v6.19 released

Wed Aug 27, 2014 12:22 pm

What's new in 6.19 (2014-Aug-26 14:05):

*) wireless - improvements for nv2 and 802.11ac
*) sstp - make sstp work on i386 as well;
*) ippool - improve performance when acquiring address without preference;
*) partitions - copying partitions did not work on some boards;
*) bridge - added "Auto Isolate" stp enhancement (802.1q-2011, 13.25.6)
*) ipsec - when peer config is changed kill only relevant SAs;
*) vpls - do not abort BGP connection when receiving invalid 12 byte
nexthop encoding;
*) dns-update - fix zone update;
*) dhcpv4 server - support multiple radius address lists;
*) console - added unary operator 'any' that evaluates to true if argument
is not null or nothing value;
*) CCR - improved performance;
*) firewall - packet defragmenting will only happen with connection tracking enabled;
*) firewall - optimized option matching order with-in a rule;
*) firewall - rules that require CONNTRACK to work will now have Invalid flag
when CONNTRACK is disabled;
*) firewall - rules that require use-ip-firewall to work will now have invalid flag
when use-ip-firewall is disabled;
*) firewall - rules that have interface with "Slave" flag specified as in-/out-interface
will now have Invalid flag;
*) firewall - rules that have interface without "Slave" flag specified as in-/out-bridge-port
will now have Invalid flag;
*) firewall - rules with Invalid flags will now be auto-commented to explain why;
*) l2tp - force l2tp to not use MPPE encryption if IPsec is used;
*) sstp - force sstp to not use MPPE encryption (it already has TLS one);
*) sstp - make it work for x86 systems
*) winbox - added dual PSU stats in health menu
*) ipv6 - Gre6 can now correctly fragment large packets
*) simple queue performance optimisation/improvement for multi-core RouterOS devices (especially CCR)

Holekm · Wed Aug 27, 2014 12:28 pm

wireless - improvements for nv2 and 802.11ac
What are the improvments? Please give some more information. It works with fp-wireless only or not?

Wed Aug 27, 2014 12:30 pm

wireless - improvements for nv2 and 802.11ac
What are the improvments? Please give some more information. It works with fp-wireless only or not?

802.11ac only is supported in wireless-fp, so yes, of course only for this package.

WirelessRudy · Wed Aug 27, 2014 12:39 pm

What's new in 6.19 (2014-Aug-26 14:05):

*) wireless - improvements for nv2 and 802.11ac

Does the "and" means the improvements are for NV2 working on a 'ac' radio? Or does it mean improvements are for nv2 systems (a/b/c/g/n/ac) AND also for 802.11ac working radio link?
In other words, do these improvements have effect on nv2 working on 'n' protocol?

Wed Aug 27, 2014 12:40 pm

What's new in 6.19 (2014-Aug-26 14:05):

*) wireless - improvements for nv2 and 802.11ac
Does the "and" means the improvements are for NV2 working on a 'ac' radio? Or does it mean improvements are for nv2 systems (a/b/c/g/n/ac) AND also for 802.11ac working radio link?
In other words, do these improvements have effect on nv2 working on 'n' protocol?

both. there are many changes in wireless. improvements for nv2, for ac, and for nv2 on ac

WirelessRudy · Wed Aug 27, 2014 12:44 pm

*) CCR - improved performance;
*) simple queue performance optimisation/improvement for multi-core RouterOS devices (especially CCR)

I'm dying to hear results or comments from the field of users. My CCR is the only device not running 6.18 yet (its on 6.13) and I dare not to change it before I am confident it works...... its my gateway to my connection to the internet and can't afford to loose its functionality... so please anyone let me know how it works on the CCR.

MadEngineer · Wed Aug 27, 2014 1:04 pm

Just updated an RB951-2n and all the firewall filter rules were gone. They all came back on a second restart

fposavec · Wed Aug 27, 2014 1:07 pm

[Ticket#2014082766000509] - SXT ac in station mode rebooted on nstream protocol.
nv2 is ok, bud half oh bandwitch
802.11 also ok, bandwitch cca 75Mbit on 20 MHz channel

infused · Wed Aug 27, 2014 1:21 pm

My CCR is on 6.3.... Could this finally be the release???

Wed Aug 27, 2014 1:27 pm

Just updated an RB951-2n and all the firewall filter rules were gone. They all came back on a second restart

Most likely you just had to wait a while, until the list populated

stmx38 · Wed Aug 27, 2014 1:33 pm

Upgraded RB951G-2HnD - all is ok!

Thank you, Mikrotik Team!

MadEngineer · Wed Aug 27, 2014 1:40 pm

Just updated an RB951-2n and all the firewall filter rules were gone. They all came back on a second restart
Most likely you just had to wait a while, until the list populated

i did give it some time, was looking at what backups i had and even i tested a filter rule. the firewall rules were 100% gone and not applied - the router was open to the world until I reset it again.

rextended · Wed Aug 27, 2014 1:41 pm

I just upgrade my x86 with user-manager...

WHERE IS GONE USER MANAGER AND ALL THE USERS DATABASE????

Why on changelog that is not present?

THE WEB INTERFACE DO NOT APPEAR AND "new" USER-MANAGER IS DISABLED BY DEFAULT AND ALL THE USERS AND PROFILE ARE GONE...

Reverting to 6.18 have solved the problem...
I love backups...

NOw I check on another board what mess you have done with usermanager.

MadEngineer · Wed Aug 27, 2014 1:51 pm

If you check logs after an update it will tell you why the user manager import failed.

efondevila · Wed Aug 27, 2014 1:55 pm

You can download the latest versions from the site, any problem normis?

Wed Aug 27, 2014 1:56 pm

You can download the latest versions from the site, any problem normis?

looks like it works for others. what is the problem? did you refresh the page?

efondevila · Wed Aug 27, 2014 1:57 pm

when I want to download the zip file returns me to the main page of mikrotik

Wed Aug 27, 2014 1:59 pm

when I want to download the zip file returns me to the main page of mikrotik

you are clicking on the wrong thing

here is the download section

Screenshot 2014-08-27 13.58.37.png

rextended · Wed Aug 27, 2014 2:00 pm

If you check logs after an update it will tell you why the user manager import failed.

It also explain why

http://192.168.3.4/userman

no more exist and

ros code

[admin@MikroTik] > tool
[admin@MikroTik] /tool>              
bandwidth-server  e-mail  graphing  mac-server  netwatch  sms  sniffer  traffic-generator  traffic-monitor  user-manager  bandwidth-test  dns-update  export  fetch  flood-ping  ip-scan  mac-scan  mac-telnet  ping-speed  profile  torch  traceroute  wol
[admin@MikroTik] /tool> 
..  /  :
[admin@MikroTik] /tool> user-manager 
[admin@MikroTik] /tool user-manager> 
limitation  profile  profile-limitation  router  session  user  user-counters  user-profile  edit  export  get  print  set
[admin@MikroTik] /tool user-manager> 
..  /  :
[admin@MikroTik] /tool user-manager> set 
accounting-port  authentication-port  enabled
[admin@MikroTik] /tool user-manager> set enabled=
no  yes
[admin@MikroTik] /tool user-manager> set enabled=yes 
[admin@MikroTik] /tool user-manager> 
limitation  profile  profile-limitation  router  session  user  user-counters  user-profile  edit  export  get  print  set
[admin@MikroTik] /tool user-manager> pri
                 enabled: yes
     authentication-port: 1812
         accounting-port: 1813
  global-active-sessions: 0
[admin@MikroTik] /tool user-manager>

database entry no longer exist, user-manager must be activated, etc.???

efondevila · Wed Aug 27, 2014 2:02 pm

when I want to download the zip file returns me to the main page of mikrotik
you are clicking on the wrong thing

here is the download section

Screenshot 2014-08-27 13.58.37.png

no la que no me deja bajar es la version --> all_files_6.19rc.zip

efondevila · Wed Aug 27, 2014 2:03 pm

when I want to download the zip file returns me to the main page of mikrotik
you are clicking on the wrong thing

here is the download section

Screenshot 2014-08-27 13.58.37.png

no it wont let me down is the version -> all_files_6.19rc.zip

Wed Aug 27, 2014 2:03 pm

where do you see such file?
such file doesn't exist.

rextended · Wed Aug 27, 2014 2:05 pm

WARNING: USER-MANAGER IS TOTALLY UNUSABLE ON 6.19 DO NOT UPGRADE OR UPDATE!!!

PLEASE TAKE URGENT MEASURES BEFORE ANYONE PUT THE FILE ON WORKING ENVIRONMENTS AND LOST ALL USER-MANAGER DATABASE.

THANKS.

efondevila · Wed Aug 27, 2014 2:06 pm

where do you see such file?
such file doesn't exist.

when I click on the link me back to the main page of mikrotik!!!

rextended · Wed Aug 27, 2014 2:13 pm

Secondary BUG/ Missing feature:
/interface wireless reset-configuration DO NOT ASK FOR CONFIRMATION ON CLI
(using Winbox and pressing the button ask the confirmation)

Vanta · Wed Aug 27, 2014 2:14 pm

where do you see such file?
such file doesn't exist.

For some reason he tries to download the RC...
http://goo.gl/E6DyVl

when I click on the link me back to the main page of mikrotik!!!

You click on the wrong link

http://goo.gl/tLjuv3

rextended · Wed Aug 27, 2014 2:15 pm

[post deleted from author]

Wed Aug 27, 2014 2:18 pm

I already posted a screenshot where the v6.19 links are located, please see above.

rextended · Wed Aug 27, 2014 2:20 pm

I already posted a screenshot where the v6.19 links are located, please see above.

Tle "old" 6.19rc link left on the page cause confusion.

Wed Aug 27, 2014 2:21 pm

I already posted a screenshot where the v6.19 links are located, please see above.
Tle "old" 6.19rc link left on the page cause confusion.

I know, we will remove it. I already wrote where to download correct versions.

rextended · Wed Aug 27, 2014 2:23 pm

I already posted a screenshot where the v6.19 links are located, please see above.
Tle "old" 6.19rc link left on the page cause confusion.
I know, we will remove it. I already wrote where to download correct versions.

Please,

have you read or not the problem about user-manager and transmitted to the support?

Thanks.

Wed Aug 27, 2014 2:24 pm

Ticket#2014032566001217 (2014-03-25) BUG 6.12: Replicable kernel crash when try to discover why winbox not working well from 6.8 over IP obtained from pppoe-client

this is from the above ticket:

From "xxx" <info@xxx>
To: <support@mikrotik.com>
Subject: Fw: [Ticket#2014032566001217] BUG 6.12: Replicable kernel crash when try to discover why winbox not working well from 6.8 over IP [...]
Created: 04/15/2014 16:11:05

Solved on 6.12 final,
Thanks!

rextended · Wed Aug 27, 2014 2:27 pm

[post deleted from author]

Shiro · Wed Aug 27, 2014 2:30 pm

Updated a CCR1009 to from 6.18 to 6.19, having issues with dnat from WAN Interface (L2TP with static IP) to my Exchange Server.

