v6rc14 released

Tue Apr 30, 2013 8:57 am

What's new in 6.0rc14 (2013-Apr-24 11:52):

*) route - make connected routes inactive when interface has no link;
*) ipsec - changing or removing unused peer or proposal config won't flush active SAs;
*) console - add 'without-paging' to more 'print' commands;
*) route - automatically repair FIB inconsistencies;
*) ipsec mode-cfg - unity split include support;
*) ipsec policy - template matching for policy generation;
*) metarouter: fixed occasional lockups on mipsbe boards;
*) fixed crash when bridge filter rule had action=return for rule in builtin chain;
*) traffic-flow - fixed deadlock and crash on multicore;
*) fixed memory leak on CCR with PPPoE interfaces;
*) improved PPPoE interface encapsulation performance;
*) fixed queues - total amount of traffic passing through queues sometimes was about 1Gbit;

zanetti · Tue Apr 30, 2013 10:06 am

Upgraded an RB2011L from rc13 to 14. DHCP Client and DHCP Server would not issue to or bind to addresses. Downgraded to 5.24, all is back to normal.

leoktv · Tue Apr 30, 2013 11:04 am

Upgraded an RB2011L from rc13 to 14. DHCP Client and DHCP Server would not issue to or bind to addresses. Downgraded to 5.24, all is back to normal.

HI i can confirm this problem.

elgrandiegote · Tue Apr 30, 2013 3:46 pm

send backup by mail script, still does not work, in the v6rc11 version works fine...
any news on this?

mrz · Tue Apr 30, 2013 3:49 pm

send backup by mail script, still does not work, in the v6rc11 version works fine...
any news on this?

Debug your script which part of it exactly is not working.

DirectX9 · Tue Apr 30, 2013 4:12 pm

Hi Normis
Very nice to see the updated ROS 6rc14
Very interesting bug with Atheros 9344. What's new in 3.07 firmware
Only one question - and when we can change the DTIM interval for SSID (default is 1 and cannot be changed by software - no option) to 2 or higher?
A lot of client side devices looses the battery power very fast because of Mikrotik DTIM is selected 1.
In 5.25 release I had asked the same before.
Please answer.

BETEPAH · Tue Apr 30, 2013 4:53 pm

Bug with SNMP not fixed.

jozwiak · Tue Apr 30, 2013 5:11 pm

Problem with IPsec Peer and "Generate policy" Not working.
Downgraded til v6rc13 and it worked again

raz · Tue Apr 30, 2013 5:37 pm

Im running rc14 since few Days, MT made here a good Job.

Thanks also for fixing this traffic-flow Deadlock.

btw: Is there a Model planned with Dual CPU? or 72 Core tilegx?

fsfranco · Tue Apr 30, 2013 9:23 pm

After for Release 6rcv14 Update, i am lost connection in RB 1100AHx2

Connection via console, reset-configuration and downgrade for version 6rcv13.

Its Work OK!!

[]s
Franco

majkel · Wed May 01, 2013 9:57 am

BUG on RC14 Official from 24/Apr ~~11:00 on CCR 1036

There is big CPU usage of Profiling (5-6% of all cores)

This bug doesnt exist in pre-release version of rc14 from Apr/23/2013 16:37:18

check this..

majkel · Wed May 01, 2013 10:03 am

Im running rc14 since few Days, MT made here a good Job.

Thanks also for fixing this traffic-flow Deadlock.

btw: Is there a Model planned with Dual CPU? or 72 Core tilegx?

why 2CPU ? 72 cores?

THere still NO OPTIMIZATION on 36 cores and you wont 2 CPU or 72 cores?

any ddos on TCP attack about 400mbit killing CCR in two seconds.. and for me I need to buy some another firewall.. becouse ccr are useless

NetworkPro · Wed May 01, 2013 10:58 am

Intel PRO/1000 (e1000?) VLAN tagging fail

2 device="02:01.0" name="82547EI Gigabit Ethernet Controller (rev: 0)"
vendor="Intel Corporation" category="Ethernet controller"
vendor-id="0x8086" device-id="0x1019"
(OnBoard Intel Server Board S875WP1-E http://downloadmirror.intel.com/19459/e ... 5WP1-E.zip)

