strings and IP addreses in options value;
*) bgp - don't show community 'internet' in BGP advertisements;
*) ipsec - enable hardware acceleration for aes-cbc + md5|sha1|sha256 aead on tilera cpus;
*) ospf - fixed checksum calculation for OSPFv3 AS-external-LSAs;
*) default configuration - changed dhcp server lease time to 10 minutes;
*) fixed port isolation on CRSs (bug introduced in v6.6);
*) smb - added support for SMB 2.002
*) timezone information updated;
*) ppp - fixed ppp bridging (did not work since v6.6);
*) improved speed of PPP, PPPoE, PPTP & L2TP on multicore routers;
*) address-list - fix crash when adding two identical address list entries (not allowed any more);
*) fixed multicast forwarding on CCRs;
*) firewall - improved address-type matcher, and added it for ipv6 aswell;

Simply push "Check for Updates" in your router,
or download from http://www.mikrotik.com/download

rmmccann · Fri Jan 31, 2014 4:24 pm

If I'm reading correctly, from now on any bridge I create will have RSTP enabled by default?

If that's the case, I certainly hope the priority has been defaulted to F000 than as well. I can see problems with this implementation on xSTP enabled networks. IMO, this was fine the way it was before and should be a setting we choose to enable.

MrYan · Fri Jan 31, 2014 4:28 pm

*) ppp - fixed ppp bridging (did not work since v6.6);

This is now working for me again. It would be nice however if the PPP interface added to the bridge didn't show as '(unknown)' in /interface bridge ports however. This did used to work a few version back.

TonyJr · Fri Jan 31, 2014 4:39 pm

Working fine on RB493G, RB751U-2HnD and OmniTIK UPA-5HnD with latest included boot firmware.

karina · Fri Jan 31, 2014 5:04 pm

rb433 worked fine

Fri Jan 31, 2014 5:13 pm

GUI support for IPsec users, mode-cfg and policy groups/templates seems to have been added to both WinBox and WebFig. Why is it not mentioned in the ChangeLog?

Sleeth · Fri Jan 31, 2014 5:23 pm

Watchdog it's enabled after upgrading by default, and my RB are going crazy rebooting after 5 minutes.

ndbjorne · Fri Jan 31, 2014 5:58 pm

Hi,
I tested 6.9 updating 3 RB:

RB750 (removed headers but first one)

while ( true ) do={ system resource cpu print ;delay  10}
 # CPU                                           LOAD         IRQ        DISK
 0 cpu0                                           26%          3%          0%
 0 cpu0                                            9%          2%          0%
 0 cpu0                                           36%          5%          0%
 0 cpu0                                            7%          3%          0%
 0 cpu0                                            9%          1%          0%
 0 cpu0                                            6%          2%          0%
 0 cpu0                                           19%          5%          0%
 0 cpu0                                            1%          0%          0%
 0 cpu0                                           36%          6%          1%
 0 cpu0                                           39%          5%         28%
 0 cpu0                                           52%          7%          0%
 0 cpu0                                           28%          4%          0%
 0 cpu0                                           11%          2%          0%
 0 cpu0                                           15%          2%          0%
 0 cpu0                                            8%          1%          0%
 0 cpu0                                           12%          2%          0%
 0 cpu0                                           13%          3%          0%
 0 cpu0                                            5%          1%          0%
 0 cpu0                                           11%          2%          0%
 0 cpu0                                            3%          2%          0%
 0 cpu0                                           11%          2%          0%
 0 cpu0                                            8%          4%          0%
 0 cpu0                                           11%          3%          0%
 0 cpu0                                           21%          4%         15%

on a RB711UA-2HnD

while ( true ) do={ system resource cpu print ;delay  10}
 # CPU                                           LOAD         IRQ        DISK
 0 cpu0                                           45%         43%          0%
 0 cpu0                                           45%         45%          0%
 0 cpu0                                           43%         42%          0%
 0 cpu0                                           42%         42%          0%
 0 cpu0                                           41%         40%          0%
 0 cpu0                                           42%         42%          0%
 0 cpu0                                           41%         41%          0%
 0 cpu0                                           45%         44%          0%
 0 cpu0                                           42%         42%          0%
 0 cpu0                                           40%         40%          0%
 0 cpu0                                           46%         46%          0%
 0 cpu0                                           39%         39%          0%
 0 cpu0                                           39%         39%          0%
 0 cpu0                                           43%         43%          0%
 0 cpu0                                           37%         37%          0%
 0 cpu0                                           38%         38%          0%
 0 cpu0                                           38%         38%          0%
 0 cpu0                                           44%         44%          0%
 0 cpu0                                           41%         41%          0%
 0 cpu0                                           41%         41%          0%
 0 cpu0                                           41%         41%          0%
 0 cpu0                                           40%         40%          0%
 0 cpu0                                           43%         43%          0%
 0 cpu0                                           40%         39%          0%
 0 cpu0                                           44%         43%          0%
 0 cpu0                                           36%         36%          0%
 0 cpu0                                           44%         44%          0%
 0 cpu0                                           45%         45%          0%
 0 cpu0                                           42%         42%          0%
 0 cpu0                                           40%         40%          0%
 0 cpu0                                           43%         43%          0%
 0 cpu0                                           42%         42%          0%
 0 cpu0                                           41%         41%          0%
 0 cpu0                                           38%         38%          0%
 0 cpu0                                           43%         43%          0%

on a 951-2n:

while ( true ) do={ system resource cpu print ;delay  10}
 # CPU                                           LOAD         IRQ        DISK
 0 cpu0                                            1%          0%          0%
 0 cpu0                                            1%          0%          0%
 0 cpu0                                           11%         10%          0%
 0 cpu0                                            1%          0%          0%
 0 cpu0                                            1%          0%          0%
 0 cpu0                                            2%          0%          0%
 0 cpu0                                            1%          0%          0%
 0 cpu0                                            0%          0%          0%
 0 cpu0                                            0%          0%          0%
 0 cpu0                                            0%          0%          0%
 0 cpu0                                            0%          0%          0%
 0 cpu0                                            1%          0%          0%
 0 cpu0                                            0%          0%          0%
 0 cpu0                                            9%          0%          0%
 0 cpu0                                            1%          0%          0%
 0 cpu0                                            9%          9%          0%
 0 cpu0                                            1%          1%          0%
 0 cpu0                                            1%          0%          0%
 0 cpu0                                            1%          1%          0%
 0 cpu0                                            0%          0%          0%
 0 cpu0                                            0%          0%          0%
 0 cpu0                                            0%          0%          0%
 0 cpu0                                            0%          0%          0%
 0 cpu0                                            2%          1%          0%
 0 cpu0                                            1%          0%          0%
 0 cpu0                                            1%          0%          0%
 0 cpu0                                           23%          0%          0%
 0 cpu0                                            2%          0%          0%
 0 cpu0                                            1%          0%          0%
 0 cpu0                                            1%          0%          0%
 0 cpu0                                            0%          0%          0%

which is pretty fair, yet...

> system routerboard print 
...
  current-firmware: [b]3.10[/b]
  upgrade-firmware: [b]3.10[/b]
> system resource print 
                   uptime: 19m52s
                  version: 6.9
...
>

No system is supposed to work hard (all of them had a <5% CPU before upgrading).

Happening to anyone else?

Bye

i4jordan · Fri Jan 31, 2014 6:17 pm

Bug in ipsec phase 2 AES-256 on CCR??

I have updated our CCR1036-12G-4EM from 6.7 to 6.9.
Al was working fine except for a lot of our VPN tunnels.

The standard ipsec tunnels with AES-256 in phase 2 (proposal) all send packages but did not receive.
After changing from AES-256 to AES-128 they went fine. All SA where correct but only outbound data not inbound.

So I do think there is a bug in the AES-256 implementation.

The other side of the tunnels were no Mikrotik devices but for example a SonicWall SRA 3500.

After changing from AES-256 to AES-128 in Phase 2 things were working OK. Also the tunnels seem much (factor 2x).

Campano · Fri Jan 31, 2014 7:05 pm

Hi, SMB not working well ...

can't access to shared folder, windows show me "not permision for this folder..."

efaden · Fri Jan 31, 2014 7:53 pm

I also still have high "management" CPU usage on the profiler (5 to 8%) while idle. Doesn't happen on my 2011s.

estdata · Fri Jan 31, 2014 8:04 pm

There is nothing to stop one thing rather than another does not run from one of the version, I offered here that do not could be Denial of Service thing to develop in.
However, the increase in the cpu, why this version of the RB2011UAS-the 2Hnd-IN for more than a RB951-AN on this.
Nonsense
I do not want to cash out ahead of a new router for a new version if the blow goes just after the

mars · Fri Jan 31, 2014 8:10 pm

sxt not accesible after upgrade

Lakis · Fri Jan 31, 2014 9:06 pm

sxt not accesible after upgrade

Hm thats interesting you know cos after upgrade (from 6.5) on two RB433AH address list was empty - can u check if this is a case

noyo · Fri Jan 31, 2014 9:13 pm

Lite5

Upgrade 6.7 -> 6.9

18:05:50 system,error,critical System rebooted because of kernel failure

the device works

and

jan/02/1970 01:00:20 wireless,info xx:xx:xx:xx:xx:CE@wlan1 established connection on 5500000, SSID moja_siec

Yekver · Fri Jan 31, 2014 9:55 pm

After upgrade SXT 5HnD from 6.7 to 6.9 graph lines (Rx signal/Tx signal )is not rendering at index page.
I have 2 SXT's, configured in briged mode, but none of them rendering graph lines!
In first SXT with wireless bridge mode, there is no 'wireless' section at all.

Check this please.

JanezFord · Fri Jan 31, 2014 10:26 pm

Upgraded the following routerboards to 6.9 + firmware (a few of each): rb751u-2hnd, rb751g-2hnd, rb711u-2hnd, rb951g-2hnd, rb800, rb493g, rb2011uas-rm, rb450g

no issues so far ... upgrade went smooth ...

JF

cheeze · Fri Jan 31, 2014 11:26 pm

From the looks of things, would yalls say that for what it's worth....6.9 is a pretty good release with minimal bugs?

I downloaded it but I haven't deployed it yet. From the posts here, it seems that this release is a good release.

YakCo · Fri Jan 31, 2014 11:42 pm

Nice!

6.9 installed smoothly on the Dell 1950, with LSI / Perc6i (2 x 15K drives) and onboard Broadcom NIC's. I have not been able to thoroughly test many functions, but so far it looks very good. I put a 1GB file on it, and then FTP downloaded at 650 Mb/s.

The only odd behaviour I've noticed so far, is that it will not reboot on command, via winbox.

Dean
(Yakco)

ofca · Fri Jan 31, 2014 11:47 pm

*) bridge - default protocol-mode changed to RSTP for new bridges,

Really? Seriously? :/

karina · Fri Jan 31, 2014 11:49 pm

Ok so which of us Ccr owners is going to becfirst to have a go

ditonet · Fri Jan 31, 2014 11:55 pm

Just upgraded CCR1036-12G-4S, everything looks good.

Regards,

karina · Fri Jan 31, 2014 11:59 pm

At last it looks like I can dust off my Ccr sat on the shelf. Just checked the Ccr thread and not a murmur. No news is good news

ndbjorne · Sat Feb 01, 2014 12:15 am

Hi,
upgraded another RB951-2n and like the previous one show a 3.10 and cpu seems ok.

Back to 711UA-2HnD: tried downgrade to 6.7 but cpu is 43% like with 6.9, but notice that

> while ( true ) do={ system resource cpu print ;delay  10}
 # CPU                           LOAD         IRQ        DISK
 0 cpu0                           47%         45%          0%
 0 cpu0                           47%         47%          0%
 0 cpu0                           39%         39%          0%
interrupted
> system resource irq print                
Flags: ro - read-only 
 #    IRQ USERS                  CPU ACTIVE-CPU         COUNT
 0      2 wlan1                 auto          0    32 316 309
 1      3 usb1                  auto          0             0
 2      5 switch0               auto          0        30 271
 3     16 beeper                auto          0         1 604
 4     18 GPIO                  auto          0       247 704
> interface wireless disa  wlan1           
> while ( true ) do={ system resource cpu print ;delay  10}
 # CPU                           LOAD         IRQ        DISK
 0 cpu0                            1%          0%          0%
 0 cpu0                            0%          0%          0%
 0 cpu0                            0%          0%          0%
interrupted
> interface wireless ena  wlan1            
> while ( true ) do={ system resource cpu print ;delay  10}
 # CPU                           LOAD         IRQ        DISK
 0 cpu0                           40%         37%          0%
 0 cpu0                           47%         47%          0%
interrupted

then I disabled wlan1 and rebooted, when back

system resource irq print                
Flags: ro - read-only 
 #    IRQ USERS                  CPU ACTIVE-CPU         COUNT
 0      2 wlan1                 auto          0             0
...
> system clock print 
            time: 22:51:35
...
> interface wireless ena  wlan1            
> system resource irq print     
Flags: ro - read-only 
 #    IRQ USERS                  CPU ACTIVE-CPU         COUNT
 0      2 wlan1                 auto          0       384 766
...
> system clock print            
            time: 22:51:56
...
> system resource irq print 
Flags: ro - read-only 
 #    IRQ USERS                  CPU ACTIVE-CPU         COUNT
...
> system clock print        
            time: 22:52:06
...
> system resource irq print 
Flags: ro - read-only 
 #    IRQ USERS                  CPU ACTIVE-CPU         COUNT
 0      2 wlan1                 auto          0     1 135 661
...
> system clock print        
            time: 22:52:14
...