The issue is that a dnat to port 443 fails to reach the destination host, if something connected to 443 from the wan side, there is no traffic leaving the CCR to the destination host. So far this only affects port 443 (for ActiveSync/OAW), port 25 (for smtp) works.

I played a bit with MTU settings on my L2TP WAN, but nothing helps. Going back to 6.18 solve this problem.

I had the same issue with 6.19rc6 and made a rollback.

Any idea whats causing this?

Update: found the error on my side, it was an old layer 7 blocking rule that made trouble with 6.19, removed it and everything is ok.

Wed Aug 27, 2014 2:52 pm

Ticket#2014032566001217 (2014-03-25) BUG 6.12: Replicable kernel crash when try to discover why winbox not working well from 6.8 over IP obtained from pppoe-client
this is from the above ticket:

From "xxx" <info@xxx>
To: <support@mikrotik.com>
Subject: Fw: [Ticket#2014032566001217] BUG 6.12: Replicable kernel crash when try to discover why winbox not working well from 6.8 over IP [...]
Created: 04/15/2014 16:11:05

Solved on 6.12 final,
Thanks!

The problem happen again after 6.12 "non-final" (simply check all the 6.12->6.18 topic, everytime I repost the bug), simply do the test I have posted on the forum.
If is needed I submit again the same email for open another ticket.

Please remove "from" on the mail...

Why did you not respond to the ticket?

rextended · Wed Aug 27, 2014 3:17 pm

The version I have use to test again that kernel error are the 6.19 build time "Aug/26/2014 11:10:11",
but now I download again 6.19 for security: build time are "Aug/26/2014 14:05:51"

Now whit this version I'm not able to obtain kernel panik (and obviously I'm very happy for that),
but really I'm disoriented because I do not have notice the new "new" version...

Shiro · Wed Aug 27, 2014 3:19 pm

*) winbox - added dual PSU stats in health menu

Still missing PSU stats on CCR1009 in Winbox, see attached screenshot.

/system routerboard print
routerboard: yes
model: CCR1009-8G-1S-1S+
serial-number: 48FF01207EFD
current-firmware: 3.18
upgrade-firmware: 3.18

/system health print
fan-mode: auto
use-fan: main
active-fan: main
cpu-overtemp-check: yes
u-overtemp-threshold: 100C
cpu-overtemp-startup-delay: 1m
voltage: 23.7V
current: 873mA
temperature: 32C
cpu-temperature: 54C
power-consumption: 20.7W
psu1-state: ok
psu2-state: fail
fan1-speed: 5514RPM

Wed Aug 27, 2014 3:19 pm

we only released one version. you probably had the pre-release installed.

rextended · Wed Aug 27, 2014 3:20 pm

we only released one version. you probably had the pre-release installed.

Seeing that I have re-download all the other packages, same build of Aug 26 @ 14:05 but I still have problem with user-manager.

Do not work and do not display any webpage.

What is changed???
What are the new commands for use it?

Thanks.

Wed Aug 27, 2014 3:21 pm

Looks like Usermanager has a bug. Nothing was changed.

Chupaka · Wed Aug 27, 2014 3:23 pm

Normis, since 6.16, there's no CHANGELOG file in torrent - is it expected behaviour? why did you remove it?

rextended · Wed Aug 27, 2014 3:25 pm

Looks like Usermanager has a bug. Nothing was changed.

Finally one reply about that

I'm still happy of RouterOS, no matter how many bug.

Why?

From one version to another I see everytime improvements (6.19 go on both side of my production 23,5Km NV2 wireless-fp PTP and now I have only 26ms latency vs google.it with 40/45Mbit/s of traffic... [at the Interoute optical fibre I have 24ms]).

The bug are made to be solved, RouterOS to be improved

Wed Aug 27, 2014 3:29 pm

Normis, since 6.16, there's no CHANGELOG file in torrent - is it expected behaviour? why did you remove it?

Changelog has moved from file to database, check download page.

Shiro · Wed Aug 27, 2014 3:35 pm

Looks like Usermanager has a bug. Nothing was changed.

Yes something is wrong, tried to setup a new usermanager with 6.19, state is enabled, but database stay at "backup", tried 2 reboots without success, webinterface is not working.

rextended · Wed Aug 27, 2014 3:37 pm

Looks like Usermanager has a bug. Nothing was changed.
Yes something is wrong, tried to setup a new usermanager with 6.19, state is enabled, but database stay at "backup", tried 2 reboots without success, webinterface is not working.

I'm not the only, I'm happy for not do some wrong.

vipnet · Wed Aug 27, 2014 3:38 pm

ThunderCache 7.1

on update the route static MARK in blue color no active
soluction reboot now rb or change MARK

Alex R. Endres
ThunderCache team
skype live:adcd_comprasonline
licence https://under-linux.org/showthread.php?t=164219

Brazil.

Shiro · Wed Aug 27, 2014 3:40 pm

Looks like Usermanager has a bug. Nothing was changed.
Yes something is wrong, tried to setup a new usermanager with 6.19, state is enabled, but database stay at "backup", tried 2 reboots without success, webinterface is not working.
I'm not the only, I'm happy for not do some wrong.
I think is some related to:
*) partitions - copying partitions did not work on some boards;
but on /tool user-manager are missing database and customer fields.

oh yes, that partition thing is now working very well on CCR1009, i was able to split and copy the partition + config, finaly a fix for that

Wed Aug 27, 2014 4:59 pm

Looks like Usermanager has a bug. Nothing was changed.
Yes something is wrong, tried to setup a new usermanager with 6.19, state is enabled, but database stay at "backup", tried 2 reboots without success, webinterface is not working.
I'm not the only, I'm happy for not do some wrong.

We fixed the issue. We had accidentally released the incorrect userman package, please re-download the userman package now.

lelmus · Wed Aug 27, 2014 5:02 pm

For the peeps that have trouble with website still showing 6.18 and the "RouterOS v6.19 release candidate special release" section jumping to the home page. Use the below links to get the updates.

http://download2.mikrotik.com/routeros/ ... e-6.19.zip

http://download2.mikrotik.com/routeros/ ... mipsbe.npk

rextended · Wed Aug 27, 2014 5:49 pm

I'm not the only, I'm happy for not do some wrong.
We fixed the issue. We had accidentally released the incorrect userman package, please re-download the userman package now.

I have upload now the user-manager package from site and from downoad from auto upgrade...

The file still 80kb, the regular user manager is over 700Kb.

User-manager still broken.

savage · Wed Aug 27, 2014 6:06 pm

We fixed the issue. We had accidentally released the incorrect userman package, please re-download the userman package now.

So changes was made to userman that's not in the changelog? What else was changed that isn't documented -sigh-

Each and every single release, it's the same thing over and over... It's actually becoming laughable now.

Chupaka · Wed Aug 27, 2014 6:27 pm

The file still 80kb, the regular user manager is over 700Kb.

at least in the current torrent all user-manager's files are more than 700k =)

ndbjorne · Wed Aug 27, 2014 6:32 pm

2014-08-27T170630.png

Upgraded a RB951G-2HnD from 6.17 to 6.19.
It seemed ok until I shut it down: after a while it booted up and on log I found (see attached image).
I unplugged power, then powered it again and now

ros code

/system shutdown

works.

Bye

Chupaka · Wed Aug 27, 2014 6:37 pm

Normis, why are there two Netinstall files in torrent (netinstall-6.19.zip and netinstall-6.19-tile.zip) if they are identical? moreover, download link in 'Tile' section points to netinstall-6.19.zip

wispwest · Wed Aug 27, 2014 7:00 pm

Well this is a tough one. I have 2 PtP links, one is 30 and the other is 40 miles. They've always been fine with NV2, however I notice I get more 'real' throughput when using Nstreme now (used to be the opposite). So I switch to Nstreme and my speedtests on laptop at the edge of the links get 70/30 (compared to 50/20 with Nv2). So I think all is well. Then starting at 11:30PM all the way till this morning, there are random disconnects for 'no apparent reason'. When it works, it works great! Then out of no where, station drops registration for a few seconds, then comes back.

So.. back to NV2 for now. Wish Nstreme would be more reliable like it used to be.

CyberTod · Wed Aug 27, 2014 8:11 pm

+1 for nstreme. Do you still work on it ? On latest versions only changes for nv2 are mentioned. Nstreme needs only 1 improvement - no disconnections.
Maybe too soon to say, havent done testing, but my smokeping graphs show less packet loss in nv2 today.

Bigfoot · Wed Aug 27, 2014 8:37 pm

Userman Problem started in 6.19rc6 !!!
Can I have the correct files please !!!

Stillhard · Wed Aug 27, 2014 9:13 pm

It's from rc5, check my report here http://forum.mikrotik.com/viewtopic.php ... 24#p441524

poizzon · Wed Aug 27, 2014 10:01 pm

21:41:30 ipsec,error phase1 negotiation failed due to time up. 01d4db047a60b9bc:0000000000000000
21:42:30 ipsec,error phase1 negotiation failed due to time up. 77ef7ebbe6072755:0000000000000000
21:43:30 ipsec,error phase1 negotiation failed due to time up. 84b2038687046a7f:0000000000000000
21:44:30 ipsec,error phase1 negotiation failed due to time up. 42deffb07e877772:0000000000000000
21:45:30 ipsec,error phase1 negotiation failed due to time up. 51399cc76d9cf438:0000000000000000
21:46:30 ipsec,error phase1 negotiation failed due to time up. 06bded99cac5be7f:0000000000000000
21:47:40 ipsec,error phase1 negotiation failed due to time up. 89a97510102ce77c:0000000000000000
21:48:40 ipsec,error phase1 negotiation failed due to time up. 3f89e001a721cbab:0000000000000000
21:49:50 ipsec,error phase1 negotiation failed due to time up. 218823fc05c11ed0:0000000000000000
21:50:50 ipsec,error phase1 negotiation failed due to time up. b6863d33d03abd58:0000000000000000
21:51:50 ipsec,error phase1 negotiation failed due to time up. 56afaac137289cf2:0000000000000000
21:52:50 ipsec,error phase1 negotiation failed due to time up. 3ba12327f25ba182:0000000000000000
21:53:50 ipsec,error phase1 negotiation failed due to time up. 9772820fef218c03:0000000000000000
21:54:50 ipsec,error phase1 negotiation failed due to time up. e0f9cea084fc72f2:0000000000000000
21:55:50 ipsec,error phase1 negotiation failed due to time up. f699872d6c861076:0000000000000000
21:56:50 ipsec,error phase1 negotiation failed due to time up. d14bf652f5792b1d:0000000000000000
21:57:50 ipsec,error phase1 negotiation failed due to time up. 29f50f1bcfb494ed:0000000000000000
21:58:50 ipsec,error phase1 negotiation failed due to time up. 0d818de75296c237:0000000000000000

RafGan · Thu Aug 28, 2014 1:35 am

We fixed the issue. We had accidentally released the incorrect userman package, please re-download the userman package now.