Temporary workaround: create new bridge, put the problematic ether interface in that bridge. Add the VLANs with interface=that bridge so that bridge software would now be taking care of vlan tagging etc. instead of the NIC driver. (Note: the VLANs are not bridge ports of the mentioned bridge. They can be bridge ports of another bridge)

@MT I will send you the supout.rif file taken when problem is observed.

p.s. it may be vlan offload bug or otherwise Intel bug in my opinion !!!! (where MT could choose the driver, tweaks, they have in ROS, maybe NIC firmware or if they can - patch the damn driver (patching of NIC firmware is also a possibility))

P.S. 2: Intel PRO/1000 NIC driver in RouterOS 5.22 did not have this problem

honzam · Wed May 01, 2013 11:59 am

All V6rc show bad frequency in LOG

odie · Wed May 01, 2013 4:39 pm

after upgrade from rc13 no more connectivity via Vlan on RB1100AHx2
it is a shame that MT does not really test those releases !!!
i would like to get some money for testing and wasting time with this buggy stuf !
it seems that Mikrotik OS gets worse from release to release
sorry but i have enought !!

hey mikrotik why dont you try to fix all the errors first an then bring new features ????

onnoossendrijver · Wed May 01, 2013 8:16 pm

All V6rc show bad frequency in LOG

That's the frequency in kilohertz, just like the atheros driver under linux reports it.

raz · Wed May 01, 2013 9:36 pm

THere still NO OPTIMIZATION on 36 cores and you wont 2 CPU or 72 cores?

any ddos on TCP attack about 400mbit killing CCR in two seconds.. and for me I need to buy some another firewall.. becouse ccr are useless

Moar pps is what i need, and yes i would pay for that 2-3k $.

AsciiWolf · Wed May 01, 2013 10:50 pm

it seems that Mikrotik OS gets worse from release to release
sorry but i have enought !!

It's not a final release which means it's not ready for production use!

odie · Thu May 02, 2013 2:45 am

it is a release candidate for sure - but major things like Vlans should work....
and also with 5.24 we have even more sudden reboots of the boxes and no answers or fixes
and MT sell routers where you cant use a final ROS ( CCR )

osvaldotcf · Thu May 02, 2013 4:07 am

it seems that Mikrotik OS gets worse from release to release
sorry but i have enought !!
It's not a final release which means it's not ready for production use!

Yes but is not a Beta!
This is a Release Candidate! and ... 14!!!!!

NetworkPro · Thu May 02, 2013 4:43 am

So what big deal with the naming? Are you going to explain what the problem is and CONTRIBUTE or are you going to be annoyed and chatty on the forums?

Time to be annoyed has long passed. Nothing you can do about that.

Thu May 02, 2013 9:01 am

it is a release candidate for sure - but major things like Vlans should work....
and also with 5.24 we have even more sudden reboots of the boxes and no answers or fixes
and MT sell routers where you cant use a final ROS ( CCR )

We are using vlans, bonding and vlans on bonding with no issues.

Are you sure your config is right?

odie · Thu May 02, 2013 9:11 am

the config works until rc13....
seems to be not so wrong...
and also works on >100 other MT boxes

Thu May 02, 2013 9:12 am

the config works until rc13....
seems to be not so wrong...
and also works on >100 other MT boxes

please make a new topic and post "this config" that works only up to rc13. we will see what is so special about it.

odie · Thu May 02, 2013 12:59 pm

sent it to support this morning !!
including supout

zyzelis · Thu May 02, 2013 1:53 pm

after upgrade from rc13 no more connectivity via Vlan on RB1100AHx2
it is a shame that MT does not really test those releases !!!
i would like to get some money for testing and wasting time with this buggy stuf !
it seems that Mikrotik OS gets worse from release to release
sorry but i have enought !!

hey mikrotik why dont you try to fix all the errors first an then bring new features ????

Hello, i am using Vlans with RB1100AHx2 since 6.0rc11. And after upgrade from rc13 to rc14, Vlan's still working.