Then tried downgrade to 5.26 but it behaved the same.
Then back to 6.7 but used main package instead of all those previously installed... nothing again.
Downgraded bootloader to 3.10 and disabled unused packages...

  
> while ( true ) do={ system resource cpu print ;delay  10}
 # CPU                           LOAD         IRQ        DISK
 0 cpu0                           58%         51%          0%
 0 cpu0                           41%         41%          0%
 0 cpu0                           42%         42%          0%
interrupted
> int wire disable wlan1                   
> while ( true ) do={ system resource cpu print ;delay  10}
 # CPU                           LOAD         IRQ        DISK
 0 cpu0                           26%         17%          0%
 0 cpu0                            0%          0%          0%
 0 cpu0                            0%          0%          0%
interrupted
> int wire ena wlan1                       
> while ( true ) do={ system resource cpu print ;delay  10}
 # CPU                           LOAD         IRQ        DISK
 0 cpu0                           18%         11%          0%
 0 cpu0                           40%         40%          0%
interrupted

Only thing I can say is that cpu usage is related to 711 wireless.

Bye

tournet · Sat Feb 01, 2014 2:55 am

Hi, SMB not working well ...

can't access to shared folder, windows show me "not permision for this folder..."

Samba works fine for me. Routeros 6.9 and windown 7.

gnuttisch · Sat Feb 01, 2014 3:01 am

Have updated rb1000, rb1100, rb1100ah2, rb450g, rb450,

so far so good.

This is ore first step towards ros6 in critical equipment, it feels like 6.9 is finally ready.

WirelessRudy · Sat Feb 01, 2014 4:54 am

What's new in 6.9 (2014-Jan-31 11:18):

*) wireless - improved 11n and nv2 stability;

So, what were these changes?
Did many have problems with previous versions?

I mean, wireless is very dynamic. Links are at times not as stable as other times and basically can always be improved. So its hard to tell if a link is 100% or just 90%
But if what I thought was 100% now could be improved to what is now 120% (so a new 100%) that would be nice.

Anybody noticed any improvements?

I just finished upgrading my complete network to 6.7 and was happy. Hardly any issues and a pretty stable network.
I had only one forgotten unit to go and suddenly tonight found it upgraded to 6.9!

I'm working as the only MT user in a valley full of ubnt operators.
Basically all 5Ghz frequencies are used, either by me or by any other operator. We all use tdma (Airmax versus NV2) and mimo 'n' wireless nowadays.
After the first tdma (=nv2) issues in the early days the wireless is pretty stable and robust in v6.x family. But heho, with so much interference sources some of my links can always do better. Should I now really hurry to do a new full upgrade round so all my wireless become even more robust?

mahnet · Sat Feb 01, 2014 4:59 am

/damn

RADIUS CRASHED AFTER UPGRADE TO 6.9.
I was having some issues with radius accounting, they say they have solved it in 6.9
Damn what a way to ruin my day.

Restarting radius server says DATABASE CORRUPTED.
Does not even accept my customer name and password.

manuelsbc · Sat Feb 01, 2014 6:52 am

I update my groove from 6.7 to 6.8, and does not start .. that I can do? turns but does not finish starting, I can not access, not reflected in winbox

Sat Feb 01, 2014 8:08 am

I update my groove from 6.7 to 6.8, and does not start .. that I can do? turns but does not finish starting, I can not access, not reflected in winbox

This is the 6.9 Thread. Netinstall 6.9.

Sat Feb 01, 2014 8:16 am

What's new in 6.9 (2014-Jan-31 11:18):

*) wireless - improved 11n and nv2 stability;
So, what were these changes?
Did many have problems with previous versions?

I mean, wireless is very dynamic. Links are at times not as stable as other times and basically can always be improved. So its hard to tell if a link is 100% or just 90%
But if what I thought was 100% now could be improved to what is now 120% (so a new 100%) that would be nice.

Anybody noticed any improvements?

I just finished upgrading my complete network to 6.7 and was happy. Hardly any issues and a pretty stable network.
I had only one forgotten unit to go and suddenly tonight found it upgraded to 6.9!

I'm working as the only MT user in a valley full of ubnt operators.
Basically all 5Ghz frequencies are used, either by me or by any other operator. We all use tdma (Airmax versus NV2) and mimo 'n' wireless nowadays.
After the first tdma (=nv2) issues in the early days the wireless is pretty stable and robust in v6.x family. But heho, with so much interference sources some of my links can always do better. Should I now really hurry to do a new full upgrade round so all my wireless become even more robust?

There was a bug where modulations stay at the lowest rate under unknown conditions. May be this is solved now.
There was a report that with 6.8 nv2 latency has improved and is more constant.

I also would like to hear more on what the improvements are.

Sat Feb 01, 2014 9:38 am

Done some testing on SXT, Omnitik, 951, 411AH. So far no problem.
Did not notice any changes in wireless behavior/performance.

estdata · Sat Feb 01, 2014 11:13 am

This is ROS 6.9

Shiro · Sat Feb 01, 2014 11:32 am

Updated two 493g and two 411ah from 6.7 to 6.9, no issues so far.

PPTP, IPSec, L2TP tunnels working, No change on The Wireless side, everything is fine so far.

I wonder why some of you talk about Firmware 3.11. In Routerboard/Upgrade there is only 3.10 and upgrade 3.10.
When pressing upgrade and reboot the device, I see some error message about ram disk invalid. But the Boards starts normally, but it's still on 3.10.

steen · Sat Feb 01, 2014 11:48 am

Hello Folks!

Done some upgrades CRS, RB750, RB433, RB2011UAS-2hnd no issues so far.

Observations, putential remaining bugs:

CRS, it still leaks traffic between vlans (followed mikrotik manual and instructions howto do it), same issue as before.
CRS, in winbox, still all swicthed interfaces show same traffic speed, for example if one session on ether10 is 60Mbit/s all else swicthports also show 60Mbit/s
CRS, is still practically impossible to understand swicthing and vlans setup if you are not doctor i silicon and layer1, layer2, ethernet protocols.

Other observations:
Firmware update on RB750 from 3.10 to 3.12, no firmware update on any other device was observed.

Zavi · Sat Feb 01, 2014 12:09 pm

Hi, SMB not working well ...
can't access to shared folder, windows show me "not permision for this folder..."

I had the same problem after upgrade from 5.26 -> 6.7 on the 751G
It was enough to unmount network drive and mount it again, now it works.
And finally, now it works even from ES explorer on Android. Although no SMB in changelog.

payday · Sat Feb 01, 2014 1:40 pm

Hi, SMB not working well ...
can't access to shared folder, windows show me "not permision for this folder..."

I have the same problem (Win 7 and RB2011). I removed and recreated all shares and it is slightly better now. My default share pub was targeted to directory "/pub" that did not exist, so I created it.
Still sometimes copying to SMB share hangs when I try to create/rename other folders in meantime, sometimes message appears that network folder is inaccessible.

Also, errors appear when I select more then one folder and right click it (I have 2 shares usb and pub):

Feb/01/2014 11:55:39 disk smb, info user Kamillo connected to share usb
Feb/01/2014 11:55:39 disk smb, info usb active sessions: 1
Feb/01/2014 11:55:41 disk smb, info user Kamillo connected to share pub
Feb/01/2014 11:55:41 disk smb, info pub active sessions: 1
Feb/01/2014 11:55:46 disk smb, error cannot connect to invalid share pu
Feb/01/2014 11:55:46 disk smb, error cannot connect to invalid share pu
Feb/01/2014 11:58:40 disk smb, error cannot connect to invalid share us
Feb/01/2014 11:58:40 disk smb, error cannot connect to invalid share us

Generally, SMB is 2x faster then it was in 6.7 but only when writing to share

. It is interesting - writing is faster then reading.

ropeba · Sat Feb 01, 2014 3:02 pm

6.9 still has bug same as the 6.7 version. Ip addresses are disappeared, dhcp stops working, pppoe is not working, more then a half of services is not usable and reboot is necessary. It happens on all RBs!!!! Now it's happened on CCR which is core router in my network. It's wrong when 10.000 users is left without service only because Mikrotik is not fixing reported bugs!

brainy · Sat Feb 01, 2014 3:08 pm

What's new in 6.9 (2014-Jan-31 11:18):

*) lcd - added option to change the color-scheme;
*) updated bootloader firmware;
*) ppp: fixed RADIUS accounting;
*) ppp: fixed IPV6-Prefix assigning;
*) ppp: fixed dial-on-demand;

What's new in 6.8 (2014-Jan-29 15:52):

*) bridge - default protocol-mode changed to RSTP for new bridges,
fixed bridge mac address changing when port (with lowest mac address) goes down
*) userman - improve startup time;
*) sstp client - support server name verification from certificate;
*) wireless - improved 11n and nv2 stability;
*) dhcp client - support interfaces in bridge;
*) dhcp - parse decimal strings and IP addreses in options value;
*) bgp - don't show community 'internet' in BGP advertisements;
*) ipsec - enable hardware acceleration for aes-cbc + md5|sha1|sha256 aead on tilera cpus;
*) ospf - fixed checksum calculation for OSPFv3 AS-external-LSAs;
*) default configuration - changed dhcp server lease time to 10 minutes;
*) fixed port isolation on CRSs (bug introduced in v6.6);
*) smb - added support for SMB 2.002
*) timezone information updated;
*) ppp - fixed ppp bridging (did not work since v6.6);
*) improved speed of PPP, PPPoE, PPTP & L2TP on multicore routers;
*) address-list - fix crash when adding two identical address list entries (not allowed any more);
*) fixed multicast forwarding on CCRs;
*) firewall - improved address-type matcher, and added it for ipv6 aswell;
Simply push "Check for Updates" in your router,
or download from http://www.mikrotik.com/download

Is DHCP-Server on vlan-interface on bonding-interface fixed?

artsmolkin · Sat Feb 01, 2014 3:41 pm

Looks like there is a problem with compatibility of old and new ppp packages...

I'm trying to connect via winbox to mikrotik with 6.9, which is behind mikrotik with 6.3. So, winbox crashes with message "could not connect" from 30-40 second since I connected. I can't do anything on it.
Scheme:
me -> pptp -> mikrotik ros 6.3 -> pptp -> mikrotik 6.7

What should I do with this?

zyzelis · Sat Feb 01, 2014 3:46 pm

What's new in 6.9 (2014-Jan-31 11:18):

*) lcd - added option to change the color-scheme;
*) updated bootloader firmware;
*) ppp: fixed RADIUS accounting;
*) ppp: fixed IPV6-Prefix assigning;
*) ppp: fixed dial-on-demand;

What's new in 6.8 (2014-Jan-29 15:52):

*) bridge - default protocol-mode changed to RSTP for new bridges,
fixed bridge mac address changing when port (with lowest mac address) goes down
*) userman - improve startup time;
*) sstp client - support server name verification from certificate;
*) wireless - improved 11n and nv2 stability;
*) dhcp client - support interfaces in bridge;
*) dhcp - parse decimal strings and IP addreses in options value;
*) bgp - don't show community 'internet' in BGP advertisements;
*) ipsec - enable hardware acceleration for aes-cbc + md5|sha1|sha256 aead on tilera cpus;
*) ospf - fixed checksum calculation for OSPFv3 AS-external-LSAs;
*) default configuration - changed dhcp server lease time to 10 minutes;
*) fixed port isolation on CRSs (bug introduced in v6.6);
*) smb - added support for SMB 2.002
*) timezone information updated;
*) ppp - fixed ppp bridging (did not work since v6.6);
*) improved speed of PPP, PPPoE, PPTP & L2TP on multicore routers;
*) address-list - fix crash when adding two identical address list entries (not allowed any more);
*) fixed multicast forwarding on CCRs;
*) firewall - improved address-type matcher, and added it for ipv6 aswell;
Simply push "Check for Updates" in your router,
or download from http://www.mikrotik.com/download
Is DHCP-Server on vlan-interface on bonding-interface fixed?