Still waiting for this file... Almost 12 hours is not equel to now... It is still not fixed...

dayswaste · Thu Aug 28, 2014 2:40 am

Upgraded RB2011UAS-2HnD via "Check for Updates" in package list remotely. Had me worried but after +15mins it came back online.

jayd2k · Thu Aug 28, 2014 7:30 am

Back on July 5th I was reporting the following issue with MT support which, as of today, still persists.

Can't assign any bandwidth greater than 4294Mbps (32 bit) to a 10GbE interface:

ros code

[admin@DCCCR01.RT] /mpls traffic-eng interface> print 
Flags: X - disabled, I - invalid 
 #   INTERFACE                  BANDWIDTH  TE-METRIC REMAIN...
 0   sfp-sfpplus1                4294Mbps          1   4.2Gbps
 1   sfp-sfpplus2                4295Mbps          1  32.7kbps

[admin@DCCCR01.RT] /mpls traffic-eng interface> print detail 
Flags: X - disabled, I - invalid 
 0   interface=sfp-sfpplus1 bandwidth=4294Mbps k-factor=3 resource-class=0 
     refresh-time=30s use-udp=no blockade-k-factor=3 te-metric=1 
     igp-flood-period=3m 
     up-flood-thresholds=15,30,45,60,75,80,85,90,95,97,98,99,100 
     down-flood-thresholds=15,30,45,60,75,80,85,90,95,97,98,99,100 
     remaining-bw=4.2Gbps 

 1   interface=sfp-sfpplus2 bandwidth=4295Mbps k-factor=3 resource-class=0 
     refresh-time=30s use-udp=no blockade-k-factor=3 te-metric=1 
     igp-flood-period=3m 
     up-flood-thresholds=15,30,45,60,75,80,85,90,95,97,98,99,100 
     down-flood-thresholds=15,30,45,60,75,80,85,90,95,97,98,99,100 
     remaining-bw=32.7kbps

This keeps us from fully utilizing our 10GbE environment with MPLS on-top. Almost two months and even though I was following up, still this wasn't fixed. That's like offering a Porsche w/o a steering wheel. Know what I mean?

This was tested on four CCR1036-8G-2S+ with ROS 6.19 installed. Since this seems to be a software bug it supposedly effects ALL 10GbE enabled boards.

Kindly fix this ...

Alessio Garavano · Thu Aug 28, 2014 8:21 am

Download from torrent only the usermanager package for ppc and upgraded succesfully, now work again!

PS: need to set "uninstall" the previous package and reboot with the new package in the root folder of Files and ready!

Thu Aug 28, 2014 9:17 am

We fixed the issue. We had accidentally released the incorrect userman package, please re-download the userman package now.
So changes was made to userman that's not in the changelog? What else was changed that isn't documented -sigh-

Each and every single release, it's the same thing over and over... It's actually becoming laughable now.

No. the user manager package that was included, was broken. We re-released the version with a working user-manager package. No changes were made.

Thu Aug 28, 2014 9:18 am

The file still 80kb, the regular user manager is over 700Kb.
at least in the current torrent all user-manager's files are more than 700k =)

Could be browser cache. Rextended, check if you don't have a proxy or browser cache.

bds1904 · Thu Aug 28, 2014 9:30 am

upgrading to 6.19 (from 6.18)broke the config on my x86 box. It broke the mangle rules (i use them for PCC load balancing) and the route rules. I confirmed the problem twice by downgrading, restoring from a backup config and upgrading again. The config remained broken even after downgrade.

I was able to backup, upgrade to 6.19 then restore the config, and it works just fine.

rextended · Thu Aug 28, 2014 9:48 am

The file still 80kb, the regular user manager is over 700Kb.
at least in the current torrent all user-manager's files are more than 700k =)
Could be browser cache. Rextended, check if you don't have a proxy or browser cache.

Try yourself, just re-downloaded:

This is the official download link on download page:

http://download2.mikrotik.com/routeros/ ... 6-6.19.zip

user-manager is only 82.001 byte Aug/27/2014 09:52

Also the auto upgrade inside the routerboard download only 80.1Kbyte file

I have download also all_packages-ppc-6.19.zip and all_packages-mipsbe-6.19.zip, never downloaded before,
and the file is 73k and 65k instead the usual 800KB.

rextended · Thu Aug 28, 2014 9:59 am

I still see on download page "RouterOS v6.19 release candidate special release"

If you use multiple servers, the IP of http://www.mikrotik.com for my DNS is 159.148.147.196

Are you s ure the webisite is updated with new files?

I do not know if torrent containing right files, but auto-upgrade and webserver still release wrong user-manager packages.

rextended · Thu Aug 28, 2014 10:06 am

I'M INVESTIGATING ON TWO MORE MEMORY AND KERNEL ERRORS:

IS not easy write it on two word, I'm preparing the adequate description, but for now:

1) Kernel Crash or Panic when using MPPE encryption
pppoe-server [default-encryption settings for 6.19] <-> EoIP [default settings for 6.19] <=== cable ===> EoIP [default settings for 6.19] <-> pppoe-client [default-encryption settings for 6.19]

2) Out of memory on both pppoe-server and pppoe-client machine when traffic is passed over EoIP.
Traffic decrease memory available and the memory is never released:
pppoe-server [default settings for 6.19] <-> EoIP [default settings for 6.19] <=== cable ===> EoIP [default settings for 6.19] <-> pppoe-client [default settings for 6.19]

Hugok1 · Thu Aug 28, 2014 10:30 am

I have upgraded 4 of our 67 Towers rb433AH routers from 6.18 to 6.19, however on 2 of the four the wireless settings was lost so we had to drive to reconfigure the boards. On the other two boards the settings was not lost... Same boards, same radios, plz plz plz Mikrotik, do NOT release something as a commercial release when it was NOT properly tested.

I am also awaiting a answer on a support ticket in regards to OSPF randomly locking up in version 6.18.

Hope to hear from you....

nkourtzis · Thu Aug 28, 2014 10:31 am

Try yourself, just re-downloaded:

This is the official download link on download page:

http://download2.mikrotik.com/routeros/ ... 6-6.19.zip

user-manager is only 82.001 byte Aug/27/2014 09:52

Also the auto upgrade inside the routerboard download only 80.1Kbyte file

I have download also all_packages-ppc-6.19.zip and all_packages-mipsbe-6.19.zip, never downloaded before,
and the file is 73k and 65k instead the usual 800KB.

Same here. Cleared cache, nothing changed. Package downloaded from torrent is ok.

Rudios · Thu Aug 28, 2014 10:32 am

I still see on download page "RouterOS v6.19 release candidate special release"

If you use multiple servers, the IP of http://www.mikrotik.com for my DNS is 159.148.147.196

Are you s ure the webisite is updated with new files?

I do not know if torrent containing right files, but auto-upgrade and webserver still release wrong user-manager packages.

Are you using a proxy? I have tried to download the newer version from two different computers, each connected differently to the internet.
One is directly connected and I get a newer (bigger) version of the zip. But the other, connected via proxy (unmanaged by myself) is still downloading smaller package.

Thu Aug 28, 2014 10:34 am

This is what I see in the server:

Screenshot 2014-08-28 10.33.59.png

and from the ZIP:

Screenshot 2014-08-28 10.36.57.png

rextended · Thu Aug 28, 2014 10:36 am

This is what I see in the server:

Screenshot 2014-08-28 10.33.59.png

And inside the ZIP on the server?

Are the same server used for auto-update inside the routerboard?

Using this link (not present on any official download page)

http://download2.mikrotik.com/routeros/ ... r-6.19.npk

Finally I got the right file, but other users know that only if go to this forum...

Thu Aug 28, 2014 10:37 am

This is what I see in the server:

Screenshot 2014-08-28 10.33.59.png
And inside the ZIP on the server?

Are the same server used for auto-update inside the routerboard?

yes.

rextended · Thu Aug 28, 2014 10:47 am

The server download2.mikrotik.com are solved as d355q2xs8kb5oj.cloudfront.net 54.230.79.3 and single user-manager package are ok, but on the zip on same server are wrong.

The server used from auto-update funcion from the routerboard is 54.230.77.67 and everytime obtain corrupted user-manager-6.19.npk of 80KB

Now I clear all DNS server cache, and retry.

Thu Aug 28, 2014 10:48 am

The server download2.mikrotik.com are solved as d355q2xs8kb5oj.cloudfront.net 54.230.79.3 and single user-manager package are ok, but on the zip on same server are wrong.

The server used from auto-update funcion from the routerboard is 54.230.77.67 and everytime obtain corrupted user-manager-6.19.npk of 80KB

they use DNS load balancing, you will get different IP each time. single package is the same as in auto update.

rextended · Thu Aug 28, 2014 10:53 am

The server download2.mikrotik.com are solved as d355q2xs8kb5oj.cloudfront.net 54.230.79.3 and single user-manager package are ok, but on the zip on same server are wrong.

The server used from auto-update funcion from the routerboard is 54.230.77.67 and everytime obtain corrupted user-manager-6.19.npk of 80KB
they use DNS load balancing, you will get different IP each time. single package is the same as in auto update.

Yes, yes, I know, but seems one server with that address >>54.230.77.67<< still not updated, I have try another server on cloudfront for the auto-update and now has the right file...

Thu Aug 28, 2014 10:54 am

The server download2.mikrotik.com are solved as d355q2xs8kb5oj.cloudfront.net 54.230.79.3 and single user-manager package are ok, but on the zip on same server are wrong.

The server used from auto-update funcion from the routerboard is 54.230.77.67 and everytime obtain corrupted user-manager-6.19.npk of 80KB
they use DNS load balancing, you will get different IP each time. single package is the same as in auto update.

Yes, yes, I know, but seems one server with that address >>54.230.77.67<< still not updated, I have try another server on cloudfront for the auto-update and now has the right file...

It could take a while until all AWS Cloudfront servers receive the "invalidation request".

dohmniq · Thu Aug 28, 2014 10:55 am

You could try adding a query string after the URL to cause a cache miss:

http://download2.mikrotik.com/routeros/ ... stion-mark

I just downloaded the zip and see this:

876625 Stored 876625 0% 08-27-14 14:28 00000000 user-manager-6.19.npk

~850KB

reshsv · Thu Aug 28, 2014 10:58 am

Hello.
We upgrade our CCR to 6.19 and get "kernel failure" and "reboot by watchdog timer"
I can send autosupout.rif file if needed.

Now we rollback to 6.14...

rextended · Thu Aug 28, 2014 10:59 am

You could try adding a query string after the URL to cause a cache miss:

http://download2.mikrotik.com/routeros/ ... stion-mark

I just downloaded the zip and see this:

876625 Stored 876625 0% 08-27-14 14:28 00000000 user-manager-6.19.npk

~850KB

Again: i do not use cache or proxy. Or you intend Cloudfront cache?

rextended · Thu Aug 28, 2014 11:00 am

Hello.
We upgrade our CCR to 6.19 and get "kernel failure" and "reboot by watchdog timer"
I can send autosupout.rif file if needed.