So i can't confirm this bug

Thu May 02, 2013 1:54 pm

Bug with SNMP not fixed.

please clarify what SNMP bug do you mean? did you contact support and report it?

napismizpravu · Thu May 02, 2013 8:07 pm

RB433UAH uptime 19h crash web-proxy (NAT forwarding off internet work)

Arnold2222 · Thu May 02, 2013 10:08 pm

after upgrade from rc13 no more connectivity via Vlan on RB1100AHx2
it is a shame that MT does not really test those releases !!!
i would like to get some money for testing and wasting time with this buggy stuf !
it seems that Mikrotik OS gets worse from release to release
sorry but i have enought !!

hey mikrotik why dont you try to fix all the errors first an then bring new features ????

Hello, i am using Vlans with RB1100AHx2 since 6.0rc11. And after upgrade from rc13 to rc14, Vlan's still working.
So i can't confirm this bug

For me also works fine. No any issues with vlans for now.

Thu May 02, 2013 11:35 pm

Hi Mikrotik staff,

I understand you are all very busy working on RouterOS. But can someone please look at Ticket#2013050266000072
It is regarding an issue with BGP pre-pends on RouterOS 6.

I also have Ticket#2013043066000404 open which may be related.

Thanks.

RafGan · Thu May 02, 2013 11:41 pm

Intel PRO/1000 (e1000?) VLAN tagging fail

2 device="02:01.0" name="82547EI Gigabit Ethernet Controller (rev: 0)"
vendor="Intel Corporation" category="Ethernet controller"
vendor-id="0x8086" device-id="0x1019"
(OnBoard Intel Server Board S875WP1-E http://downloadmirror.intel.com/19459/e ... 5WP1-E.zip)

Temporary workaround: create new bridge, put the problematic ether interface in that bridge. Add the VLANs with interface=that bridge so that bridge software would now be taking care of vlan tagging etc. instead of the NIC driver. (Note: the VLANs are not bridge ports of the mentioned bridge. They can be bridge ports of another bridge)

@MT I will send you the supout.rif file taken when problem is observed.

p.s. it may be vlan offload bug or otherwise Intel bug in my opinion !!!! (where MT could choose the driver, tweaks, they have in ROS, maybe NIC firmware or if they can - patch the damn driver (patching of NIC firmware is also a possibility))

P.S. 2: Intel PRO/1000 NIC driver in RouterOS 5.22 did not have this problem

I confirm this issue on this same RouterBoard, plus forwarding on non-vilan ports losts many of packets (input is OK). Downgrade to 6RC13 and everything works great.

Best Regards,
GaR

samsung172 · Fri May 03, 2013 12:46 am

Seems like some of the problems around vrf are partly introduced in rc14. I cannot use it due to unstable router (hanging, not showing anny error in log etc. It also had a strange, not possible to start terminal from winbox. (wrong user password) teleneted from the link in, and it was ok. Something wired are back.

Also not yet fied the vrf bug route-distinguisher

Here there should be 0.0.0.0:0

ros code

/ip route vrf
add export-route-targets=0.0.0.0:0 import-route-targets=0.0.0.0:0 \
    route-distinguisher="(unknown)" routing-mark=vrf.internet

I have also had some troble mac-telnet from later releases of v6 and 5.xx to clients in nv2 mode. Is this a common problem? switching from nv2 to 802.11 solve the problem.

Also im missing good old preamble mode, ack-timeout etc make the setup of bad links easier.

Alessio Garavano · Fri May 03, 2013 5:44 am

I found the solution to the VLAN problem on RB1100AHx2!

Upgrading an RB1100AHx2 with VLANs from rc13 to this rc14 the router was inaccesible, only connecting directly to the ether13 i can access by MAC in winbox and solve the problem changing the L2 MTU from 1598 to 1700 of the ether3 where are all the VLANs

Best regards!

BETEPAH · Fri May 03, 2013 10:34 am

Bug with SNMP not fixed.
please clarify what SNMP bug do you mean? did you contact support and report it?

No, I not contact support. I speak english very bad. Если говорят в саппорте по русски, то тогда смогу
описать проблему им очень подробно.

I have softrouter on server hardware PC. When I use 6.0rc11 graphics fine. When I upgrade to 6.0rc13:

graph-01.jpg

Fri May 03, 2013 10:35 am

Bug with SNMP not fixed.
please clarify what SNMP bug do you mean? did you contact support and report it?
No, I not contact support. I speak english very bad. Если говорят в саппорте по русски, то тогда смогу
описать проблему им очень подробно.