Hello,
DHCP server works for me on VLAN interfaces with 6.9, so in your case DHCP server not works on bonding?
What network layout with bonding, il try to reproduce it.

patrickmkt · Sat Feb 01, 2014 4:11 pm

updated from 6.7 to 6.9 on a rb1100ah and now I have on one of my sstp link to another routerboard still on 6.7:
Encryption negociation rejected.

dlj87 · Sat Feb 01, 2014 6:02 pm

OpenVPN server doesn't work! Clients log off immediately after connection establishing. Downgrade to 6.7.

npero · Sat Feb 01, 2014 6:20 pm

OpenVPN server doesn't work! Clients log off immediately after connection establishing. Downgrade to 6.7.

I try to connect to OpenVPN on x86 after upgrade to 6.9 no problem. Using OpenVPN client in windows version 2.3.1.
Also this x86 machine is PPPOE server with radius authentication and around 80 active pppoe connection and couple PPTP upgraded to 6.9 for now up time 5 hours no problems.
Seem to me PPPOE service is working better first impression.

Also upgraded two Rb750 and one 951G no problem for now up time 20 hours, 951 connected using OpenVPN to 6.7 no problem for now.
For now no problem I can see, next weekend is planing upgrade complete network to 6.9.

dlj87 · Sat Feb 01, 2014 6:34 pm

I tried another RB as a client... All RB from our corporate network cann't connect to RB with OS 6.9 installed

mahnet · Sat Feb 01, 2014 7:09 pm

6.9 still has bug same as the 6.7 version. Ip addresses are disappeared, dhcp stops working, pppoe is not working, more then a half of services is not usable and reboot is necessary. It happens on all RBs!!!! Now it's happened on CCR which is core router in my network. It's wrong when 10.000 users is left without service only because Mikrotik is not fixing reported bugs!

r u using CCR as RADIUS SERVER? 10,000 users on Mikrotik is amazing. I used 1100ah which barely goes to 200+ users.

npero · Sat Feb 01, 2014 7:23 pm

I tried another RB as a client... All RB from our corporate network cann't connect to RB with OS 6.9 installed

I try to connect for one Rb750 to x86 OpenVPN both 6.9.
Yes disconnect your are right.

Solution select on OpenVPN client chipper and authenticate different from any then working.

ropeba · Sat Feb 01, 2014 8:03 pm

6.9 still has bug same as the 6.7 version. Ip addresses are disappeared, dhcp stops working, pppoe is not working, more then a half of services is not usable and reboot is necessary. It happens on all RBs!!!! Now it's happened on CCR which is core router in my network. It's wrong when 10.000 users is left without service only because Mikrotik is not fixing reported bugs!
r u using CCR as RADIUS SERVER? 10,000 users on Mikrotik is amazing. I used 1100ah which barely goes to 200+ users.

No. routing, ospf, dhcp, etc...

mahnet · Sat Feb 01, 2014 8:19 pm

6.9 still has bug same as the 6.7 version. Ip addresses are disappeared, dhcp stops working, pppoe is not working, more then a half of services is not usable and reboot is necessary. It happens on all RBs!!!! Now it's happened on CCR which is core router in my network. It's wrong when 10.000 users is left without service only because Mikrotik is not fixing reported bugs!
r u using CCR as RADIUS SERVER? 10,000 users on Mikrotik is amazing. I used 1100ah which barely goes to 200+ users.
No. routing, ospf, dhcp, etc...

And what Radius manager r u having??

steen · Sat Feb 01, 2014 8:44 pm

Hello Folks!

Whole our environment is successfully upgraded to ros6.9, rb411, rb333, rb600, rb433,rb750, sxt, sextant, ccr, crs, rb2011. Many devices in two towns.
All working normally, no problems at all.

However we still do not have a working CRS with vlan:s and swicthes, it still leaks traffic.
CCR have still disabled snmp and lcd due to sudden death of ports otherwise, it works in that way since ros6.4.

Regards //
// Peter Steen

zyzelis · Sat Feb 01, 2014 11:30 pm

Hello Folks!

Whole our environment is successfully upgraded to ros6.9, rb411, rb333, rb600, rb433,rb750, sxt, sextant, ccr, crs, rb2011. Many devices in two towns.
All working normally, no problems at all.

However we still do not have a working CRS with vlan:s and swicthes, it still leaks traffic.
CCR have still disabled snmp and lcd due to sudden death of ports otherwise, it works in that way since ros6.4.

Regards //
// Peter Steen

Good to hear you,
so lets start to upgrade

mahnet · Sat Feb 01, 2014 11:39 pm

Any successful upgrades of the x86. I guess ROS 6.9 has issues understanding the hard disk.

madcat · Sat Feb 01, 2014 11:42 pm

address-list - fix crash when adding two identical address list entries (not allowed any more);

it also solved the crash where you rename an addresslist item!
thanks!

onnoossendrijver · Sat Feb 01, 2014 11:49 pm

Any successful upgrades of the x86. I guess ROS 6.9 has issues understanding the hard disk.

I have no issues with my x86 booting 6.9 from USB stick.

bajituka · Sun Feb 02, 2014 12:37 am

Hi men!
Sorry in advance, I'm newbie to MT and this forum.
After upgrading to 6.9, ROS doesn't save the country setting and doesn't display the list if devices connected in the quickset tab.
http://puu.sh/6Gypy.png

nosovk · Sun Feb 02, 2014 1:17 am

Please, fix vlans. I still use 6.3 becouse of it

bds1904 · Sun Feb 02, 2014 1:38 am

RB2011UAS-RM 6.7 to 6.9 breaks PCC routing with 2 WAN connections

the second interface is not accessible when the first is enabled

downgrade to 6.7 and it works just fine

dkmj · Sun Feb 02, 2014 2:21 am

Upgrade 6.7 -> 6.9

Running Wireless bridge mode between two RB SXT 5HnD. The client side looks 'fine' (as it did prior to upgrade), link is up, but on the Server side the Quick set screen lacks large portions/parameters/info: IP address netmask dns gateway etc info "on the right" is correct. But top of screen says it's in client/CPE mode which can't be the case since client side is up and 'receiving'? And below, all the wireless info is simply missing - white space on the Quick Set page.

I checked three different browsers and it's the same in all of them.

Do not want to start 'reconfiguring' for risk of loosing link.

villageworker · Sun Feb 02, 2014 2:48 am

Upgrade 6.7 -> 6.9

Running Wireless bridge mode between two RB SXT 5HnD. The client side looks 'fine' (as it did prior to upgrade), link is up, but on the Server side the Quick set screen lacks large portions/parameters/info: IP address netmask dns gateway etc info "on the right" is correct. But top of screen says it's in client/CPE mode which can't be the case since client side is up and 'receiving'? And below, all the wireless info is simply missing - white space on the Quick Set page.

Same issue on a 9512n. Attached are screens taken of the same 9512n in winbox and quickset.

TealFrog · Sun Feb 02, 2014 9:23 am

Attempted to upgrade three routers from V6.7 to V6.9
RB493 - Okay
RB493G - Okay
2011UAS-2HnD - Unsuccessful

Attempted twice to upgrade 2011UAS-2HnD router to V6.9, router became unusable running constantly at 100% CPU utilization. I would have liked to have had some diagnostics to post, but router wasn't responding. Both attempts ended with downgrading to V6.7. CPU utilization back to normal (nil) with downgrades.

Regards,

skibi82 · Sun Feb 02, 2014 10:14 am

Software 6.9 for The quick repair
After updating RB951G after some time to get the fuck out
The same thing happens with RB450G
Hardware reboot is need to start system and downgrade the device

At the moment,
- 12 x RB751G
- 6 x RB2011-L
- 20 x RB2011-UAS
- 10 x RB1100AHx2
firmware on them do not get the fuck communication

npero · Sun Feb 02, 2014 10:53 am

Any successful upgrades of the x86. I guess ROS 6.9 has issues understanding the hard disk.

I upgrade x86 on them PPPOE server around 80 active PPPOE connection, radius authentication, NAT, OSPF, MPLS, VPLS not very complicated setup so far no problem.
Up time: 21 hours for now

Software 6.9 for The quick repair
After updating RB951G after some time to get the f**k out
The same thing happens with RB450G

What is the problem ?

I typing this text connected to one 951G with 6.9, on 951G have OpenVPN clients, web proxy, virtual AP, simple queue I don't see any problem.
Up time for not: 1 day 14 hours

steen · Sun Feb 02, 2014 11:03 am

Hello Folks!

After the upgrade party yesterday which was fully successful since no device went down and needed a hard restart or reset.

But portflapping problems remains on CRS and potentially also CCR but much less, this time both between the devices and and to other devices so that problem is not solved. Portflapping has been first observed at our site after we replaced HP and Cisco swithches with RB.
Cables, patch panels, rj45 sockets are all cat 6 or better, all cabling used here is in permanent installation and they are all Cat 6 or higher, not exeeding 30 meters in any case.

Portflapping has never been observed on any RB333, RB600, RB750, RB411, RB433.

Kind of big dissapointment that CRS still leaks and does not work like a normal switch with trunks and vlans, not even stuff like etherchannels and bonding is there, if you do not use the weak cpu driven variants. It works barely as a classic hub, if ports were not flapping.

Otherwise problem with wireless or ipsec and routing on any RB device, but it has never been observed in our sites.

saaremaa · Sun Feb 02, 2014 11:15 am

Any successful upgrades of the x86. I guess ROS 6.9 has issues understanding the hard disk.
I upgrade x86 on them PPPOE server around 80 active PPPOE connection, radius authentication, NAT, OSPF, MPLS, VPLS not very complicated setup so far no problem.
Up time: 21 hours for now

I confirm. I upgrade my x86 Mikrotik from 6.2 → 6.9 (OSPF(ipv4)+OSPFV3(ipv6)+BGP+VLAN) and have no problem for me. Mikrotik installed on ssd drive Intel.

[xxxxxxxxxxx@Mikrotik-Border] /system resource> pr
                   uptime: 2d7h4m32s
                  version: 6.9
               build-time: Jan/30/2014 15:06:11
              free-memory: 1740.7MiB
             total-memory: 1893.9MiB
                      cpu: Intel(R)
                cpu-count: 8
            cpu-frequency: 3392MHz
                 cpu-load: 2%
           free-hdd-space: 54.8GiB
          total-hdd-space: 55.0GiB
  write-sect-since-reboot: 460108
         write-sect-total: 460108
        architecture-name: x86
               board-name: x86
                 platform: MikroTik

Yekver · Sun Feb 02, 2014 12:53 pm

Upgrade 6.7 -> 6.9

Running Wireless bridge mode between two RB SXT 5HnD. The client side looks 'fine' (as it did prior to upgrade), link is up, but on the Server side the Quick set screen lacks large portions/parameters/info: IP address netmask dns gateway etc info "on the right" is correct. But top of screen says it's in client/CPE mode which can't be the case since client side is up and 'receiving'? And below, all the wireless info is simply missing - white space on the Quick Set page.

I checked three different browsers and it's the same in all of them.

Do not want to start 'reconfiguring' for risk of loosing link.

+1 I have the same issue with having the same network topology

estdata · Sun Feb 02, 2014 12:56 pm

cpu not 100% and this is ROS 6.9
Routerboard: RB2011UAS-2Hnd-IN

skibi82 · Sun Feb 02, 2014 2:23 pm

Any successful upgrades of the x86. I guess ROS 6.9 has issues understanding the hard disk.
I upgrade x86 on them PPPOE server around 80 active PPPOE connection, radius authentication, NAT, OSPF, MPLS, VPLS not very complicated setup so far no problem.
Up time: 21 hours for now

Software 6.9 for The quick repair
After updating RB951G after some time to get the f**k out
The same thing happens with RB450G
What is the problem ?

I typing this text connected to one 951G with 6.9, on 951G have OpenVPN clients, web proxy, virtual AP, simple queue I don't see any problem.
Up time for not: 1 day 14 hours

The problem exists

To my eye there is something broken in IPSEC

Because the only routers that have ceased to operate (40 pieces)
They were
/ ip ipsec installed-sa flush, added in the scheduler as well as in scripts.

Another explanation I am not able to find any, behavior. But thanks to this I have learned once again that you only need to do the upgrade on 1 test router with all manner of configuration would then not run on Sundays.

Another is that they started the last month I've learned that it's no trust RouterOS device because it seems that the development of RouterOS returned to the stage before beta

At 6.7 the error does not occur

I hope that this will change and the developers finally understand forehead is meant a stable version meaning Beta

kichsolo · Sun Feb 02, 2014 3:20 pm

Hi

I tried another RB as a client... All RB from our corporate network cann't connect to RB with OS 6.9 installed
I try to connect for one Rb750 to x86 OpenVPN both 6.9.
Yes disconnect your are right.

Solution select on OpenVPN client chipper and authenticate different from any then working.

Not woking for me, cause never set to "any". Has multiple different Microtiks, some working, some lost OVPN server ability after upgrade. As OVPN client - no such problem.
Downgrading to 6.7

doridian · Sun Feb 02, 2014 3:55 pm

Somehow OpenVPN has become unstable with 6.9. It was super stable in 6.7 now I get disconnects every few hours...