Now we rollback to 6.14...

You use EoIP tunnel?

dohmniq · Thu Aug 28, 2014 11:07 am

You could try adding a query string after the URL to cause a cache miss
Again: i do not use cache or proxy. Or you intend Cloudfront cache?

ANY cache - including transparent proxy caches run by your connectivity provider, Cloudfront transparent reverse proxies, browser cache, etc.
It's just a suggestion and could help other people.

Thu Aug 28, 2014 11:08 am

We have repeatedly issued invalidation request for these files, it should be ok in a few minutes

zervan · Thu Aug 28, 2014 11:08 am

This is insane! I've updated RouterOS to 6.19 last evening in multiple organizations and today they are calling me that WiFi authentication is not working, since I'm using userman

How can you offer "stable" version with such a huge mistake?

I really need to download PPC version of fixed package ASAP, presented links are not working, there is still the old broken one, I tried http://download2.mikrotik.com/routeros/ ... sgdfgwgwer - contains broken package. I don't use any cache.

rextended · Thu Aug 28, 2014 11:09 am

Clearing DNS cache solve all problem for the "Check for Updates" feature on RouterBOARD, and also on the webiste.

For be explicit for other users: Master DNS cache, gateway DNS cache, office DNS router cache and PC DNS cache.

For the device than have acquired non-wirking IP, must be rebooted, after that when resolving again the DNS give good IP.

rextended · Thu Aug 28, 2014 11:14 am

This is insane! I've updated RouterOS to 6.19 last evening in multiple organizations and today they are calling me that WiFi authentication is not working, since I'm using userman How can you offer "stable" version with such a huge mistake?

I really need to download PPC version of fixed package ASAP, presented links are not working, there is still the old broken one, I tried http://download2.mikrotik.com/routeros/ ... sgdfgwgwer - contains broken package. I don't use any cache.

INSANE IS USE ONE "JUST OUT" VERSION TO UPGRADE MULTIPLE DISTANT DEVICE ON PRODUCTION ENVIRONMENT!!!

Not only MikroTik, but also Cam... Cis... and Ubi... haved the same problem on some release...

zervan · Thu Aug 28, 2014 11:20 am

INSANE IS USE ONE "JUST OUT" VERSION TO UPGRADE MULTIPLE DISTANT DEVICE ON PRODUCTION ENVIRONMENT!!!

Not only MikroTik, but also Cam... Cis... and Ubi... haved the same problem on some release...

I thought that since there are public RC versions to download, each "stable" version is tested enough. Now I know that I was wrong. Why I upgraded so early? Because today is the last working day of public holiday here and there is only small part of users in those organizations working. Next week it would be worse.

But this debate will not help, could somebody send me a link to working Userman package?
EDIT: It seems that the last torrent file is containing bigger working package, I am downloading...

zervan · Thu Aug 28, 2014 11:31 am

EDIT: It seems that the last torrent file is containing bigger working package, I am downloading...

Yes, it is containing big file, but now I don't know how RouterOS force to update this package. I tried to save it to router and reboot, it is not working.

EDIT: I have found hint from Alessio Garavano: need to set "uninstall" the previous package and reboot with the new package in the root folder of Files and ready!

chaf84 · Thu Aug 28, 2014 11:33 am

EDIT: It seems that the last torrent file is containing bigger working package, I am downloading...
Yes, it is containing big file, but now I don't know how RouterOS force to update this package. I tried to save it to router and reboot, it is not working.

Uninstall the wrong package then install the new one.

Thu Aug 28, 2014 11:34 am

downgrade to v6.18, then upgrade to correct v6.19

zervan · Thu Aug 28, 2014 11:36 am

Uninstall the wrong package then install the new one.

I was worrying that I will loose configuration and data, but it is working: I've set this package to uninstall, saved new one to router and than just one reboot - old package was uninstalled and the new one installed, it is working again.

rextended · Thu Aug 28, 2014 11:53 am

@zervan: If you read the forum...

I just upgrade my x86 with user-manager...

WHERE IS GONE USER MANAGER AND ALL THE USERS DATABASE????

Why on changelog that is not present?

THE WEB INTERFACE DO NOT APPEAR AND "new" USER-MANAGER IS DISABLED BY DEFAULT AND ALL THE USERS AND PROFILE ARE GONE...

Reverting to 6.18 have solved the problem...
I love backups...

NOw I check on another board what mess you have done with usermanager.

zervan · Thu Aug 28, 2014 12:03 pm

@zervan: If you read the forum...

It is the question of trust. If I trust Mikrotik (doing well working), I will not hesitate to upgrade and I will not check each update on forum if somebody had worse luck before me. Last few months I didn't experienced any problems, so I had trusted.

reshsv · Thu Aug 28, 2014 12:26 pm

Hello.
We upgrade our CCR to 6.19 and get "kernel failure" and "reboot by watchdog timer"
I can send autosupout.rif file if needed.

Now we rollback to 6.14...
You use EoIP tunnel?

Yes, we use EoIP...
Strange thing... also we have same problem on RB951Ui-2HnD with Eoip tunnel, with software 6.15, rollback to 6.14 don't solve problem, rollback to 6.13 - solve.

rextended · Thu Aug 28, 2014 12:39 pm

Hello.
We upgrade our CCR to 6.19 and get "kernel failure" and "reboot by watchdog timer"
I can send autosupout.rif file if needed.

Now we rollback to 6.14...
You use EoIP tunnel?
Yes, we use EoIP...
Strange thing... also we have same problem on RB951Ui-2HnD with Eoip tunnel, with software 6.15, rollback to 6.14 don't solve problem, rollback to 6.13 - solve.

It is why I ask you about EoIP, I'm expecting you have it because this:
http://forum.mikrotik.com/viewtopic.php ... 44#p444591

On the last 6.19 version I see memory problem with EoIP tunnel, oversized (> 14xx) UDP packet go to router memory and never be discarded, causing the router go out of memory...

I must write the correct procedure and submit it to support, the problem are easly REPLICABLE....

reshsv · Thu Aug 28, 2014 1:21 pm

It is why I ask you about EoIP, I'm expecting you have it because this:
http://forum.mikrotik.com/viewtopic.php ... 44#p444591

On the last 6.19 version I see memory problem with EoIP tunnel, oversized (> 14xx) UDP packet go to router memory and never be discarded, causing the router go out of memory...

I must write the correct procedure and submit it to support, the problem are easly REPLICABLE....

Looks correct...
on RB951Ui-2HnD BOX we upgrade to 6.19 and shut down EoIP tunnel, and box works fine...
One of our CCR with 6.18 and EoIP works fine, but RB951Ui-2HnD with 6.14 reloading...
Which releases affected this problem?

rextended · Thu Aug 28, 2014 1:29 pm

It is why I ask you about EoIP, I'm expecting you have it because this:
http://forum.mikrotik.com/viewtopic.php ... 44#p444591

On the last 6.19 version I see memory problem with EoIP tunnel, oversized (> 14xx) UDP packet go to router memory and never be discarded, causing the router go out of memory...

I must write the correct procedure and submit it to support, the problem are easly REPLICABLE....
Looks correct...
on RB951Ui-2HnD BOX we upgrade to 6.19 and shut down EoIP tunnel, and box works fine...
One of our CCR with 6.18 and EoIP works fine, but RB951Ui-2HnD with 6.14 reloading...
Which releases affected this problem?

Surely 6.19, i do not use 6.8 till 6.16.
I use 6.7 on 4 RB1100AHx2 because MPPE bug still exist, and on dozen of machines on production environment 6.17 and 6.18 and no one have this EoIP problem.

StubArea51 · Thu Aug 28, 2014 8:19 pm

Thanks MikroTik...

Running smoothly on a 2011 series..will test different CCRs in the next few days.

MadEngineer · Fri Aug 29, 2014 1:54 pm

Just updated an RB951-2n and all the firewall filter rules were gone. They all came back on a second restart
Most likely you just had to wait a while, until the list populated
i did give it some time, was looking at what backups i had and even i tested a filter rule. the firewall rules were 100% gone and not applied - the router was open to the world until I reset it again.

Checking again I've found the firewall rules had been re-arranged somewhat. Some of them seem to have been re-added in order that I last added them. For example I've always had drop everything else rules on forward and input as the last rules, but these moved up ahead of some rules I had added recently. A jump rule has moved below the whole chain it's supposed to jump to, a couple of accepts that had the same comment were next to each other but they're now split, a couple of log rules that used to be prior to their matching drop rules have also moved around.

MadEngineer · Fri Aug 29, 2014 2:11 pm

Bit of an odd one: begin creating a drop rule. Before saving it with OK or Apply, give it a comment and click OK to close the comment entry. You'll see it immediately creates an Accept rule. It correctly changes after hitting OK or Apply to close the already created rule.

Fri Aug 29, 2014 2:47 pm

Bit of an odd one: begin creating a drop rule. Before saving it with OK or Apply, give it a comment and click OK to close the comment entry. You'll see it immediately creates an Accept rule. It correctly changes after hitting OK or Apply to close the already created rule.

It has always been like this. (Well, at least since 4.something, when I first learned about RouterOS). It has always been creating a rule upon comment addition, even though the rule might have not showed up immediately in the list in previous ROS releases.

Alessio Garavano · Fri Aug 29, 2014 4:20 pm

EDIT: It seems that the last torrent file is containing bigger working package, I am downloading...
Yes, it is containing big file, but now I don't know how RouterOS force to update this package. I tried to save it to router and reboot, it is not working.

EDIT: I have found hint from Alessio Garavano: need to set "uninstall" the previous package and reboot with the new package in the root folder of Files and ready!

Good! +1 karma will be appreciated

diablonet · Fri Aug 29, 2014 6:31 pm

Hello, I own a RB2011 UiAS-2HnD and I had troubles with this update.

I had version 6.18, and applied this into it: /system ntp client set enabled=yes primary-ntp=132.248.30.3 secondary-ntp=132.248.3.73 server-dns-names="0.pool.ntp.org"

When I update it to 6.19, it reboots itself in an endless loop, then downgrade it to 6.18.

I removed the NTP setting: server-dns-names="0.pool.ntp.org", then applied the update 6.19 and then it worked.

Tried a few times and I was able to replicate this issue always. Before update remove that server-dns-names setting if you applied, please fix this Mikrotik

igb · Fri Aug 29, 2014 7:13 pm

I've just tried an upgrade on a 2011-UAS from 6.18 to 6.19. Various devices are connected to the switch ports, some of them using VLAN trunking; three VLANS are present. The hardware ethernet switch is not being used, everything is going via the CPU. There is a PPPoE connection to the Internet. The router comes back up, but many websites fail to load (load some initial content then hang). I had users to contend with, so backed out to 6.18 without much investigation.