I have softrouter on server PC. When I use 6.0rc11 graphics fine. When I upgrade to 6.0rc13:
graph-01.jpg

You can write in Russian. Some of our support enginers will be able to understand it. Please send full description so we can test.

hapi · Fri May 03, 2013 10:38 am

CCR in bridge is fail. When it will be fixed?

Fri May 03, 2013 10:39 am

CCR in bridge is fail. When it will be fixed?

"bridge is fail" is not enough to understand your problem and fix it. please describe in more detail and send it to support.

hapi · Fri May 03, 2013 11:02 am

insert 2 interface to bridge and let the traffic flow around 200Mbit. CCR is rebooted or totaly shutdown in time. half cores is running on 100%.
it says that a lot of people. You never tried it? CCR is totally unstable.

infused · Fri May 03, 2013 2:40 pm

Not unstable... depends what your doing... high throughput yes, it seems.

hapi · Fri May 03, 2013 3:04 pm

is unstable in bridge, dot. Throughput is 200Mbit.

Fri May 03, 2013 3:11 pm

is unstable in bridge, dot. Throughput is 200Mbit.

I think you have a different understanding of word "unstable". You probably mean "slow"

We are looking at this issue

hapi · Fri May 03, 2013 3:38 pm

no!! random rebooting or totally stop responding all router.

majkel · Fri May 03, 2013 3:48 pm

Still there is no hope for TCP SYN Attacks..

u Can see CPU usage 50% and attack 160mbit?

We try everything.. The same we have with 4-5 rules..

cursixx · Sat May 04, 2013 3:51 am

anyone notice a vpn performance drop? running l2tp

with rc13 I was getting around 400+kb
with rc14 100kb or less

zervan · Sat May 04, 2013 12:11 pm

Upgrading an RB1100AHx2 with VLANs from rc13 to this rc14 the router was inaccesible, only connecting directly to the ether13 i can access by MAC in winbox and solve the problem changing the L2 MTU from 1598 to 1700 of the ether3 where are all the VLANs

I can confirm VLAN bug in RC14 on RB751G which is used as an AP with VLAN, see http://forum.mikrotik.com/viewtopic.php ... 73#p362473

After upgrade to RC14, I've lost connection to other AP on ether1, which should be tagged as VLAN11, so I've spent few hours to found out what's wrong. Here is what I've found...

I'm doing ARP ping on VLAN 11, packet should be received on the router, but they are not - see the screenshot, there is no VLAN 11 tag.
When I increase L2 MTU on the incoming VLAN port (from 1600 to any higher value), it starts to work - see the second screenshot.
I can change L2 MTU back, it is still working with any L2 MTU value.
After reboot it is not working again and it doesn't matter on the value of L2 MTU before the restart, until it is increased again.

samsung172 · Sat May 04, 2013 9:05 pm

I did not say. I had to downgrade to rc13 and all was ok.

Seems like some of the problems around vrf are partly introduced in rc14. I cannot use it due to unstable router (hanging, not showing anny error in log etc. It also had a strange, not possible to start terminal from winbox. (wrong user password) teleneted from the link in, and it was ok. Something wired are back.

Also not yet fied the vrf bug route-distinguisher

Here there should be 0.0.0.0:0

ros code
/ip route vrf
add export-route-targets=0.0.0.0:0 import-route-targets=0.0.0.0:0 \
    route-distinguisher="(unknown)" routing-mark=vrf.internet
I have also had some troble mac-telnet from later releases of v6 and 5.xx to clients in nv2 mode. Is this a common problem? switching from nv2 to 802.11 solve the problem.

Also im missing good old preamble mode, ack-timeout etc make the setup of bad links easier.

oriondotnet · Sun May 05, 2013 9:13 pm

v6rc13 working with !MYSRCADD, and stop working on rc14, mean all drop

Flags: X - disabled, I - invalid, D - dynamic
0 ;;; Drop to IP Phone
chain=forward action=drop dst-address=10.10.10.1 src-address-list=!MYSRCADD in-interface=sfp1

device on ccr 36 core

SaLMoN · Mon May 06, 2013 1:47 pm

After upgrade to rc14 I'have lost DHCP Server config on RB133 and cannot setup it.