YakCo · Sun Feb 02, 2014 5:23 pm

Updating my earlier post...

Minor SMB problem. SMB share works. Win7 client. Can download files up to 1GB - download rate is 68 KB/s. When I try to upload anything larger than about 120-140 MB, the upload fails. Did not test for file sizes larger than 1GB.

Dean

Nice!

6.9 installed smoothly on the Dell 1950, with LSI / Perc6i (2 x 15K drives) and onboard Broadcom NIC's. I have not been able to thoroughly test many functions, but so far it looks very good. I put a 1GB file on it, and then FTP downloaded at 650 Mb/s.

The only odd behaviour I've noticed so far, is that it will not reboot on command, via winbox.

Dean
(Yakco)

littlebill · Sun Feb 02, 2014 6:26 pm

2011 upgraded flawlessly. had a slightly high cpu for about 10 minutes, then settled down.

fixed my 10 mbit led problem on the fast ethernet side, everything else is working just fine

spacemind · Sun Feb 02, 2014 7:47 pm

Somehow OpenVPN has become unstable with 6.9. It was super stable in 6.7 now I get disconnects every few hours...

Openvpn very unstable and with a lot of disconnects and VLANs are heaving a lots of problems in 6.9.

How can this happen in routers? So many problems. ROS 6.9 should be renamed beta before bugs are fixed.

MT guys please fix the bugs, 6.9 is unusable as hotspot with openvpn and radius with vlans. Back to 6.3

Neohunter · Sun Feb 02, 2014 8:26 pm

I update RB750G from 6.7 to 6.9 via system/package/check for updates - download & install
router reboot but not start - also all Ethernet interface led not flashed

i try repair it via netinstall - connect 1 eth on 750G to pc eth card
push rest and start router - wait so long but act not flashed !
and i can't see router in netinstall

How i can reanimate router ?

Sun Feb 02, 2014 8:41 pm

I updated RB2011UAS-2HND today from v. 6.5 to v. 6.9.

Almost no problems, just Dude was not accepting the login, so I needed to uninstall Dude and install it again (v. 3.6). After that it works with no problems so far.

Just there is something eating the memory - I suspect the Dude. It is constantly rising by the approx 2% per hour.
Will be observing it and maybe try to find something about Dude and memory...

All other things are working fine as before upgrade.

zyzelis · Sun Feb 02, 2014 9:24 pm

I update RB750G from 6.7 to 6.9 via system/package/check for updates - download & install
router reboot but not start - also all Ethernet interface led not flashed

i try repair it via netinstall - connect 1 eth on 750G to pc eth card
push rest and start router - wait so long but act not flashed !
and i can't see router in netinstall

How i can reanimate router ?

Hello,
1. try to press reset button for 3 secs and search 750G in winbox. if it wont help you then
2. press reset button for 30s and release. Device should appear in netinstall list.

ps. netinstall works only on poe-in interface (ether1?))
pps. are you sure, that netinstall prepared correctly? ie disabled firewall, set ip address

Good luck

docmarius · Sun Feb 02, 2014 9:31 pm

For me. with this FW on a RB-450 routing marks stopped working.
- mark route in mangle/prerouting
- set another default route for that routing mark (this one shows blue in the list, meaning it has no effect)
- marked packets still go out on the default gw.

The same config worked flawless in 6.7

Can I have a download link for 6.7 so I can downgrade?

dlj87 · Sun Feb 02, 2014 10:07 pm

Can I have a download link for 6.7 so I can downgrade?

Just copy the 6.9 link, change 6.9 to 6.7 and press enter)

docmarius · Mon Feb 03, 2014 1:23 am

Tnx. Did that and everything is back in working order with the same config and 6.7 FW

littlebill · Mon Feb 03, 2014 2:28 am

pptp appears broke, trying to manage winbox while in the tunnel, different sections are blank, then winbox shows disconnected

sstp also does not appear to be working at this point either. it does not, get an error at verifying username and password, then windows times out with error 628......

i downgraded back to 6.7, pptp and sstp back to working flawless and significantly faster

guess i can live for a while without the leds working right.

guys fix ppp, you seem to have broken alot

this is on a 2011

heislerb · Mon Feb 03, 2014 2:39 am

Can I have a download link for 6.7 so I can downgrade?
Just copy the 6.9 link, change 6.9 to 6.7 and press enter)

I copied 6.7 to the file area and rebooted the router. The router still shows Version 6.9. How do you down grade???

littlebill · Mon Feb 03, 2014 2:40 am

Can I have a download link for 6.7 so I can downgrade?
Just copy the 6.9 link, change 6.9 to 6.7 and press enter)
I copied 6.7 to the file area and rebooted the router. The router still shows Version 6.9. How do you down grade???

system packages, downgrade

heislerb · Mon Feb 03, 2014 2:52 am

system packages, downgrade

What is a system package? How do I get it? What do I do with it?
Sorry I know so little!

littlebill · Mon Feb 03, 2014 3:08 am

system packages, downgrade
What is a system package? How do I get it? What do I do with it?
Sorry I know so little!

winbox....

heislerb · Mon Feb 03, 2014 3:42 am

winbox....

Thanks littlebill !!!! It all worked like a charm.

You are a person of few words. I almost gave up with your response. Then I followed your instructions and I had a little laugh!!!

Bob

3bs · Mon Feb 03, 2014 3:49 am

fixed 5 bugs and added 500, super release.

jondavy · Mon Feb 03, 2014 6:02 am

why this script does not work?

/ ip hotspot user profile remove [ find default=no ]

and in pppoe profiles works perfectly?

/ ppp profile remove [find default=no ]

tested with RouterOS 6.7 and 6.9

rtfmjoey · Mon Feb 03, 2014 7:45 am

Just upgraded my RB2011

Upgrade solved some WiFi stability problems for me!

But.....

-LCD is STILL missing from webfig,
-PoE on port 10 is missing
-The quick set doesn't show anything anymore

WinBox shows everything quite nicely...

Mon Feb 03, 2014 9:14 am

Just upgraded my RB2011

Upgrade solved some WiFi stability problems for me!

But.....

-LCD is STILL missing from webfig,
-PoE on port 10 is missing
-The quick set doesn't show anything anymore

WinBox shows everything quite nicely...

make sure you upgrade the firmware also (/system RouterBOARD upgrade) and send your supout.rif file to support.

markom · Mon Feb 03, 2014 10:33 am

I have RB2011 working on 6.0
The router is ~100km away from my office.
I really want to upgrade it but I don't have intensions to travel on site to do downgrade.
Is it safe to upgrade to 6.9 RB2011?

Alfabi · Mon Feb 03, 2014 11:13 am

im have strange bug from 6.5-6.9 on 6.4 all is well
onetime double traffic each 40-50 sec. on interfaces using ospf+mpls, no difference track system on or off mtu is 1518 or 1508, double traffic is only on transport ports.... on whole types on MT boards without difference where to use....

Mon Feb 03, 2014 11:28 am

I have RB2011 working on 6.0
The router is ~100km away from my office.
I really want to upgrade it but I don't have intensions to travel on site to do downgrade.
Is it safe to upgrade to 6.9 RB2011?

Are you kidding? Wait some days until user reports say it's save and then do it.

I first update routers in my room, wait some days, then Routers easily reachable
and then remote sites.

BGP Routers still running on 5.25.

ditonet · Mon Feb 03, 2014 12:28 pm

Are you kidding? Wait some days until user reports say it's save and then do it.

It looks that only I was brave enough, to upgrade router located 210 km away from me

.
But everything went fine, so far so good.

Regards,

rtfmjoey · Mon Feb 03, 2014 12:46 pm

Just upgraded my RB2011

Upgrade solved some WiFi stability problems for me!

But.....

-LCD is STILL missing from webfig,
-PoE on port 10 is missing
-The quick set doesn't show anything anymore

WinBox shows everything quite nicely...
make sure you upgrade the firmware also (/system RouterBOARD upgrade) and send your supout.rif file to support.

Suddenly all of it is working as it should

Probably the browsercache expired and renewed the page (ctrl-f5 wasn't helping).

Strange!

Rivera · Mon Feb 03, 2014 2:20 pm

>OpenVPN server doesn't work! Clients log off immediately after connection establishing. Downgrade to 6.7.

Can confirm. Same goes for PPTP.

Mon Feb 03, 2014 2:21 pm

Anyone with Webfig and Quickset issues, please clear cache and reload the page a few times, the previous version files could be cached in your proxy or browser.

DMK · Mon Feb 03, 2014 2:40 pm

Hi, guys!

I do not recommend updating to 6.9, especially when device is 100 km away from you.
On one of my rb2011 updating lead to rewriting of firewall rule for 8291 port (winbox) and router became unreachable (another rb2011 survived updating without this issue). Another problem on this rb2011 is strange behaviour of routing rules. I use simple config with l2tp and ipsec which started to work unpredictable after update. Downgrading to 5.24 (which is the most stable firmware for my needs) immediately solved all problems.

Best regards,
Dmitry

DLNoah · Mon Feb 03, 2014 3:19 pm

However we still do not have a working CRS with vlan:s and swicthes, it still leaks traffic.

Regards //
// Peter Steen

One thing support recommended to me when I communicated with them on my testing CRS behaving like a hub was to completely factory reset the CRS after applying v6.8rc1. I assume the same advice would apply to someone who is upgrading from v6.6-6.7 to 6.9.

Hello,

RouterOS version v6.8rc1 contains a fix for CRS125 forwarding issue introduced in v6.7 which you have been experiencing:
http://www.mikrotik.com/download/share/ ... 6.8rc1.npk

It is required to reset and reconfigure CRS125 for fixes to take effect.
You can keep all non-switch related configuration from export file for easier import, but switch configuration needs to be manually reconfigured.

Regards,
Janis Becs

voxframe · Mon Feb 03, 2014 3:36 pm

CCR have still disabled snmp and lcd due to sudden death of ports otherwise, it works in that way since ros6.4.

Could you please elaborate on this problem? I have not heard about it. Normally I have SNMP and LCD running on my CCR units without any problems (So far).

Thanks!

crtee · Mon Feb 03, 2014 4:26 pm

6.7 -> 6.9 breaks LACP for me on CCR1036, anyone else?

XTX · Mon Feb 03, 2014 5:49 pm

Hi

6.7 -> 6.9 DDM info is gone in SFP interface (only wavelenght is present, temperature, tx rx power , bias, voltage all gone) CCR 1036-8G-2S, finisar sfp

All the DDM info was present in 6.7...

patrick7 · Mon Feb 03, 2014 9:43 pm

One of my routers is in a colocation 400km away

So I'll wait with upgrading and test it before on my RB750GLs

karina · Mon Feb 03, 2014 11:01 pm

pptp appears broke, trying to manage winbox while in the tunnel, different sections are blank, then winbox shows disconnected

sstp also does not appear to be working at this point either. it does not, get an error at verifying username and password, then windows times out with error 628......

i downgraded back to 6.7, pptp and sstp back to working flawless and significantly faster

guess i can live for a while without the leds working right.

guys fix ppp, you seem to have broken alot

this is on a 2011

have you checked your max mtu possible through the tunnell i find many pptp linkd need a slight drop to 1400

willie2276 · Tue Feb 04, 2014 2:12 am

HELP HELP HELP
After upgradind a RB1100AHx2 to 6.9!!!!!!!!!!!!!!!
L2TP server doesn't work! Clients log off immediately after connection establishing.

mediana · Tue Feb 04, 2014 9:13 am

Model: RB450G
Upgrade from 6.7 to 6.9, IPsec (Site-to-Site) is broken.
Downgrade from 6.9 to 6.7, IPsec works as usual.

TealFrog · Tue Feb 04, 2014 9:21 am

Downloaded RoS upgrade package again and did another reinstall on the RB2011. No issues so far and no high CPU utilization. Odd.

Attempted to upgrade three routers from V6.7 to V6.9
RB493 - Okay
RB493G - Okay
2011UAS-2HnD - Unsuccessful

Attempted twice to upgrade 2011UAS-2HnD router to V6.9, router became unusable running constantly at 100% CPU utilization. I would have liked to have had some diagnostics to post, but router wasn't responding. Both attempts ended with downgrading to V6.7. CPU utilization back to normal (nil) with downgrades.

Regards,

Tue Feb 04, 2014 9:29 am

Model: RB450G
Upgrade from 6.7 to 6.9, IPsec (Site-to-Site) is broken.
Downgrade from 6.9 to 6.7, IPsec works as usual.

I also had this problem. When I investigated I found that the IPSEC policies had been disabled by the upgrade (they were red) I simply enabled them and IPSEC works fine on 6.9

alx26 · Tue Feb 04, 2014 10:00 am

HELP HELP HELP
After upgradind a RB1100AHx2 to 6.9!!!!!!!!!!!!!!!
L2TP server doesn't work! Clients log off immediately after connection establishing.