Has anyone had similar problems? When I have the network to myself I will try disabling the firewall, as that seems to be an area of substantial change in this realise: I have firewall rules, both for IPv4 and IPv6, and my initial impression is that the websites that were failing to complete were all v4 (Google, for example, worked fine, and I normally access Google over IPv6).

samotoka · Fri Aug 29, 2014 9:05 pm

Hello, I own a RB2011UAS-2HnD-IN and I had troubles with this update too.

I had version 6.18 and after update my router become irresponsible , reset from lcd screen and router load some default config whit 192.168.88.1 but there is not connection to router even winbox , after 3-20 attempts i realized that default config starts wi-fi and via wireless connection and winbox i reset router with no default config , but router start continuous loop. Next step - netinstall , but after netinstall even i delete some config sets on netinslall app - after netinstal router again bacome iresponsibe via 192.168.88.1

Suggestions
1. netinstall app to have big button about netinstall with NO default config
2. LCD touch screen to have menu reset with NO default config

After two days 6.19 working well and smoothly , pptp is OK

AlexS · Sat Aug 30, 2014 9:57 am

Netflow is patchy

reboot -> works
reboot -> doesn't work

its a bit hit and miss weather it works, stoping and starting it doesnt seem to help either ..

rather weird

But I have a supposition it has some thing to do with my a sync routing, its not starting records unless it seems a syn packet ... i think !

A

Sat Aug 30, 2014 10:09 am

Netflow is patchy

reboot -> works
reboot -> doesn't work

But I have a supposition it has some thing to do with my a sync routing, its not starting records unless it seems a syn packet ... i think !

What netflow version do you use? V9 is supposed to report flow stats on a per-connection basis.

rjickity · Sat Aug 30, 2014 12:17 pm

i've just noticed new vlan options in the wireless interface:

/interface wireless vlan-id and vlan-mode

quite handy, just out of curiosity does it use the atheros chip or is this something the CPU does (a.k.a "/interface vlan") and will this affect card -> client tagging as well as card -> CPU ?

edit: i just tested it an it looks like its tagging both when activated. A suggestion would be, if there is no-tag specified than remove the vlan-id option from being set. As the option seems irrelevant without a vlan database. Unless it is used with capsman somehow i guess.

either way, an updated wiki would be great for this

JanezFord · Sun Aug 31, 2014 3:30 am

Upgrade from 6.18 to 6.19 erased all of my static dhcp leases. This happened on my crs125 and rb2011uas. The same thing happened on both units when I upgraded from 6.17 to 6.18. I did not have this issue with previous upgrades.

JF.

Zorro · Mon Sep 01, 2014 7:15 am

works fine. but 6.14 upgrade take almost eight/ten(not use clocks this time) minutes of looking at "booting" screen after reboot.
which slightly terrified mine(does MT employ some kind of media-check or other maintenance or anything else fail-safe related, that would be so MUCH time-consuming in recent verions ?), before it finally completed.

janisk · Mon Sep 01, 2014 11:58 am

Netflow is patchy

reboot -> works
reboot -> doesn't work

But I have a supposition it has some thing to do with my a sync routing, its not starting records unless it seems a syn packet ... i think !
What netflow version do you use? V9 is supposed to report flow stats on a per-connection basis.

definitely need more data on this.

roadracer96 · Mon Sep 01, 2014 6:33 pm

6.19 upgrade definitely plays havoc on configs. I have a very simple config on my home router (CCR1016-12G). 3x Capsman APS, one bridge as LAN interface, one wan interface, NAT, IPv6, DHCP server. Upgrading from 6.18 made it so the bridge didn't show up in interfaces, but the router still worked and forwarded traffic fine. Rebooting the router a second time made it so nothing forwarded and the router rebooted every 60ish seconds. Ended up having to net install router and put config back on. So its running fine now on 6.19 with essentially the same config, just created on 6.19 not 6.18.

standart · Tue Sep 02, 2014 3:35 am

I can not turn ntp server in this version, already copied the ntp.npk to "Files" in RBs 750GL, and RB540G CCR1016 file. But in any work if the version is 6.19.

Can you help me?

parinya1957 · Tue Sep 02, 2014 4:28 am

After update from v6.17 to v6.19 for 6 of my router board, 1 RB450G loose PPPoE config on interface and I need to go to device to restore backup. Lucky that this unit not far from my home.

Other was fine for 1 of RG450G, 2 of RB2011UiAS-2HnD, 1 of RB1100AHx2 and 1 of RB951-2n.

I have 2 RB450G and 1 was success but the other success on upgrade to v6.19 but loose some config.

pturkey · Wed Sep 03, 2014 1:18 am

I upgraded a hotspot that is connected to rhe gateway router.
When it came back on Most of my network went dead.
After recycleing the power it everything is okay.
RB333
I will not be upgrading anymore with 6.19.

meetriks2 · Wed Sep 03, 2014 11:03 am

Cosmetic problem PPTP/SSTP server:

[admin@x] > ppp active print 
Flags: R - radius 
 #   NAME         SERVICE CALLER-ID         ADDRESS         UPTIME   ENCODING                                         
<snip>
13   probe        sstp    xxxxx     0.0.0.0         1h19m42s AES256-CBC                                       
14   probe        pptp    xxxxx     0.0.0.0         1h19m42s MPPE128 stateless                                

It shows 0.0.0.0 but is not. And was not showed this way in v5.26

[admin@xxx] > ip address print
Flags: X - disabled, I - invalid, D - dynamic 
 #   ADDRESS            NETWORK         INTERFACE                                                                     
 0   91.230.51.86/27    91.230.51.64    ether1                                                                        
<snip>
14 D 172.30.252.255/32  XXX.36.170.52    <sstp-probe-13>                                                               
17 D 172.30.252.255/32  XXX.28.108.110   <sstp-probe-14>

The clients pick the own IP address. Works fine.

Hope it gets fixed in v6.20

Wed Sep 03, 2014 12:53 pm

diablonet,
please provide us with console output, when your router enters endless loop.

I have used the same configuration for “ntp client” as you specified, my router works without any reboots. Perhaps there is any other configuration that is required.

igb,
I assume you need to check mss (MTU/MRU configuration for PPP profile), and decrease mss to 1300 in /ip firewall mangle rules, by max-mtu max-mru configuration at ppp server.

AlexS,
please provide us (support@mikrotik.com) with support output file from your router, when traffic-flow does not work for you.

JanezFord,
please provide us (support@mikrotik.com) with support output file from 6.19 version, when DHCP-leases are erased.

samim,
please post your issues in the appropriate forum section, this thread is about RouterOS 6.19.

Zorro,
what hardware do you use?

roadracer96,
please provide us (support@mikrotik.com) with support output file from your router.

standart,
please post entire output of command that you use to enable NTP server, and error returned by MikroTik RouterOS.

parinya1957,
please share your configuration with us. We will try to repeat the same problem in our test networks.

pturkey,
please add more details to your problem description. It would be convenient to get console output from your router, when problem is present.

meetriks2,
share you ppp configuration for local and remote addresses, that are assigned to clients.

meetriks2 · Wed Sep 03, 2014 1:12 pm

Here is my config, is very simpel. The linux clients have a static IP configured. I do not assign them.

[admin@x] /ppp> export
/ppp profile
set 1 idle-timeout=3m local-address=172.30.252.255
/ppp secret
add name=probe password=xxxx profile=default-encryption

[admin@x] /ppp> /interface
[admin@xx] /interface> export
/interface ethernet
set [ find default-name=ether2 ] disabled=yes
/interface ovpn-server server
set mac-address=AA:AA:AA:AA:AA:AA
/interface pptp-server server
set authentication=mschap2 enabled=yes max-mru=1460 max-mtu=1460
/interface sstp-server server
set authentication=mschap2 certificate=cert1 default-profile=default-encryption enabled=yes

Thanks!

snoppy · Wed Sep 03, 2014 5:49 pm

Hello!
CCR1036-8G-2S+EM
8000 simple queue, about 3500 active users.
we tested 6.19rc6 and had a poor perfomance.
now we test 6.19 and received some improvement

conntrack disable.
on one side we have 2G bonding, on enother side we have 10G.
when traffic 600Mbit/s 60Kpps, load 20%
when traffic 950Mbit/c 90kpps, load 100%, 5 sec and reboot adn so on.

on one side we have 10G, on enother side we have 10G.
when traffic 600Mbit/s load 20%
when traffic 950Mbit/c load 100%, 5 sec and reboot and so on.

ticket 2014081166000977 from 6.19rc6
suppout sent.
so, on increase in great performance in section "simple queue" it is necessary to dream

parinya1957 · Wed Sep 03, 2014 6:28 pm

Here is my config for one of my RB450G which loose PPPoE when upgrade by click "Check for Update" in Package List. The other was done success.

/ip pool
add name=default-dhcp ranges=192.168.0.20-192.168.0.254
/ip dhcp-server
add address-pool=default-dhcp disabled=no interface=ether5 name=default
/interface pppoe-client
add ac-name="" add-default-route=no allow=pap,chap,mschap1,mschap2 comment=\
    "" dial-on-demand=no \
    disabled=no interface=ether1 keepalive-timeout=60 max-mru=1480 max-mtu=\
    1480 mrru=disabled name=pppoe-out1 password=password profile=default \
    service-name="" use-peer-dns=no user=user
/queue type
add kind=pcq name=ether5-PCQ-Upload pcq-classifier=src-address pcq-limit=40 \
    pcq-rate=1M
add kind=pcq name=ether5-PCQ-Download pcq-classifier=dst-address pcq-limit=40 \
    pcq-rate=13M
/ip address
add address=192.168.0.1/24 comment="default configuration" interface=ether5 \
    network=192.168.0.0
add address=1.1.1.1/24 interface=ether5 network=1.1.1.0
/ip cloud
set enabled=yes
/ip dhcp-server alert
add disabled=no interface=ether5 on-alert=print
/ip dhcp-server network
add address=192.168.0.0/24 comment="default configuration" dns-server=\
    192.168.0.1 gateway=1.1.1.1
/ip dns
set allow-remote-requests=yes cache-size=5000000KiB max-udp-packet-size=512 \
    servers=8.8.8.8,8.8.4.4
/ip dns static
add address=192.168.0.1 name=router
/ip firewall filter
add action=drop chain=forward disabled=yes p2p=all-p2p
add action=passthrough chain=unused-hs-chain comment=\
    "place hotspot rules here" disabled=yes
/ip firewall nat
add action=passthrough chain=unused-hs-chain comment=\
    "place hotspot rules here" disabled=yes to-addresses=0.0.0.0
add action=masquerade chain=srcnat out-interface=pppoe-out1
add action=masquerade chain=srcnat comment="masquerade hotspot network" \
    src-address=192.168.0.0/24 to-addresses=0.0.0.0
/ip proxy
set parent-proxy=0.0.0.0
/ip route
add check-gateway=ping distance=1 gateway=pppoe-out1 routing-mark=to_ether1
add check-gateway=ping distance=1 gateway=pppoe-out1
/ip service
set telnet disabled=yes
set ftp disabled=yes
set www port=81
set ssh disabled=yes
set api disabled=yes
/system clock
set time-zone-name=Asia/Bangkok
/system ntp client
set enabled=yes primary-ntp=17.82.253.7 secondary-ntp=17.72.255.11
/tool graphing interface
add interface=pppoe-out1

JanezFord · Wed Sep 03, 2014 8:53 pm

JanezFord,
please provide us (support@mikrotik.com) with support output file from 6.19 version, when DHCP-leases are erased.