2fast4youbr · Tue May 07, 2013 4:06 am

What happend with the HTML files of hotspot ? after I made the upgrade they vanished.

cheers

NetworkPro · Tue May 07, 2013 6:48 am

What happend with the HTML files of hotspot ? after I made the upgrade they vanished.

cheers

You will have to copy them from elsewhere with different version or from a backup. In RC14 they are mysteriously missing.

2fast4youbr · Tue May 07, 2013 6:55 am

thanks... just did.

cheers

Masyanich · Tue May 07, 2013 10:01 am

send backup by mail script, still does not work, in the v6rc11 version works fine...
any news on this?

HI i can confirm this problem.

lysy1033 · Tue May 07, 2013 2:11 pm

IPSEC over GRE doesnt work any more for me. I've got fatal INVALID-PAYLOAD-TYPE message in logs ipsec,debug.

When I downgrade packages to 6rc13 everything works just fine.

when needed i can post configs.

2fast4youbr · Tue May 07, 2013 6:46 pm

Another problem... Looks like that the vLan routes should be done automatic , but I created 3 and the inter vlan comunnication is not working at all.

From any vlan I can ping any gateway, but not the host.

cheers

oxicid · Wed May 08, 2013 2:28 am

Sorry, I posted in the wrong thread before by mistake.

Starting with version 6.0rc14 we found a bug regarding functioning SIP VoIP PBXs.
The problem occurs both with 3CX Phone System and Asterisk releases and with SIP devices.
It seems that the changes introduced at the level of NAT somehow cause the fall of the SIP call after a few minutes.
Note that all systems always operated with static port mapping service and SIP ALG disabled and have always worked properly from version 5.0 routeros.
Do Mikrotik introduced some changes at the level of Outbound NAT or Sip ALG (SIP firewall service) with 6rc14 ?

The problem was not present in the version 6.0rc13 and even in version 5.24.
In this situation it is impossible to properly operate an IP PBX behind routeros.
Please let me now something ?
For the moment we downgrade to 6rc13 and everything works fine.

NetworkPro · Wed May 08, 2013 9:19 am

try connection tracking yes rather than auto and some "ip->settings" tweaks possibly as well

tkgit · Thu May 09, 2013 2:21 am

running well on RB751u2HnD and RB433 with WDS mesh,OSPF routing
when stable version will be released?

2fast4youbr · Thu May 09, 2013 5:28 am

Hi all, sorry for the duplicate post, but maybe could be something related to this new v6rc14.

Here is the original post: viewtopic.php?f=2&t=72494

When I have the vLans working on the switch, my mac and some other devices like wireless ap wrt54g don't get an IP address, the DHCP server sends NAK twice and after that only discover & offer without sucess. But with Win7 & Linux Ubuntu works fine.

When I take out the vLans and do a simple network (without vLans & Trunk), every body get an IP ....

I dont know if this is something with my switch (Linksys srw224g4) or with the RouterOS, looking that Win7 works and no one has talken about this on the forum, I am pretty sure that could be RouterOS 6rc14, hope not, I will try to do more tests.

Does any one know a good test to isolate this problem ? I've attached Wireshark to my switch and mirror'd the ports and filtered all by bootp, should I filter other kind package ? which one ?

Would the /export help here ?

cheers!

mark · Thu May 09, 2013 7:19 pm

6to4 interfaces stopped working. A first setup will work and, after modifying it—even reverting the changes or deleting everything altogether—, it will break and no connection will be made.
… routes to those interfaces are rejected (address not reachable or the like is the error message).
With about 1'000 connections or more bridge-local will disappear all of the sudden.
»torch« doesn't catch all packets that the interface graph does
no L2TP, PPTP or VPN in general to IPv6 — yet the first IPv6-only deployments are made here, in Germany (and years ago in Asia)
packets which should be stopped by the firewall are leaking through it (!) using PPTP
It seems I cannot get proper 300 Mbaud (802.11n) speeds with my RB951G-2HnD. 130 Mbaud is as most as I can get.
Using the web-interface and/or stand-alone client to ping an address will work, but re-using that window to ping another will sometimes result in a blank results list.
A plus (+) sign in local-part of email addresses still results in the »Incorrect Email« error.