We observe similar behavior after upgrade, but the problem cause another. There are changes in the log topic "info". In previous versions failed pppoe authentication logging as:

11:27:03 pppoe,info PPPoE connection established from 08:60:6E:62:37:60 
11:27:03 pppoe,ppp,info <pppoe-0>: waiting for call... 
11:27:03 pppoe,ppp,info <pppoe-0>: terminating... - user 002074 authentication failed 
11:27:03 pppoe,ppp,info <pppoe-0>: disconnected

and now same authentication failure looks not so obvious:

11:32:59 pppoe,info PPPoE connection established from 08:60:6E:62:37:60 
11:32:59 pppoe,ppp,info,account 002074 logged out, 0 0 0 0 0

Dear Mikrotik Support Team, сan you say, if it changes in the "logging" or just a bug?

karina · Tue Feb 04, 2014 10:19 am

HELP HELP HELP
After upgradind a RB1100AHx2 to 6.9!!!!!!!!!!!!!!!
L2TP server doesn't work! Clients log off immediately after connection establishing.
We observe similar behavior after upgrade, but the problem cause another. There are changes in the log topic "info". In previous versions failed pppoe authentication logging as:
Code: Select all
11:27:03 pppoe,info PPPoE connection established from 08:60:6E:62:37:60 
11:27:03 pppoe,ppp,info <pppoe-0>: waiting for call... 
11:27:03 pppoe,ppp,info <pppoe-0>: terminating... - user 002074 authentication failed 
11:27:03 pppoe,ppp,info <pppoe-0>: disconnected 
and now same authentication failure looks not so obvious:
Code: Select all
11:32:59 pppoe,info PPPoE connection established from 08:60:6E:62:37:60 
11:32:59 pppoe,ppp,info,account 002074 logged out, 0 0 0 0 0 
Dear Mikrotik Support Team, сan you say, if it changes in the "logging" or just a bug?

have you tried removing the interface and reconfiguring it. maybe something got changed in the upgrade, have read similar issues with other things eg: something got disabled etc etc

JanezFord · Tue Feb 04, 2014 11:07 am

Model: RB450G
Upgrade from 6.7 to 6.9, IPsec (Site-to-Site) is broken.
Downgrade from 6.9 to 6.7, IPsec works as usual.

Check your configs ... I'm running ipsec site-to-site on multiple rb450g without issues ... also interconnecting with with ipsec to rb951g, rb751g, rb800 and rb2011 units ... no issues at all (for now, running 3 days), lots of traffic, encryption aes-128+sha1 on all. All units have been upgraded from 6.7 to 6.9.

JF

DjM · Tue Feb 04, 2014 11:18 am

>OpenVPN server doesn't work! Clients log off immediately after connection establishing. Downgrade to 6.7.

Can confirm. Same goes for PPTP.

I can also confirm issue with VPNs:
- SSTP from MK 6.9 -> MK 6.7 (server) is working
- SSTP in other cases = not working
- OVPN = not working

Downgrade back to 6.7, everything is working.

I haven't time to do more deeper tests, but it seems that not only one network topology is having problem.

willie2276 · Tue Feb 04, 2014 11:23 am

HELP HELP HELP
After upgradind a RB1100AHx2 to 6.9!!!!!!!!!!!!!!!
L2TP server doesn't work! Clients log off immediately after connection establishing.

---------------------
In my case donwgrade is not enough!!!!

After downgrade to 6.7 the result is the same.
To resolve MUST have a configuration backup of your previous 6.7 configuration and restore it.

ilnicchio · Tue Feb 04, 2014 11:28 am

pptp appears broke, trying to manage winbox while in the tunnel, different sections are blank, then winbox shows disconnected

sstp also does not appear to be working at this point either. it does not, get an error at verifying username and password, then windows times out with error 628......

i downgraded back to 6.7, pptp and sstp back to working flawless and significantly faster

guess i can live for a while without the leds working right.

guys fix ppp, you seem to have broken alot

this is on a 2011

For me the same problem. In ROS 6.7 I can reach my router through a pptp tunnel, in ROS 6.9 not (it freeze after some Kb of plugin transfer) maybe there's some problem with "change MSS"?
I can normally reach winbox with full MTU interface (like ether1). I try out with 3 RB with the same results...
Please fix.

Tue Feb 04, 2014 12:20 pm

I just notice that MRRU option in previous versions on some routers was set to 512, but now in v6.9 it is 1600.

I suggest to unset this option, it might cause the problem between versions - this option is only necessary when ppp BCP or MLPPP is used.

karina · Tue Feb 04, 2014 1:01 pm

I see a lot of posts concerning various VPN problems. I am not seeing this on my CCR 10-16 Judging by the symptoms reported I suspect its just a config issue. pay particular attention to MTU values, do a packet fragmentation test to discover the true MTU of the tunnel and adjust as required. I found that Some of my VPN's required a small drop of MTU to 1400. especially pptp tunnels.

karina · Tue Feb 04, 2014 1:08 pm

[/quote]
For me the same problem. In ROS 6.7 I can reach my router through a pptp tunnel, in ROS 6.9 not (it freeze after some Kb of plugin transfer) maybe there's some problem with "change MSS"?
I can normally reach winbox with full MTU interface (like ether1). I try out with 3 RB with the same results...
Please fix.[/quote]

When you create a PPTP interface a mangle rule is automatically created to reduce the MTU of that interface to take into account of the axtra overhead. Using winbox open the interface and on the general tab is a MAX MTU and MAX MRU. I think the default is 1420. try dropping this to 1400

roast · Tue Feb 04, 2014 3:16 pm

Somehow OpenVPN has become unstable with 6.9. It was super stable in 6.7 now I get disconnects every few hours...

V6.6 to V6.9 -> OpenVPN stopped working for me. Switched to SSTP..........

roast · Tue Feb 04, 2014 3:20 pm

For me. with this FW on a RB-450 routing marks stopped working.
- mark route in mangle/prerouting
- set another default route for that routing mark (this one shows blue in the list, meaning it has no effect)
- marked packets still go out on the default gw.

The same config worked flawless in 6.7

Can I have a download link for 6.7 so I can downgrade?

I have the same problem!

I mark route in mangle for prerouting, goes out the default gateway...

littlebill · Tue Feb 04, 2014 4:11 pm

>OpenVPN server doesn't work! Clients log off immediately after connection establishing. Downgrade to 6.7.

Can confirm. Same goes for PPTP.
I can also confirm issue with VPNs:
- SSTP from MK 6.9 -> MK 6.7 (server) is working
- SSTP in other cases = not working
- OVPN = not working

Downgrade back to 6.7, everything is working.

I haven't time to do more deeper tests, but it seems that not only one network topology is having problem.

well i have the identical setup and the identical issue. thus this is bugged big time

littlebill · Tue Feb 04, 2014 4:12 pm

For me the same problem. In ROS 6.7 I can reach my router through a pptp tunnel, in ROS 6.9 not (it freeze after some Kb of plugin transfer) maybe there's some problem with "change MSS"?
I can normally reach winbox with full MTU interface (like ether1). I try out with 3 RB with the same results...
Please fix.[/quote]

When you create a PPTP interface a mangle rule is automatically created to reduce the MTU of that interface to take into account of the axtra overhead. Using winbox open the interface and on the general tab is a MAX MTU and MAX MRU. I think the default is 1420. try dropping this to 1400[/quote]

pptp works 1 in 3 times. why should i need to adjust an mtu, when it didn't change from 6.7

everyone is having the same issue. MTU is not the problem here

sstp is broke period for windows. this is more then a coincidence

pcunite · Tue Feb 04, 2014 4:20 pm

For those of you having issues with 6.9: after updating and also updating the firmware to 3.10, then you'll need to do a system reset-configuration and then manually apply your custom settings.

Neohunter · Tue Feb 04, 2014 5:24 pm

I update RB750G from 6.7 to 6.9 via system/package/check for updates - download & install
router reboot but not start - also all Ethernet interface led not flashed

i try repair it via netinstall - connect 1 eth on 750G to pc eth card
push rest and start router - wait so long but act not flashed !
and i can't see router in netinstall

How i can reanimate router ?
Hello,
1. try to press reset button for 3 secs and search 750G in winbox. if it wont help you then
2. press reset button for 30s and release. Device should appear in netinstall list.

ps. netinstall works only on poe-in interface (ether1?))
pps. are you sure, that netinstall prepared correctly? ie disabled firewall, set ip address

Good luck

I do it all, but router looks like brick ;(
I don't know what make update if after that router not booting and all LED not lights
Only PWR LED lightnings

patrickmkt · Tue Feb 04, 2014 5:45 pm

For those of you having issues with 6.9: after updating and also updating the firmware to 3.10, then you'll need to do a system reset-configuration and then manually apply your custom settings.

You mean that for every upgrade we have to completely reset the router, reprogram manually all the config (because reimporting the config will create the same issue, and a script will not work as for every version update Mikrotik made undocumented change on small syntax that creates error in scripts). That's a few hours job for each router + a few weeks of survey to debug any potential error.

It's like saying every time you have a windows update you need to reformat your hard drive and reinstall all softwares. How stupid does it sound?

mediana · Tue Feb 04, 2014 6:14 pm

Model: RB450G
Upgrade from 6.7 to 6.9, IPsec (Site-to-Site) is broken.
Downgrade from 6.9 to 6.7, IPsec works as usual.
I also had this problem. When I investigated I found that the IPSEC policies had been disabled by the upgrade (they were red) I simply enabled them and IPSEC works fine on 6.9

You are right! I didn't notice that and now it works. Thanks a lot!

pcunite · Tue Feb 04, 2014 7:22 pm

You mean that for every upgrade we have to completely reset the router, reprogram manually all the config (because reimporting the config will create the same issue, and a script will not work as for every version update Mikrotik made undocumented change on small syntax that creates error in scripts). That's a few hours job for each router + a few weeks of survey to debug any potential error.

It's like saying every time you have a windows update you need to reformat your hard drive and reinstall all softwares. How stupid does it sound?

I don't disagree with you - just offering a solution. Note that you can do an export file=MyFile and mostly copy and paste that back into a terminal window.

pjhb34fg · Tue Feb 04, 2014 8:36 pm

Dunno if right topic.

Got 951G-2HnD, bought it with l4 license and 5.25 OS version.
After upgrading to 6.7 and firmware to 3.10 got frequent internet disconnects (wan port traffic dies in statistics) during high bandwith utilization, (only reboot or 15 min of waiting helps, strange though) I've reseted config and made new manually, i ve got simple ap mode with nat and wifi.
Wan port is pure ethernet with dhcp, nothing in logs.
Upgraining to 6.9 didnt resolve my issue, reseting config and again making it manually again didnt help, l dont see anything strange in logs, traffic just stops on wan port.
Got no POE enabled, Had to downgrade to 5.26.
Anyone saw similar problems?

ffernandes · Tue Feb 04, 2014 8:40 pm

\0 ...
my rb 435G just went down with kernel failure.....
had to take power down/on and after a few tries it came up again...
ros 6.9 of 30/01/14 11:18:19 release....

anyone else had this?!

Tue Feb 04, 2014 9:59 pm

Model: RB450G
Upgrade from 6.7 to 6.9, IPsec (Site-to-Site) is broken.
Downgrade from 6.9 to 6.7, IPsec works as usual.
I also had this problem. When I investigated I found that the IPSEC policies had been disabled by the upgrade (they were red) I simply enabled them and IPSEC works fine on 6.9
You are right! I didn't notice that and now it works. Thanks a lot!

Mikrotik you might want to fix this for 6.10

paoloaga · Wed Feb 05, 2014 2:37 am

Hello!

I am upgrading customer devices from 6.1 to 6.9 and the provisioning script stopped working;

I want to configure an account if it doesn't already exist:

{
:local uname
:set uname [/user get [find name=someaccount] name]
:if ($uname = "someaccount") do={
/log info message="Updating account"
/user set someaccount group="full" password=$adminSecret comment="Helpdesk default access"
} else={
/log info message="Adding account"
/user add name=someaccount group="full" password=$adminSecret comment="Helpdesk default access"
}
}

When the account does not already exists, in v6.1 it was working properly. With 6.7 and 6.9 the script fails with a "no such item" error.

Is it a new behaviour or is it a bug?

Wed Feb 05, 2014 8:53 am

You mean that for every upgrade we have to completely reset the router, reprogram manually all the config (because reimporting the config will create the same issue, and a script will not work as for every version update Mikrotik made undocumented change on small syntax that creates error in scripts). That's a few hours job for each router + a few weeks of survey to debug any potential error.

It's like saying every time you have a windows update you need to reformat your hard drive and reinstall all softwares. How stupid does it sound?
I don't disagree with you - just offering a solution. Note that you can do an export file=MyFile and mostly copy and paste that back into a terminal window.