I have already made new static dhcp leases .... if this situation occurs again with 6.20 (as it did with 6.18 and 6.19) I will provide you with support out file for sure.

both routers are in production environments and I can't just downgrade them and play around when I wish to. I have to find a free time slot and announce maintenance it to my customers ... If this weekend is free I will try to downgrade and upgrade again to provide you support out file with 6.19 version.

JF.

roadracer96 · Wed Sep 03, 2014 10:35 pm

roadracer96,
please provide us (support@mikrotik.com) with support output file from your router.

I couldn't get one. Any time I tried to do that or export config, it would reboot. I had to net-install it.

ALEJANDROgordon · Thu Sep 04, 2014 4:45 pm

I have serious problem with the beta release, because the wireless link is unstable, the ccq down to 17/10%, in the 6.17 version using nv2 and wireless fp package, we have 99/99% ccq and the ping response is low.

Thu Sep 04, 2014 5:11 pm

snoppy,
support is working on your issue. Thank you for the cooperation.

parinya1957,
pppoe-client configuration is lost, is it correct?

JanezFord,
Thank you for the update. I’m sorry for any inconvenience caused.
We had multiple routers with big number of static and dynamic leases, and none of them had issues during the upgrade.

parinya1957 · Thu Sep 04, 2014 6:50 pm

parinya1957,
pppoe-client configuration is lost, is it correct?

Yes, correct.

pturkey · Thu Sep 04, 2014 7:03 pm

I was unable to look.
My network was not accessible.

pturkey · Thu Sep 04, 2014 7:59 pm

RB711-5HnD
I upgraded to 5.19 and then tried a btest.
Within 3-4 seconds it lost wireless connection.
I changed to several AP's and got no connections.
I also tried a scan for a AP, there was nothing listed.
When I rebooted everything started working.
Tried two times with same resolts.
I did make a supout.rif.
After downgrading I ran the same btest and it was working okay.

ALEJANDROgordon · Thu Sep 04, 2014 9:55 pm

RB711-5HnD
I upgraded to 5.19 and then tried a btest.
Within 3-4 seconds it lost wireless connection.
I changed to several AP's and got no connections.
I also tried a scan for a AP, there was nothing listed.
When I rebooted everything started working.
Tried two times with same resolts.
I did make a supout.rif.
After downgrading I ran the same btest and it was working okay.

5.19 or 6.19

joserudi · Thu Sep 04, 2014 11:00 pm

PayPal dont work in this versión. After upgrade from 6.13 show this error:
"PayPal - ssl connection error: handshake failed [unable to get certificate CRL] (6)"

pturkey · Fri Sep 05, 2014 1:42 am

joserudi · Fri Sep 05, 2014 6:11 am

With 6.19 paypal not works, Another friend say that until 6.18 it works fine.

Chupaka · Fri Sep 05, 2014 12:38 pm

We had multiple routers with big number of static and dynamic leases, and none of them had issues during the upgrade.

Sergejs, a friend of mine called me a week ago when he upgraded his router to 6.19, and all his static leases disappeared, so definitely problem does exist

JFYI

morf · Fri Sep 05, 2014 11:33 pm

Not working lease-script if using ip-pool in the address field. In other cases, works without problems.
Example:

/ip dhcp-server lease
add address="TEST_POOL" comment=\
    id_E0282F67-ED0E-4116-A743-03435D123444 mac-address=00:0F:02:68:43:7C \
    server="TEST_SERVER"
add address="TEST_POOL" comment=\
    id_B506877C-D021-4335-AC7A-C4AA03E2EAFE mac-address=20:89:84:61:30:7F \
    server="TEST_SERVER"
add address="TEST_POOL" comment=\
    id_7C845A2E-870F-48A8-83CE-6675E9F3F3B4 mac-address=00:8E:F2:EB:0C:F1 \
    server="TEST_SERVER"

Sob · Sat Sep 06, 2014 5:00 pm

NTP client server-dns-names reboot loop

1) Connect device (I used SXT Lite5) to ethernet, configure network address, default route and dns server (just minimal config).

2) Set /system ntp client set enabled=yes server-dns-names="0.pool.ntp.org" to enable NTP client. It will get correct time.

3) Reboot device, while still connected to network. Everything will be ok.

4) Reboot device, but this time without network connectivity (disconnect ethernet completely / connect it directly to notebook / put it into isolated VLAN / ...). Enjoy reboot loop.

Or for quick reboot:

Take configured device (from step 1), connect it directly to notebook and set /system ntp client set enabled=yes server-dns-names="0.pool.ntp.org" from there. Wait few seconds for reboot. It will come up fine, because it reboots before it saves settings. Log will say that is was rebooted by watchdog timer.

littlebill · Sun Sep 07, 2014 5:34 am

RBMAp

bunch of issues found

http://forum.mikrotik.com/viewtopic.php ... 96#p445996

steen · Sun Sep 07, 2014 2:10 pm

Hello Folks!

We updated a couple of CRS125, RB411, RB433 and SEXTANT(RB711G) from ROS6.15, on Sextant and RB433 we did also change to wireless-fp, all updates was successful, nothing unusal was observed.

However, the configurations on all devices contains nothing exotic, just simple static routing and queues plus vlans and trunk links.

We are still stuck to RoS6.7 on the core routers & firewalls, because of that routing marks (used for policy based routing) is still broken on all later releases we tested. Unfortunately that has lead to the business has decided to invest in some more Cisco ASA5510 and Cisco routers instead of the new CCR, but that is life.

Chupaka · Sun Sep 07, 2014 10:52 pm

because of that routing marks (used for policy based routing) is still broken on all later releases we tested.

what exactly did you test? is it repeatable on v6.19? how?

eavictor · Mon Sep 08, 2014 9:11 am

The firewall package is complete down.

I'm using 3 VPSs with RouterOS x86 to make my private TV Stream Server available from several place.
(That shitty ISP only provides private IPs)

After upgrade to 6.19, src-nat only function when src-address=0.0.0.0/0
if I set the src-address=1.2.3.4 or dst-address=1.2.3.4-1.2.3.10 the rule doesn't function.

firewall drop rules like src-address=!1.2.3.4 are malfunctioning, it blocks 1.2.3.4 but not other IPs.
firewall accept rule like src-address=1.2.3.4 doesn't functioning either.

This problem appears on all of my 3 VPSs,2 of my CCR1036-12G-4S, RB1100AHx2, RB951-2n, CRS125-24G-1S-2HnD-IN.
If support team members need my VPS for digging where the problem is, please PM or send me e-mail, I will open one of my VPS for you.

========
sorry for miss typing, "dst" should be replaced as "src"
I am awake without closing my eyes for over 20 hours.
========

rextended · Mon Sep 08, 2014 11:43 am

Please help:

http://forum.mikrotik.com/viewtopic.php?f=2&t=88902

For MikroTik staff:

Can you please consider to add one section of the forum dedicated esclusively for the prevention of the Internet threats?
The section is for communicate between various WISP / ISP to prevent and/or detecting threats, spam bot, brute force attack, DDoS and DoS, etc.

Thanks.

honzam · Mon Sep 08, 2014 11:56 am

Upgrade rb800 from 6.12 to 6.19 fail.
Power cycle needed

steen · Mon Sep 08, 2014 4:37 pm

because of that routing marks (used for policy based routing) is still broken on all later releases we tested.
what exactly did you test? is it repeatable on v6.19? how?

For years we have tagged source address: scmgateway protocol: tcp/smtp with mailroute.
In routing we route all traffic tagged mailroute to smtp asa5510
Default gateways we have two using classic mikrotik failover, asa5510 is not any default gateway.

Suddenly after upgrade, one customer call in and say, hey where is my emails, customer complains.

After investigating where mail traffic went, all has been queued up in scmgateway, routerboard had not passed traffic on to asa5510, it is not going anywere and stay in mailqueue.

Rollback to RoS6.7 immediate solved the problem.

Chupaka · Mon Sep 08, 2014 4:52 pm

steen, can you repeat this situation in a test lab?

picacho99 · Mon Sep 08, 2014 7:26 pm

1- Updated RouterBoof from 3.14 to ar9344_3.18.fwf before.
2- Upgraded from 6.12 to 6.19 without issues in a configuration that is summarize below:
- PPP across a fiber provider 100/10Mbits.
- IGMP forwarding with several clients.
- ABout 50 fireall rules.
- Multiple VLANs managed in both the ISP side and internally: VoIP, IPTV decoder and regular internet traffic.
- Authomated Script to update external DNS server.
- L2TP server to access internal Intranet.

After a week, nothing wrong....good job Mikrotik !!!!

WirelessRudy · Mon Sep 08, 2014 8:08 pm

Maybe if he would give us a printout of the firewall mangle, nat and route tables and rules we can see or repeat the issue ourselves..?

miguel1182 · Tue Sep 09, 2014 4:58 am

I don't know if its hardware problem but since i'm upgraded my RB2011 simple queue start to have some problems, they work fine but like 5 seconds they give you 200% bandwidth speed you configured, then it only gives you 50% and at randoms intervals do the same like some kind of burst even if isn't configured on simple queue
Sorry for my bad english

AlexS · Tue Sep 09, 2014 2:33 pm

Netflow is patchy

reboot -> works
reboot -> doesn't work

But I have a supposition it has some thing to do with my a sync routing, its not starting records unless it seems a syn packet ... i think !
What netflow version do you use? V9 is supposed to report flow stats on a per-connection basis.

So does that mean if it doesn't see the syn it will not start the counters ?

A

adrcomms · Wed Sep 10, 2014 12:14 am

We loaded 6.19 on our ccr 36, simple queues very fast and cpu load very low.

problems:
customers on pppoe started to get double ip's as we permit 2 ips per account, one for the cpe at home and the other to be used on the hotspot, never seen two ip's on other versions.
hotspot started to block authorised clients or bypassed ip ranges.

backed down to v6.6 as this stable but the simple queues are slower.

Adrian

mramos · Wed Sep 10, 2014 6:14 am

Hi ...

I upgraded a RB411AR from 5.26 => 6.19 and since then the built in 2.4GHz interface started to disconnect from AP from time to time (e.g 3 ... 6 times a day).