HaPe · Thu May 09, 2013 7:27 pm

MT has started to make rc15.

Stillhard · Thu May 09, 2013 8:20 pm

Tested 6.0rc15, I can't hardly use it with 100% CPU usages almost every time, back to 6.0rc11 for now

majkel · Thu May 09, 2013 9:56 pm

MT has started to make rc15.

its started 29 april ;P

semenko · Thu May 09, 2013 10:56 pm

With about 1'000 connections or more bridge-local will disappear all of the sudden.

Hey Mark -- I'm very interested in this problem.

Using an RB2011L-RM on 6.0rc14, I experience very intermittent (~weekly) loss of the router (completely unresponsive to ping, won't route traffic, etc.).

I'm wondering if this is an issue with our bridge-local interface disappearing -- for you, does the interface only disappear in the gui, or does the bridge itself seem to disappear & stop passing traffic?

HaPe · Thu May 09, 2013 11:23 pm

MT has started to make rc15.
its started 29 april ;P

Today morning there weren't any entries

mark · Fri May 10, 2013 1:10 am

With about 1'000 connections or more bridge-local will disappear all of the sudden.
Hey Mark -- I'm very interested in this problem.

Using an RB2011L-RM on 6.0rc14, I experience very intermittent (~weekly) loss of the router (completely unresponsive to ping, won't route traffic, etc.).

I'm wondering if this is an issue with our bridge-local interface disappearing -- for you, does the interface only disappear in the gui, or does the bridge itself seem to disappear & stop passing traffic?

Bridge-local disappears in GUIs and traffic won't be passed after that happens; »unknown interface« (or the like) appears where you normally could read »bridge-local«.

I've lost connection to the next hop and have logged in immediately to learn what went wrong. I can only guess what would've happened if the error went unnoticed for hours. Maybe the device will not even respond to ARP requests and therefore appear completely unresponsive. It could be something other that you stumbled upon, though. I've set up watchdogs on my devices and they're being restarted regularly—which, of course, isn't a fix…

Bergante · Sat May 11, 2013 12:39 am

Some things I would fix (rc14):

- Using BGP-VPLS (l2vpn) there's no way (or I haven't found it) to see the BGP advertisements

- The web GUI for BGP-VPLS doesn't work

- For routing->bgp it would be nice to have a l2vpn routes tab

- Logging isn't very consistent. I have noticed that, for instance, some OSPF state transitions are logged, but not all.

- IP->Neighbors->discovery: the web GUI doesn't offer the option to change default discovery settings. It can be set from the command line, though.

napismizpravu · Mon May 13, 2013 3:07 pm

error compilation all_files_6.0rc15.zip? update print log > system,error all broken package ...

samulosan · Mon May 13, 2013 3:24 pm

send backup by mail script, still does not work, in the v6rc11 version works fine...
any news on this?
HI i can confirm this problem.

Hi.Delete tls=yes of the script and enable in /tools/email.

Ex:
/tool e-mail send from= xxxxx@mail.com to=xxxx@mail.com subject=title file=backup
hope that helps

napismizpravu · Mon May 13, 2013 3:46 pm

send backup by mail script works 6.0rc15 build time May/03/2013 11:50:25

Masyanich · Mon May 13, 2013 5:05 pm

How can i download rc15?

HaPe · Mon May 13, 2013 5:18 pm

Don't use it on production. It is in alfa stage, and every day is released an update.
If you need, contact to mikrotik support.

kobuki · Tue May 14, 2013 12:00 am

RB450G, rc14, WinBox: ethernet interface Overall/Rx/Tx stats are completely empty or partially empty. On my router, ether1 is blank on every stats pages, and ether2 (2-4 ports switched together) is partially blank on the stats pages. ether1 has a changed MAC (ISP MAC restrictions), maybe this has something to do with the problem?

LennonNZ · Tue May 14, 2013 11:41 am

I don't know if its a BUG on RC14 but I have a RB2011-UAS and I have a 3G USB Stick Attached to it via MicroUSB -> USB Adapter.

When I go into PPP Client to set up the Dialer for it I don't get an option for the USB Interface.. only the serial interface.