This was only necessary for CRS125 to prevent speed issues and behaviour of a hub.
Alternatively you can enter this command:

/interface ethernet switch port set [find] learn-restricted-unknown-sa=yes

Wed Feb 05, 2014 10:58 am

For those of you having issues with 6.9: after updating and also updating the firmware to 3.10, then you'll need to do a system reset-configuration and then manually apply your custom settings.
You mean that for every upgrade we have to completely reset the router, reprogram manually all the config (because reimporting the config will create the same issue, and a script will not work as for every version update Mikrotik made undocumented change on small syntax that creates error in scripts). That's a few hours job for each router + a few weeks of survey to debug any potential error.

It's like saying every time you have a windows update you need to reformat your hard drive and reinstall all softwares. How stupid does it sound?

Like Becs says, this is not true and is not needed at all.

Polky · Wed Feb 05, 2014 11:46 am

Hi, after upgrade to 6.9 from 6.8rc1, ceased to function gateway with routing mark.
Unable to select routing mark! - Is empty.
Donwgrade back 6.8rc1 works.

THX

p2systems · Wed Feb 05, 2014 1:06 pm

x86 Server
Upgrade 6.1 to 6.9

After the upgrade, DHCP Server is using 100% CPU on one core.
If you disable IPv6 DHCP PD in PPPoE Profiles an reboot, everything is fine. But then there is no IPv6 anymore for the customers.

Ticket#2014020466000563

Update: Support is Aware of this bug. A fix will be included in 6.10

Wed Feb 05, 2014 1:16 pm

Dunno if right topic.

Got 951G-2HnD, bought it with l4 license and 5.25 OS version.
After upgrading to 6.7 and firmware to 3.10 got frequent internet disconnects (wan port traffic dies in statistics) during high bandwith utilization, (only reboot or 15 min of waiting helps, strange though) I've reseted config and made new manually, i ve got simple ap mode with nat and wifi.
Wan port is pure ethernet with dhcp, nothing in logs.
Upgraining to 6.9 didnt resolve my issue, reseting config and again making it manually again didnt help, l dont see anything strange in logs, traffic just stops on wan port.
Got no POE enabled, Had to downgrade to 5.26.
Anyone saw similar problems?

pjhb34fg, if your RB951G-2HnD works with RouterOS v5.26 but does not work properly with RouterOS version 6.x, please report this to MikroTik Technical Support and send supout.rif file from RouterOS v6.9.

bjorncmtec · Wed Feb 05, 2014 2:37 pm

>OpenVPN server doesn't work! Clients log off immediately after connection establishing. Downgrade to 6.7.

Can confirm. Same goes for PPTP.
I can also confirm issue with VPNs:
- SSTP from MK 6.9 -> MK 6.7 (server) is working
- SSTP in other cases = not working
- OVPN = not working

Downgrade back to 6.7, everything is working.

I haven't time to do more deeper tests, but it seems that not only one network topology is having problem.

Try disabling all auth and cypher options in OVPN Setup window.

Worked for me. REALLY stupid and is making me lose confidence in Mikrotik with each 6.x release

doush · Wed Feb 05, 2014 2:55 pm

PPPoE Auth failures shows in the logs as follows

14:52:27 pppoe,ppp,info,account [username] logged out, 1 0 0 0 0

Sorry but what a mess this version is !

bysard · Wed Feb 05, 2014 3:02 pm

PPPoE Auth failures shows in the logs as follows
Code: Select all
14:52:27 pppoe,ppp,info,account [username] logged out, 1 0 0 0 0 
Sorry but what a mess this version is !

Big mess. I'll stay at 5.26 with all our routers except RB2xxx, which I rather put in the closed and out of production as they are not stable at all. v6.x so far is the biggest dissapointment after bloated capacitors ob RB4xxx series. I'm sure we will be disappointed even more in the future as it seems that the quality of the software is now also degrading, not only hardware.

mietus · Wed Feb 05, 2014 4:26 pm

PPPoE Auth failures shows in the logs as follows
Code: Select all
14:52:27 pppoe,ppp,info,account [username] logged out, 1 0 0 0 0 
Sorry but what a mess this version is !

i confirm http://forum.mikrotik.com/viewtopic.php?f=2&t=81398
it will be fixed in 6.10 (maybe).

I think that such major changes should be relased as test package for some time.

koshak83 · Wed Feb 05, 2014 4:50 pm

Hmm: I'm install RouterOS v6.9 (hardreset-> netinstall) to my RB951G-2HnD, but dont have bootloader v3.12 firmware update...

bryans2k · Wed Feb 05, 2014 7:31 pm

nvm, switch related.

hugleo · Wed Feb 05, 2014 8:13 pm

Openvpn is not working in this version...

3bs · Thu Feb 06, 2014 1:21 am

Big mess. I'll stay at 5.26 with all our routers except RB2xxx, which I rather put in the closed and out of production as they are not stable at all. v6.x so far is the biggest dissapointment after bloated capacitors ob RB4xxx series. I'm sure we will be disappointed even more in the future as it seems that the quality of the software is now also degrading, not only hardware.

+1

noyo · Thu Feb 06, 2014 1:25 am

Hello!
I did upgrade to 6.9.
Is the upgrade to 3.12 will kill my omni? Yes/no ?

FW.png

ndbjorne · Thu Feb 06, 2014 1:38 am

I tried a few downgrades and then tried to upgrade at least to 6.7 but CPU usage issue remains.
Now unistalled all unused packages on RB711UA-2HnD and it works properly running 5.14:

> system package print
Flags: X - disabled
 #   NAME                    VERSION                    SCHEDULED
 0   system                  5.14
 1   routeros-mipsbe         5.14
 2   dhcp                    5.14
 3   wireless                5.14
 4   ppp                     5.14
 5   routerboard             5.14
 6   routing                 5.14
 7   advanced-tools          5.14
 8   security                5.14
> while ( true ) do={ system resource cpu print ;delay  10}
 # CPU                           LOAD         IRQ        DISK
 0 cpu0                            1%          0%          0%
 0 cpu0                            1%          0%          0%
 0 cpu0                            1%          0%          0%
 0 cpu0                            1%          0%          0%
 0 cpu0                            1%          0%          0%
 0 cpu0                            1%          0%          0%
 0 cpu0                            1%          0%          0%
 0 cpu0                            1%          0%          0%
 0 cpu0                            1%          0%          0%
 0 cpu0                            1%          0%          0%
 0 cpu0                            1%          0%          0%
 0 cpu0                            1%          0%          0%
interrupted
>  system resource print    
                   uptime: 1d14h58m48s
[...]

I'll give 6.9 another chance in a few days.

All other systems I upgraded to 6.9 (1*1200, 6*2011L, 1*2011UAS-2HnD, 3*951-2n, 1*x86 over VMWare) seem to work fine, all have simple configurations.

Bye

ndbjorne · Thu Feb 06, 2014 1:51 am

Hello!
I did upgrade to 6.9.
Is the upgrade to 3.12 will kill my omni? Yes/no ?

All my RB upgraded to 6.9 that work fine show a 3.10 firmware:

> system routerboard print 
       routerboard: yes
             model: 1200
     serial-number: [...]
  current-firmware: 3.10
  upgrade-firmware: 3.10

The only one that showed a 3.12 is the 711 with the CPU usage issue which I had to downgrade to 5.14 to solve:

>  system routerboard print 
       routerboard: yes
             model: 711UA-2HnD
     serial-number: [...]
  current-firmware: 3.10
  upgrade-firmware: 2.38

I'd wait a while.

Bye

zyzelis · Thu Feb 06, 2014 9:11 am

Hello!
I did upgrade to 6.9.
Is the upgrade to 3.12 will kill my omni? Yes/no ?
FW.png

noyo,
after upgrade both ros and bootloader, my Omnitiks still works:)
and im observing better nv2 behavior and less cpu usage, when opened winbox.

So it will not kill your omni

JanezFord · Thu Feb 06, 2014 9:56 am

Hello!
I did upgrade to 6.9.
Is the upgrade to 3.12 will kill my omni? Yes/no ?
All my RB upgraded to 6.9 that work fine show a 3.10 firmware:
Code: Select all
> system routerboard print 
       routerboard: yes
             model: 1200
     serial-number: [...]
  current-firmware: 3.10
  upgrade-firmware: 3.10
The only one that showed a 3.12 is the 711 with the CPU usage issue which I had to downgrade to 5.14 to solve:
Code: Select all
>  system routerboard print 
       routerboard: yes
             model: 711UA-2HnD
     serial-number: [...]
  current-firmware: 3.10
  upgrade-firmware: 2.38
I'd wait a while.

Bye

I'm running six 711-UA-2HnD with 6.9 and firmware 3.12 ... they serve as public/private access points with multiple ssids bridged to proper vlans ... about 20 clients each, runing 5 days now, no CPU issues, cpu load is maybe a notch lower than it was with 6.7 and firmware 3.10.

JF

Raf · Thu Feb 06, 2014 10:32 am

Hello!
I did upgrade to 6.9.
Is the upgrade to 3.12 will kill my omni? Yes/no ?
FW.png

No. Firmware should ALWAYS be updated. Check RouterBOOT changelog if You want.

Masyanich · Thu Feb 06, 2014 10:33 am

Am I the only one who see some port flapping on CCR once again?

noyo · Thu Feb 06, 2014 11:32 am

Hello!
I did upgrade to 6.9.
Is the upgrade to 3.12 will kill my omni? Yes/no ?
FW.png
noyo,
after upgrade both ros and bootloader, my Omnitiks still works:)
and im observing better nv2 behavior and less cpu usage, when opened winbox.

So it will not kill your omni

thanks, it works
nv2, nstream is much better, stable

killersoft · Thu Feb 06, 2014 1:49 pm

v6.9 works on my Metal 2SHPn (8 Km link @ 802.11 mode ) no problem yet !!

TealFrog · Thu Feb 06, 2014 2:17 pm

The RB2011UAS-2HnD ran on 6.9 for a couple of days. Today, it lost its primary gateway connection and would not get a DHCP lease from ISP. Rebooted router and 100% CPU utilization issue occurred. I could not get the usage to drop and attempted several more reboots. I also attempted to get a support.rif file, but was unsuccessful. Downgraded from 6.9 -> 6.7 everything working again.

Chupaka · Thu Feb 06, 2014 2:51 pm

Please, fix vlans. I still use 6.3 becouse of it

VLANs are heaving a lots of problems in 6.9.

what VLAN problems are you talking about?

I am upgrading customer devices from 6.1 to 6.9 and the provisioning script stopped working;

I want to configure an account if it doesn't already exist:

{
:local uname
:set uname [/user get [find name=someaccount] name]
:if ($uname = "someaccount") do={
/log info message="Updating account"
/user set someaccount group="full" password=$adminSecret comment="Helpdesk default access"
} else={
/log info message="Adding account"
/user add name=someaccount group="full" password=$adminSecret comment="Helpdesk default access"
}
}

When the account does not already exists, in v6.1 it was working properly. With 6.7 and 6.9 the script fails with a "no such item" error.

Is it a new behaviour or is it a bug?

I think it's a new behaviour. Change

:set uname [/user get [find name=someaccount] name]

to something like

:foreach i in=[/user find name=someaccount] do={ :set $uname [/user get $i name]; }

chris6671980309 · Thu Feb 06, 2014 3:18 pm

In the 6.9 release, there is a "updated bootloader firmware". But I didn't see any new version then the current 3.10.

planetcaravan · Thu Feb 06, 2014 5:18 pm

I have RB2011 working on 6.0
The router is ~100km away from my office.
I really want to upgrade it but I don't have intensions to travel on site to do downgrade.
Is it safe to upgrade to 6.9 RB2011?

I'm using RB2011 as router for adsl, vpn, hotspot, sstp-client. Now is running 6.9 and all is working fine. I've updated from 6.7.

planetcaravan · Thu Feb 06, 2014 5:57 pm

Does someone noticed CCQ changes on RouterOS 6.9?

beamer · Thu Feb 06, 2014 6:37 pm

pptp appears broke, trying to manage winbox while in the tunnel, different sections are blank, then winbox shows disconnected

Same problem here! The PPTP session itself stays on, as well as a RDP session. Winbox in the RDP session is also fine, but a Winbox session through the VPN tunnel (PPTP) is disconnected after a short time.

Changing the "Max MTU" and "Max MRU" from 1450 to 1400 on the PPTP server didn't help.

Sometimes there's no traffic at all passing the PPTP tunnel when connecting from Mac OS X. The PPTP session is there but zero traffic going through! This seems not to happen when connecting with Win7.

i downgraded back to 6.7, pptp and sstp back to working flawless and significantly faster

Is downgrading as easy as upgrading? Just copy the packages and reboot? Config stays?

guys fix ppp, you seem to have broken alot

+1

tom211 · Thu Feb 06, 2014 8:01 pm

SSTP not working, slow VLANs, what's next...