With 5.23 It remains connected for months (AP uptime now ~180 days ROS 5.26 only G mode, 10MHz channel).

On 6.19 there is not the Rate Selection aditional feature LEGACY versus ADVANCED.

Which one is used?

Regards;

adyb76 · Wed Sep 10, 2014 9:43 am

Installed on our CCR1036 which is our QoS, with many mangle rules and a comprehensive queue tree, passing over 800mbit/s at peak times. All went ok, not sure about performance improvements yet - tonight will be the best time to check this out.

Adyb

PastuhMedvedey · Wed Sep 10, 2014 1:46 pm

What's new in 6.19 (2014-Aug-26 14:05):
*) bridge - added "Auto Isolate" stp enhancement (802.1q-2011, 13.25.6)

Please tell me where you can find a description of this function in the documentation. Follow the link below, found nothing. http://wiki.mikrotik.com/wiki/Manual:Interface/Bridge

Wed Sep 10, 2014 2:56 pm

What's new in 6.19 (2014-Aug-26 14:05):
*) bridge - added "Auto Isolate" stp enhancement (802.1q-2011, 13.25.6)
Please tell me where you can find a description of this function in the documentation.

In the aforementioned 802.1q-2011 standard, obviously. You can get yourself a copy of the standard document here.

Paragraph 13.25.6 says:

13.25.6 AutoIsolate
A Boolean. Set by management if isolate (13.25.20) is to be set, causing a Designated Port to transition to
Discarding if both AdminEdge (13.25.1) and AutoEdge (13.25.5) are FALSE and the other bridge presumed
to be attached to the same point-to-point LAN does not transmit periodic BPDUs—either as a Designated
Port or in response to BPDUs with the Proposal flag set (see 13.21). The recommended default of this
parameter is FALSE. AdminEdge and AutoEdge are both reset only on ports that are known to connect to
other bridges.

isai · Wed Sep 10, 2014 10:19 pm

Motherboards and firmware ect compatible with v6.xx?
many new boards out there...and your wiki only shows v3 and v4

Clauu · Thu Sep 11, 2014 11:38 am

I was modifing some dhcp settings/networks and after that bum all settings/networks from dhcp server section disappeared.. same for ip pools, all my 10pools had gone away.. i'm kind a tired of this very unpredictable behaviour every time i'm having emotions that something will go bad and i must reboot or it will crash etc

pchott · Thu Sep 11, 2014 1:35 pm

We are using Tftp over MikroTik and after upgrade to 6.19 we experience that we cannot transfer filenames with filename lenght 6? Does anyone alse uses TFTP over Mikrotik?

EDIT:
6x8bit = 32bit filename length and with uboot there is no way to transfer such filename. Looks like i have to go back to 6.15

zervan · Thu Sep 11, 2014 7:45 pm

It seems like there is something wrong with dhcp-server lease address-list parameter. Here is a line of my last backup - as an example:

/ip dhcp-server lease
  add address=192.168.1.91 address-list=klient mac-address=00:1E:74:DD:A3:9F server=dhcp-local

Everything was working for years. Today one of local admins contacted me about a new PC problem - he added new address, but something was not working. I found out that he didn't add address-list parameter as he should - well, but he just made copy of other leases and there is no address-list parameters at all and they were there till now. Did he delete them? I don't think so, I have also checked other of my routers where I use address-list (RB1200, RB751, RB2011) - none of them has address-list parameters set - they were lost probably after 6.19 update!

Here is the same line when I made backup now:

/ip dhcp-server lease
  add address=192.168.1.91 mac-address=00:1E:74:DD:A3:9F server=dhcp-local

The strange thing is that the old leases do add addresses to the list despite the fact that they are not listed in configuration.

Does somebody have this problem or I am the first one? This is the second serious bug in 6.19 that affected me

zervan · Thu Sep 11, 2014 8:24 pm

Another question about dhcp lease address-list problem is - how to fix that? I tried command

/ip dhcp-server lease set address-list=klient numbers=[find server=dhcp-lan2]

- but it is not displaying it neither and there are messages in the log like

dhcp,error dhcp-lan2: failed to add address-list klient for IP 192.168.3.62: already have such entry (6)

lambert · Thu Sep 11, 2014 9:45 pm

I believe the format of the set command is :

set {entry number(s)} things to set

Your command should have been in the format of:

/ip dhcp-server lease set [find server=dhcp-lan2] address-list=klient

But the error you are getting says that the address-list entry already existed in /ip firewall address-list

You can probably ignore that since some of your entries probably already had address-list settings.

zervan · Thu Sep 11, 2014 10:59 pm

The format is correct, the same error appears if I set address-list to particular lease number.

The first problem is that since manual setting of Address List in WinBox will remain and seems to work correctly (despite the same error in log), setting using command in Terminal is not working correctly. Maybe better said: it may work somehow "hidden", but the value is not displayed in Winbox, nor export. The second problem is that even working values (which are displayed in Winbox) are lost after reboot.

Is there anybody who can confirm this problem? I am the only one which is using dynamic address list from DHCP clients? I found a very simple way to reproduce the problem: add an Address List to some DHCP lease, reboot, check the DHCP lease - there is no Address List mentioned in 6.19.

tkgit · Fri Sep 12, 2014 6:37 am

i use rb751u2hnd, if i enable ppp-out 3g use usb 3g modem, cpu process suddenly rise to 100%
if i disable ppp-out, cpu process back to normal again, why? this problem began after RoS 6.xx

gvitalik · Fri Sep 12, 2014 8:50 am

i use rb751u2hnd, if i enable ppp-out 3g use usb 3g modem, cpu process suddenly rise to 100%
if i disable ppp-out, cpu process back to normal again, why? this problem began after RoS 6.xx

I am not using 3g modem, and my router is RB951G-2HnD, but on each 7 day after reboot cpu process rise to 100%.
I don't know how to back it to normal, except rebooting.

zervan · Fri Sep 12, 2014 10:07 am

I found a very simple way to reproduce the problem: add an Address List to some DHCP lease, reboot, check the DHCP lease - there is no Address List mentioned in 6.19.

There is no need to reboot - closing Winbox and opening again is enough. I have reported to support.

zervan · Fri Sep 12, 2014 10:12 am

i use rb751u2hnd, if i enable ppp-out 3g use usb 3g modem, cpu process suddenly rise to 100%
if i disable ppp-out, cpu process back to normal again, why? this problem began after RoS 6.xx

No problem with that using RB951Ui-2HnD + ZTE modem, nor RB2011UiAS-2HnD + Huawei modem, both with RouterOS 6.19 and previous. Check Tools -> Profile, which process is using CPU, maybe it is related to firewall rules or something like that.

gvitalik · Fri Sep 12, 2014 11:37 am

i use rb751u2hnd, if i enable ppp-out 3g use usb 3g modem, cpu process suddenly rise to 100%
if i disable ppp-out, cpu process back to normal again, why? this problem began after RoS 6.xx
No problem with that using RB951Ui-2HnD + ZTE modem, nor RB2011UiAS-2HnD + Huawei modem, both with RouterOS 6.19 and previous. Check Tools -> Profile, which process is using CPU, maybe it is related to firewall rules or something like that.

Hi zervan!

May be you know what is better to do in case ipsec show 96% of cpu in profile, but at this moment I have no any ipsec connection? I don't know why, but I have configured l2tp/ipsec vpn server before, but now there no any ipsec proposal and peers. If i open PPP->l2tp configuration and save it again, then the system re-create ipsec proposal, but after some period of time, this proposal will be "deleted" again.

zervan · Fri Sep 12, 2014 1:06 pm

May be you know what is better to do in case ipsec show 96% of cpu in profile, but at this moment I have no any ipsec connection? I don't know why, but I have configured l2tp/ipsec vpn server before, but now there no any ipsec proposal and peers. If i open PPP->l2tp configuration and save it again, then the system re-create ipsec proposal, but after some period of time, this proposal will be "deleted" again.

Sorry, but I don't know the reason. Try to export configuration and check if there is really no forgotten ipsec / l2tp mentioned. Is this problem related to RouterOS 6.19? Was it working in previous versions? If not, you should create new post / topic in appropriate section of this forum and somebody will know to help, I think.

I have problems with high CPU load only when I am connected to Winbox and Profile show non-senses like "flash". When I disconnect, CPU load is low again (I can see it on graph). This problem I have for about 10 versions back. But it is probably not the problem you have.

zervan · Fri Sep 12, 2014 1:09 pm

I found a very simple way to reproduce the problem: add an Address List to some DHCP lease, reboot, check the DHCP lease - there is no Address List mentioned in 6.19.
There is no need to reboot - closing Winbox and opening again is enough. I have reported to support.

I've got answer from Mikrotik support - this problem should be fixed in 6.20rc6. I will try in the evening, but I worry - what if there are some new bugs? Early implementation of 6.19 brought serious troubles to me.

rextended · Fri Sep 12, 2014 6:20 pm

What's new in 6.20rc6 (2014-Sep-08 10:16):

*) pppoe client - increase connection timeout to make connection establishment
possible on busy pppoe server;
*) dhcp server - change default lease time from 3 days to 10 minutes
to avoid running out of IPs;
*) ipsec - allow binding modeconf address to username;

Winbox v3 is required for router’s configuration.

SERIOUSLY? IS STILL FULL OF BUG AND IS NEEDED FOR 6.20!!!???

zervan · Fri Sep 12, 2014 6:37 pm

What's new in 6.20rc6 (2014-Sep-08 10:16):
...
Winbox v3 is required for router’s configuration.
SERIOUSLY? IS STILL FULL OF BUG AND IS NEEDED FOR 6.20!!!???

Unfortunately, yes. I have upgraded one of my routers to 6.20rc6 to test if Address List of DHCP Lease is fixed (it is not) and it is not possible to connect with the old good Winbox

CyberTod · Fri Sep 12, 2014 6:41 pm

I am connecting through Dude, because it is more convenient to work with a lot of devices. If I will not be able to that anymore because it is using old winbox version that is a major problem for me.

P.S. I just found a workaround for that by making a custom tool in Dude and if indeed winbox 3 is required for administering v6.20+ I will post a short guide with some pictures.

noyo · Fri Sep 12, 2014 7:10 pm

I am connecting through Dude, because it is more convenient to work with a lot of devices. If I will not be able to that anymore because it is using old winbox version that is a major problem for me.

solution is:
1. Create a directory "Dude" in the C:\
2. Copy the new winbox3 to C:\Dude\
3. Go to Dude > Tools > + end
3.1. name: Winbox3
3.2. command: "C:\Dude\winbox3.exe" [Device.FirstAddress] [Device.UserName] [Device.Password]

Connect to new winbox:
Right click on the device > tools > winbox3
Works on windows, how and on Linux under Wine.

good luck

CyberTod · Fri Sep 12, 2014 7:14 pm

I just did that

With the exception that my path points to where my winbox3 is currently saved.
The downside is that the integrated in Dude winbox connects a lot faster. And the known bug of not remembering session with the new one.

noyo · Fri Sep 12, 2014 7:19 pm

t's terribly sad that we have to pay for underdeveloped software.
Also a shame that mkrotik tests its software on the clients

mperdue · Fri Sep 12, 2014 11:26 pm

I decided to be carefull before deploying 6.19. I have some spare's here at the office so I've upgraded 12 units and bricked 4. Two Rb133c's, one RB SXT 2nd (new out of the box), and one RB912UAG-2HPnd (new).