If I go into System/Resources/USB I can see the RB2100 has identified it

I have used the Same 3G device on a RB750 and its worked fine on the USB port.

Stillhard · Tue May 14, 2013 2:35 pm

Tested 6.0rc15 (05/13/2013), RX/TX rates working, send email working, 100% cpu usage problem is gone now, but still error in usermanager web-ui (web-ui ignoring timezone setting) - in cli (/tool user-manager session print) its correctly show.

GOOD JOB MT team

Deac · Tue May 14, 2013 5:45 pm

Agree with Stillhard.
I can add that bonding still broken.

mrz · Tue May 14, 2013 5:56 pm

What exactly with bonding is broken.

Deac · Tue May 14, 2013 5:58 pm

ARP link monitoring don't work. Bonding iface immediately going to down.

mrz · Tue May 14, 2013 6:00 pm

Did you send supout file to support?

Deac · Tue May 14, 2013 6:46 pm

Done.

infused · Wed May 15, 2013 1:13 pm

I had that bonding issue on Rc5 or 6 if I remember correctly.

RC13 seems to be the most stable at the moment.

mrz · Wed May 15, 2013 4:58 pm

Some things I would fix (rc14):

- Using BGP-VPLS (l2vpn) there's no way (or I haven't found it) to see the BGP advertisements

You can't and most likely in current implementation such feature will not be added.

- The web GUI for BGP-VPLS doesn't work

Confirmed, will be fixed.

- For routing->bgp it would be nice to have a l2vpn routes tab

See the first answer

- Logging isn't very consistent. I have noticed that, for instance, some OSPF state transitions are logged, but not all.

It is a known problem that down->init, init->2-way etc are not logged, but those state changes are not so important to spam the logs. In the future we could change that.

- IP->Neighbors->discovery: the web GUI doesn't offer the option to change default discovery settings. It can be set from the command line, though.

Click on entry in the "Discovery interface" table, there will be checkbox "enabled" uncheck it and discovery will be disabled.

mark · Wed May 15, 2013 9:48 pm

sending emails using IPv6 SMTP server doesn't work
restricting access to IPv6 subnets in IP → Services doesn't work
L7-filters not available for IPv6
Accounting (the CGI script) is not available via HTTPS

Thu May 16, 2013 9:09 am

sending emails using IPv6 SMTP server doesn't work

restricting access to IPv6 subnets in IP → Services doesn't work

L7-filters not available for IPv6

Accounting (the CGI script) is not available via HTTPS

Those are not bugs, but feature requests. Many settings still don't accept IPv6. Thanks for the requests.

Bergante · Thu May 16, 2013 10:05 am

Some things I would fix (rc14):
- Using BGP-VPLS (l2vpn) there's no way (or I haven't found it) to see the BGP advertisements
You can't and most likely in current implementation such feature will not be added.

Understood. However, please do consider adding it as soon as possible after the 6.0 milestone is reached. If the l2vpn feature is widely used, that debugging information will be very important.

- Logging isn't very consistent. I have noticed that, for instance, some OSPF state transitions are logged, but not all.
It is a known problem that down->init, init->2-way etc are not logged, but those state changes are not so important to spam the logs. In the future we could change that.

I would seriously consider an overhaul of the logging policies. It's much better to "spam" the logs, allowing the user to filter what is needed and what not, rather than limiting logging. Such information can be critical to track bugs, regression changes, _and_ security incidents. You never really know what message is important (or not) until you miss it when you need it.

- IP->Neighbors->discovery: the web GUI doesn't offer the option to change default discovery settings. It can be set from the command line, though.
Click on entry in the "Discovery interface" table, there will be checkbox "enabled" uncheck it and discovery will be disabled.

[/quote]
Yes. That allows you to enable and disable the discovery on each interface.

What I mean is a feature more like the Cisco "no cdp run", that is, "I don't want the discovery running at all".

From the command line you offer a close equivalent.

[admin@rb2011-despa] /ip neighbor discovery settings> print
default: no
default-for-dynamic: no

So that you can be sure that for any newly created interfaces discovery will be indeed disabled. I don't see these default settings available from the web gui. If I relied just on the web gui I would need to disable the interfaces one by one, explicitly.