I really doubt whether it was a good idea to replace all our switches and routers with MT...

ndbjorne · Thu Feb 06, 2014 8:22 pm

I tried a few downgrades and then tried to upgrade at least to 6.7 but CPU usage issue remains.
Now unistalled all unused packages on RB711UA-2HnD and it works properly running 5.14
[...]
I'll give 6.9 another chance in a few days.

Had a few time: 5.14 -> 6.9 (same packages): no CPU issue
Added all packages loaded when running 6.7: no CPU issue

Only missing step is upgrade firmware to 3.12 but why if all of the other systems I upgraded to 6.9 with no issue show a 3.10 and only this one shows a 3.12?

skibi82 · Thu Feb 06, 2014 9:38 pm

I have to admit the pain in my heart that release 6.9 is the biggest disgrace.
Where to start:
1. As before predecessors wrote that the SSTP server is not running properly. I need only confirm the continuous reconect and of the initialization phase of the tunnel cpu going to head.
2 ipsec flusch sa make he router unstable (kill the router when the daily execution of the script, thanks god as a watchdog on the ip set it will then reboot, I wish that I had descended Sunday on the ends of the vpn set saved me the $ 100 on fuel)
3 For some reason no ip packing rb951g does not work after downgrade to 5.x works and answer from Mikrotik that did not observe the problem to be honest ...

My last words to Release 6.9. Was supposed to be good and work out how to always suck.
PS.
When is 6.10 with bug fixes?
Regards
Jaroslaw Skibiński

mspeed · Fri Feb 07, 2014 12:49 am

We are seeing crashes over multiple devices with CCR1036-12G-4S i'm afraid - appears to be related to VRRP.

Kernel error and reboot, then both routers attempt to act as master but are unable to pass traffic. Manually removing/readding vrrp or disabling VRRP allows traffic again, however we've then observed ongoing crashes due to kernel panic, or the watchdog timer which can no longer reach its watch address.

Have sent urgent sup outs to Mikrotik but am pretty disheartened. We upgraded on their recommendation to fix another crash in 6.8. The supouts can't even be read on mikrotik.com supout viewer.

Seems like Mikrotik may not be stable enough for our type of environment.. it's a real shame as we've now over 30 of the CCR's and really wanted to believe in the platform.

We're using CCR for a very simple deployment - BGP + VRRP + VLAN + Basic filtering + Netflow

sdv · Fri Feb 07, 2014 9:22 am

Upgrade: 6.6 -> 6.9
RouterBOARD: RB 1100AHx2
Tested bugs:
а. [L2TP-Server]:
After upgrade router reboots while first l2tp-session established with watchdog exeption message on serial console.
L2tp-server (and router) crash caused by the use of MRRU (value doesn't matter) parameter in /interface l2tp-server server configuration, so multilink-PPP over single link seems to be broken.

Upgrade: 6.6 -> 6.9
RouterBOARD: RB750G
Tested bugs:
a. [L2TP-Client]:
After upgrade router not always mark as active default route learned via OSPF / Static via L2TP-tunnel.
This issue caused by the use of routing-marks (with out them all seems to be working), reboot fix this situation.

DjM · Fri Feb 07, 2014 9:46 am

>OpenVPN server doesn't work! Clients log off immediately after connection establishing. Downgrade to 6.7.

Can confirm. Same goes for PPTP.
I can also confirm issue with VPNs:
- SSTP from MK 6.9 -> MK 6.7 (server) is working
- SSTP in other cases = not working
- OVPN = not working

Downgrade back to 6.7, everything is working.

I haven't time to do more deeper tests, but it seems that not only one network topology is having problem.
Try disabling all auth and cypher options in OVPN Setup window.

Worked for me. REALLY stupid and is making me lose confidence in Mikrotik with each 6.x release

Hello bjorncmtec,

thank you for hint, I will try it. As you have mentioned, it will be really mess-making in case that this configuration will work - no cypher option selected as allowed and it will start working

Jetrider · Fri Feb 07, 2014 10:09 am

6.9 still has bug same as the 6.7 version. Ip addresses are disappeared, dhcp stops working, pppoe is not working, more then a half of services is not usable and reboot is necessary. It happens on all RBs!!!! Now it's happened on CCR which is core router in my network. It's wrong when 10.000 users is left without service only because Mikrotik is not fixing reported bugs!

I can second this. Still hanging on v6.5 because of this. The way MT treats them bugreports is just ridiculous. When I email their support, replies I get makes me think that they truly believe I have to turn production router for quite a bunch of customers into ROS testbed. To be honest, it is kind of scary. One thing is if you upgrade some chain of wireless backbone or customer CPE and something goes wrong - no big damage and barely anyone cares. Now that they have introduced CCR, they also have introduced the amount of damage they can do with a single buggy version. I truly dont know how do MT tests those supout.rif files when sent over, what bugs me is this scenario:
we do a nightly upgrade CCR to newest version -> primary tests shows basic functions are OK -> morning comes, specific but critical bug hits in, customer line goes hot -> general emergency issues, we immediately downgrade -> now after the dust have settled, we make supout, detailed description of behavior and provide it to MT.
Then the fun part begins - when asked for bug to be fixed, fancy requests are being made on behalf of MT - like "ok, upgrade it again and when bug hits you, make supout and send it to us". Why ? Is this the way it is meant to be ? Is there really no way for MT to test that themselves in the lab (now that they know the problem), instead of us experimenting on our ranting customers ?

karina · Fri Feb 07, 2014 10:45 am

Upgrade party through the night, 433's sxt's 411's omnitiks RB1100ah CCR's all 100% no problems all upgrades smooth. well done mikrotik Nice release for me. No problems with Pptp Tunnels all working well

Jetrider · Fri Feb 07, 2014 10:48 am

Upgrade party through the night, 433's sxt's 411's omnitiks RB1100ah CCR's all 100% no problems all upgrades smooth. well done mikrotik Nice release for me. No problems with Pptp Tunnels all working well

Does your CCR have any DHCP server on any VLAN ?

Sander · Fri Feb 07, 2014 11:02 am

pptp appears broke, trying to manage winbox while in the tunnel, different sections are blank, then winbox shows disconnected
Same problem here! The PPTP session itself stays on, as well as a RDP session. Winbox in the RDP session is also fine, but a Winbox session through the VPN tunnel (PPTP) is disconnected after a short time.

Same problem. WinBox through the PPTP is disconnected within one minute after upgrade from 6.7 to 6.9 on RB2011L.

honzam · Fri Feb 07, 2014 12:33 pm

Upgrade from v5.26 to 6.9 - After upgrade are all packages enabled.

For example. I have in 5.26 disabled - routing, ppp, MPLS, hotspot.
After upgrade are this packages enabled. Problem is in all v6.x versions

karina · Fri Feb 07, 2014 1:14 pm

this is not a bug this is normal simply disable the required packages again

karina · Fri Feb 07, 2014 1:15 pm

Upgrade party through the night, 433's sxt's 411's omnitiks RB1100ah CCR's all 100% no problems all upgrades smooth. well done mikrotik Nice release for me. No problems with Pptp Tunnels all working well
Does your CCR have any DHCP server on any VLAN ?

No.

karina · Fri Feb 07, 2014 1:21 pm

pptp appears broke, trying to manage winbox while in the tunnel, different sections are blank, then winbox shows disconnected
Same problem here! The PPTP session itself stays on, as well as a RDP session. Winbox in the RDP session is also fine, but a Winbox session through the VPN tunnel (PPTP) is disconnected after a short time.
Same problem. WinBox through the PPTP is disconnected within one minute after upgrade from 6.7 to 6.9 on RB2011L.

I have a bunch of pptp tunnels on CCR no issues other than tweaking MTU values for max performance. Make sure keep alive packets are not being blocked by firewall rule. I had this issue after an upgrade.

Address lists may be an issue as V6.9 does not allow any duplicate entries where older versions did. I do not know what would happen if an upgrade is performed on a router with duplicate entries

beamer · Fri Feb 07, 2014 2:45 pm

I opened a ticket and just got his reply but haven't tested it yet:

Should be fixed

links removed by admin

honzam · Fri Feb 07, 2014 3:13 pm

this is not a bug this is normal simply disable the required packages again

This is not normal.
From 3.30 -> 4.17 packages stay disabled
from 4.17 -> 5.26 packages stay disabled
Only from 5.26 -> 6.x all packages are enabled.

It's additional work (wasted) when you migrate complete network to V6

honzam · Fri Feb 07, 2014 3:32 pm

I opened a ticket and just got his reply but haven't tested it yet:

The following RouterOS development versions should be for internal testing only. Please do not distribute or publish for public in any way. You can test these versions in your own, or customer networks

Fri Feb 07, 2014 3:35 pm

These beta (6.10) versions are internal FOR A REASON. They WILL contain problems, because we are not done with them. If you want to test, that's fine, but please don't confuse other forum users into thinking that this is a stable release.

lavv17 · Fri Feb 07, 2014 6:29 pm

For some reason after upgrade from 6.7 to 6.9 bonding stopped working, interfaces don't join together.

Flags: X - disabled, R - running
0 R name="bonding1" mtu=1500 mac-address=D4:CA:6D:77:6D:34 arp=enabled
slaves=ether1,ether2 mode=802.3ad primary=none link-monitoring=mii
arp-interval=100ms arp-ip-targets="" mii-interval=100ms down-delay=0ms
up-delay=0ms lacp-rate=30secs transmit-hash-policy=layer-2-and-3

Group Port-channel Protocol Ports
------+-------------+-----------+-----------------------------------------------
1 Po1(SD) LACP Gi0/14(I) Gi0/15(I)
I - stand-alone

Any hints?

mspeed · Fri Feb 07, 2014 6:59 pm

Then the fun part begins - when asked for bug to be fixed, fancy requests are being made on behalf of MT - like "ok, upgrade it again and when bug hits you, make supout and send it to us". Why ? Is this the way it is meant to be ? Is there really no way for MT to test that themselves in the lab (now that they know the problem), instead of us experimenting on our ranting customers ?

Just manually generate the supout before you downgrade, takes a minute and you can retrieve it after you've restored service and send the report. Autosupout doesn't always work reliably either.

My problem is that I do this and send in the report, but it is often 3-4 days before any Mikrotik response. Sadly we're a few crashes away from moving off the MT platform alltogether.

It seems to have so much potential, but not the support to back up the product in any real customer facing/production environment that can't deal with extended downtime.

pmuch · Fri Feb 07, 2014 7:14 pm

We are seeing crashes over multiple devices with CCR1036-12G-4S i'm afraid - appears to be related to VRRP.

Kernel error and reboot, then both routers attempt to act as master but are unable to pass traffic. Manually removing/readding vrrp or disabling VRRP allows traffic again, however we've then observed ongoing crashes due to kernel panic, or the watchdog timer which can no longer reach its watch address.

The same issue on RB750GL with VRRP on 6.7.
For me the solution was to disable TCP Invalid packages filtering - with this release VRRP packages became invalid(?)

pcunite · Fri Feb 07, 2014 7:40 pm

Sadly we're a few crashes away from moving off the MT platform alltogether ... It seems to have so much potential, but not the support to back up the product in any real customer facing/production environment that can't deal with extended downtime.

Agreed, fingers crossed MikroTik can pull out of this and create a proper internal dev/testing environment.

beamer · Fri Feb 07, 2014 7:54 pm

These beta (6.10) versions are internal FOR A REASON. They WILL contain problems, because we are not done with them.

I really don't understand MikroTik's software development approach. Why does a new release (minor version) usually contains a lot of bugs that weren't present in the previous version? You should maybe limit the introduction of new features to major versions and focus on BUGS in minor versions. It seems, if you add a "feature" like "added lcd option to change the color-scheme" then essential features like VPNs or auto-negotiation are broken?

It's such a risk to update a core router, I'm really getting sick of it. The change log of 6.8/6.9 sounded like many bugs are solved (including some newly "introduced" in 6.6), but on the other side a lot of NEW BUGS are introduced again... and so on... and so on... never-ending story?

It would be much better if there's a version which is marked as STABLE/PRODUCTION, and other releases are labeled as BETA or something. In case of CCR there's not even a previous major version that could be considered stable...

So if I update to the "internal" 6.10 now to make PPTP work again, to which version should I upgrade later without risking a desaster? The "final" 6.10? An even better 6.11? Or 7.0 sometime? Surely not. 7.x? When? Which version will be STABLE??

djdrastic · Fri Feb 07, 2014 11:29 pm

While I cannot condone the way Mikrotik goes along with their software release business do none of you test anything ? We always have some spare physical hardware and old laptop with virtualbox somewhere in the lab to test this configuration on the new version.