They will light up quickly one beep and then nothing.

How can I unbrick these units?

-Michael

athlonxp78 · Fri Sep 12, 2014 11:30 pm

What's new in 6.20rc6 (2014-Sep-08 10:16):

*) pppoe client - increase connection timeout to make connection establishment
possible on busy pppoe server;
*) dhcp server - change default lease time from 3 days to 10 minutes
to avoid running out of IPs;
*) ipsec - allow binding modeconf address to username;

Winbox v3 is required for router’s configuration.

WTF!!!??? The Beta3 of Winbox v3 is buggy and you "lock us" to work with this version?

Are you crazy?

It´s a JOKE?

steen · Sat Sep 13, 2014 10:40 am

Hello Folks!

We successfully updated about 1/3 of of our MT equipment to RoS6.19 and firmware, we also enabled wireless-fb package on them. RB411, RB433, SXT, SEXTANT and a couple of switches CRS125.

We did it in small steps like, first a baseline reboot then update RoS + reboot, then update firmware + reboot and last enable wireless-fb + reboot.

Still the routing marks for policy based routing is broken so we are stick to RoS 6.7 on those.

kobuki · Sat Sep 13, 2014 5:01 pm

What's new in 6.20rc6 (2014-Sep-08 10:16):

*) pppoe client - increase connection timeout to make connection establishment
possible on busy pppoe server;
*) dhcp server - change default lease time from 3 days to 10 minutes
to avoid running out of IPs;
*) ipsec - allow binding modeconf address to username;

Winbox v3 is required for router’s configuration.
WTF!!!??? The Beta3 of Winbox v3 is buggy and you "lock us" to work with this version?

Are you crazy?

It´s a JOKE?

I must agree. The v3 beta3 cannot reliably re-open the windows and their arrangements when I open an old session auto-imported from winbox2. This effectively makes it completey useless in production. Please save us from testing your half-baked code. There are a lot of small problems and mistakes one can tolerate when forced upon. This is certainly not among them.

bajodel · Sun Sep 14, 2014 9:54 am

What's new in 6.20rc6 (2014-Sep-08 10:16):

*) pppoe client - increase connection timeout to make connection establishment
possible on busy pppoe server;
*) dhcp server - change default lease time from 3 days to 10 minutes
to avoid running out of IPs;
*) ipsec - allow binding modeconf address to username;

Winbox v3 is required for router’s configuration.
SERIOUSLY? IS STILL FULL OF BUG AND IS NEEDED FOR 6.20!!!???

if confirmed I think ros 6 series will stop at 6.19 for me ..

.. pure crazyness

hashbang · Sun Sep 14, 2014 6:00 pm

+1
winbox 3 is buggy. It doesnt run properly on wine. The cursor will blink on the extreme right of ip address but if u use back space it will delete the 1st octet of ip address. So painful for anyone who uses keyboard and doesnt watch the screen

. Specially when u have a long list of devices in your winbox list.

xcom · Mon Sep 15, 2014 2:05 am

If this will be the case for all firmwares going forward, I will take my switch off my network and set it on fire...
I am a linux user and having to install wine to run another buggy ass application... Hell no!... You can keep the spoon.

lambert · Mon Sep 15, 2014 2:48 am

We're complaining that pre-release RouterOS requires pre-release Winbox?

Maybe MikroTik just wants to ensure that they get some Winbox 3 feedback from the people who are already demonstrably risk takers. If you are running RCs, you are a risk taker.

Anyone who cares should e-mail support and ask them if they intend for Winbox 3 to be production ready before RouterOS 6.20 is released.

Personally, I think it would be a mistake for MikroTik to require Winbox 3 to manage any RouterOS 6.x release. They can require whatever they want of the RCs. They can require whatever they want for 7.x. But the 6.x releases should not require untested management methods.

And none of that has anything to do with RouterOS v6.19. Start another topic already.

AlexS · Mon Sep 15, 2014 4:11 am

If this will be the case for all firmwares going forward, I will take my switch off my network and set it on fire...
I am a linux user and having to install wine to run another buggy ass application... Hell no!... You can keep the spoon.

I'm presuming you can still ssh to the box.

If not then 6.19 will also be my final firmware..

A

kapulan · Mon Sep 15, 2014 12:17 pm

Remote mac-telnet not work(connection timeout) all v6 version change the wireless mode to station bridge !Change the mode to simple station work again the mac-telnet.I use NV2 .
Mt team please fix this problem.

zyzelis · Mon Sep 15, 2014 12:45 pm

Just upgraded a whole network to v6.19.
Devices are CCR1036, 1100AHx2, SXT lite5;lite2, SA5r2, G2HnD, G-5HPacD-SA, G-5HPacD, RB433GL, OmniTik-U and UPA,
RB750GL, 750, 750UP, RB2011-LS;RM;UAS, RB951G;U, RB711-5h,RB911-5Hnd and grooves A52HN.
works stable so far, no "kernel failure" crashes, all wireless devices with fp packages
Thanks Mikrotik for stable package.

snoppy · Mon Sep 15, 2014 12:57 pm

Hi all!
we tested 6..20rc6
4000 simple queue
active users from 900 for 3800

conntrack ON
load 100%,traffic up to 1.7Gb/s
when load to 90-100% (1.7G) router freeze on 5-10 seconds and traffic down to 700Mb/s
connrack OFF
load 10-15% traffic up to 2.3Gb/s
sometimes load 100% and router freeze for 5-10 seconds , but traffic not down.
sometimes load up to 40-50% and then slowly going down to 10-15%

suppout was sent.
Ticket#2014081166000977 with Martins S.

voxframe · Mon Sep 15, 2014 4:07 pm

normis!

This really bothers me!
Why do you lock us into beta firmware/software before stabilizing the current branch!

Before you did it with the 2011 board and 6.x firmware. Some boards worked with 5.x and some (after a still undocumented hardware change) require 6.x.

Now you are doing it with a buggy version of winbox!

Guys... STOP! STABILIZE THE DAMN THING! Then slowly implement changes. winbox3 should not make ANY appearance in 6.x. It should be reserved for 7.x

Come on! All I want is for things to be stable like 5.26, but I can't run it on half of your hardware, and currently no version of 6.x is trustable.

adyb76 · Mon Sep 15, 2014 6:54 pm

Hi,

We have recently upgraded to 6.19 from 6.7 on couldcore router with ubiquiti AF24 and now have FCS errors on interfaces that have an AF24 connected to them.
We have this issue previously and you guys managed to fix it but it's reared its ugly head again....

Thanks,

Adyb

hashbang · Tue Sep 16, 2014 8:40 am

adyb76 please post the screenshot
vox +1

steen · Tue Sep 16, 2014 1:37 pm

Hello Folks!

We now updated even more RB411 and SXT and SEXTANT:s to RoS6.19 and wireless-fp, no problems so far.

The only observation is that the RB devices takes long time to come back on the network after updating RoS, the same is after the firmware is updated and when switching package from wireless to wireless-fb.

Well, then we have to broken routing marks that prevents us from upgrading several edge routers, business has decided to pick another brand to replace them....

Policy based routing is a must for us.

OlegM · Tue Sep 16, 2014 1:44 pm

Version 6.19 CCR. Accordingly CCR serves as PPTP server. The role of the client VPN connection that uses torrents. Noted in the PPTP connection, that when working torrents connection is not broken, but the packets lost as long as you do not restart PPTP. And it was seen in the earlier versions.

zyzelis · Tue Sep 16, 2014 2:16 pm

It was early to be happy with 6.19.
On CCR1036 which is connected via SFP to RB260GSP now i regullary get FSC error on link.
Already tried to turn off autonegotiation and set to FDX 1G, problem still persist
Before CCR was on 6.12 and such problem wasn't......

So be carefull.

Robinson · Wed Sep 17, 2014 12:01 am

After upgrading Ros 5.20 to Ros 6.19 Sextant (RB711-5HnD) lost WLAN port. After disconnecting and connecting the power supply interface has returned, but with defoult settings. 1 hour on a 6 km. remote tower to return the the link.

slebrun · Wed Sep 17, 2014 1:12 am

We upgrade a pair of 493AH routers to 6.19, and saw EoIP start dropping packets > 1422 bytes. Rolled back to 6.17, no problems.

We did not, however, upgrade the far end of the EoIP tunnels, which may or may not have wound up working better.

bajodel · Wed Sep 17, 2014 8:47 am

6.20rc9 works with winbox2 ..thanx

gemwk2 · Wed Sep 17, 2014 11:30 am

There is a problem with RB SXT G-5HPnD r2. This board restarting itself because of kernel failure and log is like this "00:42:21 system,error,critical router was rebooted without proper shutdown by watchdog timer ".

Wed Sep 17, 2014 11:36 am

There is a problem with RB SXT G-5HPnD r2. This board restarting itself because of kernel failure and log is like this "00:42:21 system,error,critical router was rebooted without proper shutdown by watchdog timer ".

how is your watchdog configured? maybe it is pinging some address

hedele · Wed Sep 17, 2014 12:31 pm

Winbox v3 is required for router’s configuration.

Welp... no more rc testing (or upgrading, for that matter) for me then...
I'll stay on 6.18 on all of my devices, since EoIP (which I use it a lot) is broken on 6.19.
6.18 was a great release (the first really stable and usable one with no major bugs for my use-cases in the whole 6.x branch) and I hold on to it dearly.

Wed Sep 17, 2014 12:38 pm

Winbox v3 is required for router’s configuration.
Welp... no more rc testing (or upgrading, for that matter) for me then...
I'll stay on 6.18 on all of my devices, since EoIP (which I use it a lot) is broken on 6.19.
6.18 was a great release (the first really stable and usable one with no major bugs for my use-cases in the whole 6.x branch) and I hold on to it dearly.

why?

hedele · Wed Sep 17, 2014 12:57 pm

Well as you can read from some of the feedback in this thread, winbox 3 does not seem mature enough for production use yet.
Also, I don't think that there is any technical reason that winbox 3 is required for new rc builds, it's probably just politics to force people into beta-testing winbox 3.

gemwk2 · Wed Sep 17, 2014 4:06 pm

There is a problem with RB SXT G-5HPnD r2. This board restarting itself because of kernel failure and log is like this "00:42:21 system,error,critical router was rebooted without proper shutdown by watchdog timer ".
how is your watchdog configured? maybe it is pinging some address

Watchdog is configured by default, no configuration is set.