We always keep a good eye on the forums and probably wait roughly 2 months or so after release before we upgrade.

rodneygci · Fri Feb 07, 2014 11:45 pm

Hello,

We use Mikrotiks for edge devices for our IP phone deployments.

One of the requirements for these devices is that they provide a DHCP option (66) as ascii.

I upgraded a device from 6.7 to 6.9 and this option no longer works.

Here is the exact config:

/ip dhcp-server option
add code=66 name=AllworxVGW value="'192.168.2.254'"

Reverting back to 6.7 corrects the issue.

Thanks!

doush · Sat Feb 08, 2014 12:17 am

PPPoE Auth failures shows in the logs as follows
Code: Select all
14:52:27 pppoe,ppp,info,account [username] logged out, 1 0 0 0 0 
Sorry but what a mess this version is !

I am quoting my self here but still I am having a very diffucult time to understand how come such a bug can be introduced !
Even if they did minimal testing, they could spot this stupid bug in a second.

My advice to Mikrotik:
Please, please, please STOP adding new features ! JUST STOP ! Concentrate on bug fixing + multicore enhancements and thats it ! Give us a rock solid version with no showstoppers like VPN problems, CPU problems, boot problems etc... Test it 1000 times before the release and publish.
Just one stable branch in v6.x and thats it ! We are not asking much.

beamer · Sat Feb 08, 2014 2:26 am

Should be fixed

OK, I updated the CCR to "internal" 6.10 and it's NOT fixed. Winbox sessions through VPN (PPTP) are still disconnecting. Try to transfer a file with a couple 100 KB, very unlikely that it'll finish.

I tested with MaxMTU/MaxMRU of PPTP server set to 1400 and 1450, same result.

heislerb · Sat Feb 08, 2014 4:39 am

RB750GL upgraded ROS from 6.7 to 6.9 (firmware from 3.10 to 3.12). Script no longer works. Down grade to 6.7 and script works. I am not good at doing scripts. Can anyone quickly tell me what 6.9 found to be a problem. Note; value of > 330000 is only for testing,
PLEASE ANSWER in this thread http://forum.mikrotik.com/viewtopic.php?f=9&t=81722

:local traffic
:local nam
:local macblock
:local date
:local day
:local month
:local year
:local tim

:set date [system clock get date]
:set month [:pick $date 0 3]
:set day [:pick $date 4 6]
:set year [:pick $date 7 11]
:set tim [system clock get time]

/queue simple
:for i from=253 to=254 do={
:set traffic [get [find target="192.168.88.$i/32"] total-bytes]
:if ($traffic < 330000) do={
:set nam [get [find target="192.168.88.$i/32"] name ]

/ip dhcp-server lease
:set macblock [get [find address="192.168.88.$i/32" ] mac-address ]

/ip firewall filter add action=reject chain=forward out-interface=ether1-gateway protocol=tcp reject-with=tcp-reset src-mac-address=$macblock

/queue simple reset-counters $nam

/system scheduler add interval=1d name=$macblock on-event= test policy= ftp,reboot,read,write,policy,test,winbox,password,sniff,sensitive,api start-date="$month/$day/$year" start-time=$tim

/queue simple
}
}
EDIT - support fixed the problem. I had no queue for target=192.168.88.253/32
- 6.9 has better error checking then 6.7
- this was a test script and that is the reason the queue was not present

Sander · Sat Feb 08, 2014 4:49 am

It's so weird. After I upgraded my RB2011L from 6.7 to 6.9. Routing Mark in "/ip route" doesn't work. It just like add a standby route in Main table. The color is blue in WinBox. I can't find any workaround for it, so I downgrade my RB2011L to 6.7. However, There is no problem on my RB951-2n after I upgraded RouterOS from 6.7 to 6.9. Routing Mark in "/ip route" works on RB951-2n.

Neilson · Sat Feb 08, 2014 5:07 am

@doush

For the below, I understand that this release the changes in PPP was to enable it for multicore properly (as in a clean rewrite of the deployment)

Therefore this is a multicore improvement as you ask for.

The output for an incorrect login is unusual however this is the first full production release of this so we can install it in our lab setups, find these differences, ask Mikrotik to change things if we want them to, then when we are ready we can deploy these things on our production systems.

Regards
Alexander

PPPoE Auth failures shows in the logs as follows
Code: Select all
14:52:27 pppoe,ppp,info,account [username] logged out, 1 0 0 0 0 
Sorry but what a mess this version is !
I am quoting my self here but still I am having a very diffucult time to understand how come such a bug can be introduced !
Even if they did minimal testing, they could spot this stupid bug in a second.

My advice to Mikrotik:
Please, please, please STOP adding new features ! JUST STOP ! Concentrate on bug fixing + multicore enhancements and thats it ! Give us a rock solid version with no showstoppers like VPN problems, CPU problems, boot problems etc... Test it 1000 times before the release and publish.
Just one stable branch in v6.x and thats it ! We are not asking much.

ferywu · Sat Feb 08, 2014 12:10 pm

Just info, Using 6.9 on RB1000, running metarouter, every 5 minutes reboot itself.
With metarouter disabled, everything fine.

I have to downgrade now.

jackwing · Sat Feb 08, 2014 3:23 pm

dhcp does not work with trunk port on switch

if no vlan and trunk port , it can work

here is the Capture

heislerb · Sun Feb 09, 2014 4:33 am

RB750GL upgraded from 6.7 to 6.9.

The following command works in 6.7 and not in 6.9.

6.7 on RB450G or RB750
[admin@RB450G-Main] /queue simple> get [find target="192.168.1.199"] total-bytes
[admin@RB450G-Main] /queue simple>

6.9
[admin@RB750GL-Bob] /queue simple> get [find target="192.168.88.254"] total-bytes
no such item

This is why my script is broken and at this moment in time I do not see a fix.

steen · Sun Feb 09, 2014 4:21 pm

Hello Folks!

A new bug was found in RoS6.9 and we had to emegreny rollback RoS6.7 all our email had stopped working for 24 hours due to it.

Mangle and Routing BUG!
We have complex mangle and routing rules for email that stopped to work fully after years flawless operation!

Mangle rules, chain: prerouting source address: scmgateway protocol: TCP source port: 25, action mark route mailnat
Mangle rules, chain: prerouting source address: scmgateway protocol: TCP destination port: 25, action mark route mailnat

Ip Route dest 0.0.0.0/0 gateway: asa5510 Routing mark: mailnat

The mikrotik router has two ISP:s as default gateway with classic mikrotik failover, but it is not default gateway for scmgateway, routing is tagged for that device out over the asa5510.

Network analysis show that traffic never leaves the mikrotik device, and telnetting to port 25 from the scmgateway hangs forever.

Rollbak to RoS6.7 solved the problem direcly.

The device is in full production for thousands of people so no supout and further analysis is possible, this is a severe incident, email incidents like this gives calls out 24 / 7 with responce time 1 hour! I will now have to sit writing a report to upper management on what have happened.

timberwolf · Sun Feb 09, 2014 5:44 pm

The device is in full production for thousands of people so no supout and further analysis is possible, this is a severe incident, email incidents like this gives calls out 24 / 7 with responce time 1 hour! I will now have to sit writing a report to upper management on what have happened.

I really don't understand dangerfreaks like you, that put new releases directly into production without any testing. And upper management ist right in requesting a report on such unprofessional behaviour.
Also, route mangle bugs have already been reported in this thread, so by reading it you could have avoided this sitatuation even without testing.

Bongo · Sun Feb 09, 2014 9:44 pm

Hi
can you add in new release v7 support for pppoe dialout adsl/vdsl static ip connection L2tp

Server IP: 10.10.10.10/32

Username: Your DSL Username

Password: Your DSL Password

Authentication Type: PAP/CHAP/MSCHAP

Tunnel Authentication: Yes/No

Secret: 123456

Active as Default Route: Yes

IPSec: Yes/No

bysard · Mon Feb 10, 2014 8:17 am

v6.9 and still in beta state. Good job Mikrotik, i'm really happy to announce we lost all confidence in you and your product. Downgraded all to 5.26. What a mess... AGAIN and AGAIN and AGAIN,...

onnoossendrijver · Mon Feb 10, 2014 9:19 am

It is really a shame, but do not overestimate the big ones like Juniper and Cisco.
Most of the releases from Juniper from last year have one or more serious bugs (MX, T and SRX routers).
Cisco about the same..

BTW: I personally have no problems with 6.9

npero · Mon Feb 10, 2014 9:34 am

I also don't have problem with 6.9.
Simple configuration routing mark work, OpenVPN server, PPTP server work, PPPOE server with radius work.
No problem for now, all network now running 6.9, x86, sxlite, groove, Rb433, Rb433AH, RB750, RB711.

In upgrade process lost one x86 after upgrade log is full some IPSEC error but on this machine is simple setup with static routing and OpenVPn server not use IPSEC, restart CPU go 100% and broke.
Reboot say same about unable to uncompress and don't boot, re install for iso image to 6.4 and reconfiguration upgrade to 6.9 and work again.

For now 6.9 is ok on my network.

mrz · Mon Feb 10, 2014 11:23 am

Hello Folks!

A new bug was found in RoS6.9 and we had to emegreny rollback RoS6.7 all our email had stopped working for 24 hours due to it.

Mangle and Routing BUG!
We have complex mangle and routing rules for email that stopped to work fully after years flawless operation!

Mangle rules, chain: prerouting source address: scmgateway protocol: TCP source port: 25, action mark route mailnat
Mangle rules, chain: prerouting source address: scmgateway protocol: TCP destination port: 25, action mark route mailnat

Ip Route dest 0.0.0.0/0 gateway: asa5510 Routing mark: mailnat

The mikrotik router has two ISP:s as default gateway with classic mikrotik failover, but it is not default gateway for scmgateway, routing is tagged for that device out over the asa5510.

Network analysis show that traffic never leaves the mikrotik device, and telnetting to port 25 from the scmgateway hangs forever.

Rollbak to RoS6.7 solved the problem direcly.

The device is in full production for thousands of people so no supout and further analysis is possible, this is a severe incident, email incidents like this gives calls out 24 / 7 with responce time 1 hour! I will now have to sit writing a report to upper management on what have happened.

Send a supout file to support.

steen · Mon Feb 10, 2014 1:18 pm

Hello Folks!

A new bug was found in RoS6.9 and we had to emegreny rollback RoS6.7 all our email had stopped working for 24 hours due to it.

Mangle and Routing BUG!
We have complex mangle and routing rules for email that stopped to work fully after years flawless operation!

Mangle rules, chain: prerouting source address: scmgateway protocol: TCP source port: 25, action mark route mailnat
Mangle rules, chain: prerouting source address: scmgateway protocol: TCP destination port: 25, action mark route mailnat

Ip Route dest 0.0.0.0/0 gateway: asa5510 Routing mark: mailnat

The mikrotik router has two ISP:s as default gateway with classic mikrotik failover, but it is not default gateway for scmgateway, routing is tagged for that device out over the asa5510.

Network analysis show that traffic never leaves the mikrotik device, and telnetting to port 25 from the scmgateway hangs forever.

Rollbak to RoS6.7 solved the problem direcly.

The device is in full production for thousands of people so no supout and further analysis is possible, this is a severe incident, email incidents like this gives calls out 24 / 7 with responce time 1 hour! I will now have to sit writing a report to upper management on what have happened.
Send a supout file to support.

That will be a problem, system is in production and I can not upgrade to RoS6.9 again simply for making a supout file, thay will break traffic again and cause call outs. What I will do is to clone the device to another exactly the same routerboard hardware and make a supout file from that.
But such action is more than a week away.

A further reading in this RoS6.9 thread show that many other have same problems, it is caused by tagging traffic and routing by the tag is not respected, traffic goes out the default gateway regardless the tag and routing.

jansat · Mon Feb 10, 2014 2:21 pm

I have installed 9/2/2014 RouterOS v6.9 to my RB2011UiAS-2HnD, but dont have bootloader v3.12 firmware update,
today 10/2/2014 on a second RB2011UAS-2HnD and a rb1100ahx2 and also not the new bootloader/firmware

Jetrider · Mon Feb 10, 2014 3:14 pm

That will be a problem, system is in production and I can not upgrade to RoS6.9 again simply for making a supout file, thay will break traffic again and cause call outs. What I will do is to clone the device to another exactly the same routerboard hardware and make a supout file from that.
But such action is more than a week away.

Welcome to the club, man !
Now, I would gladly hear someone explain me, why exactly should Steen and Jetrider clone device (or disturb customers interwebz with yet another downtimes on current device) and fap around, if we can provide them those supouts and configs and MT can themselves replicate the situation with "another exactly the same routerboard hardware" ? Dont they have testlabs better suited for the job than our own workshops and datacenters ?

Mon Feb 10, 2014 3:18 pm

In the 6.9 release, there is a "updated bootloader firmware". But I didn't see any new version then the current 3.10.

Only some RB models have new firmware, not all.