That said, something like "/ip services set discovery disabled=yes". But this is not a bug in itself, but an enhancement request.

The discrepance between the command line and the web GUI (default discovery settings) is more like a bug to me

Bergante · Thu May 16, 2013 10:49 am

I forgot, sorry for the double post:

Regarding logging, I really miss a detailed logging of router advertisement/withdrawals for the routing protocols in debug mode. Not just state transitions.

honzam · Thu May 16, 2013 2:07 pm

After upgrade from 5.25 to 6.0 = all packages all enabled.

Frekvency in log is still displayed in kHz

Thu May 16, 2013 2:30 pm

Frekvency in log is still displayed in kHz

this is a feature, because in "advanced channels" you can specify channels with decimal places. khz will give you precise

spire2z · Thu May 16, 2013 8:05 pm

I seem to have a problem where the hotspot login page wont open. All I did was upgrade from 5.24 to 6.14 and the hotspot page wont open? No setting changed.

DirectX9 · Thu May 16, 2013 8:20 pm

Hello to most respected support team. Please tell me, when can i expect final version ROS v6 for stable working CCR and all mikrotik wireless products?
All my (a lot of and expensive) custom is stopped, before the stability version will be released.
Please Answer and may be in separate topic.

rpingar · Thu May 16, 2013 8:46 pm

MT,
when Firewall will get the benefit of manycores Platform?

In an old post Normis wrote into Feb. now we are in May.....

regards
Ros

majkel · Thu May 16, 2013 11:11 pm

MT,
when Firewall will get the benefit of manycores Platform?

In an old post Normis wrote into Feb. now we are in May.....

regards
Ros

The same question.. when? .. We need some protection of our servers..

Fri May 17, 2013 8:35 am

firewall has many core support. only some specific matchers don't have multicore support, but most popular matches supprt it. just look at how load is evenly distributed through all cores on the CCR

if you have the CCR and see only one core loaded with firewall and big traffic, send us your rules.

majkel · Fri May 17, 2013 9:08 am

We have access to really big booters (min. 1Gbit/s of SSYN TCP) (2mln/packets) now and testing CCR with MANY RULES AND EXACLY ONE RULE for tested attack.. always the same.. he filtering of coures but not big attacks..

CCR give up when he must filtering about 700k/p.. (350mbit/s TCP).. many people ddosing everything and testing testing testing.. They without problem Can get acces to booters with 1mln/packets.. and our CCR is going to 100% on all cores..

Why there is no fastpatch support?

tracepath · Fri May 17, 2013 11:36 am

IP routing rules for PBR no longer work.

Fri May 17, 2013 11:39 am

We have access to really big booters (min. 1Gbit/s of SSYN TCP) (2mln/packets) now and testing CCR with MANY RULES AND EXACLY ONE RULE for tested attack.. always the same.. he filtering of coures but not big attacks..

CCR give up when he must filtering about 700k/p.. (350mbit/s TCP).. many people ddosing everything and testing testing testing.. They without problem Can get acces to booters with 1mln/packets.. and our CCR is going to 100% on all cores..

Why there is no fastpatch support?

100% on all cores means that all cores are supported for firewall and working. What you are asking is for more speed in general. Did you enable SYN cookie in RouterOS? What kind of rule is that exactly?

majkel · Fri May 17, 2013 12:46 pm

ah ah ah.. I wrote to You this half year ago.. yes i have SYN cookies enable..

I didnt say that they are no support for all cores for firewall.. but there is no good optimization for firewall..

Fri May 17, 2013 12:50 pm

ah ah ah.. I wrote to You this half year ago.. yes i have SYN cookies enable..

I didnt say that they are no support for all cores for firewall.. but there is no good optimization for firewall..

Depends on your rule. Is that a Tarpit rule, or Drop? Based on what criteria?

half year ago we didn't have CCR. today the software is much improved.

majkel · Fri May 17, 2013 1:34 pm

all the rules all droping packets.. but i`m testing the attacks with one rule.. like droping packets on port 80 TCP.. and its the same when i Have 50 different rules.. always max packets is 600-700k/s and I get 100% CPU..

I bought my CCR in November 2012 so I dont understand

why U dont have it..

I dont blame You that CCR is not good for DDOS now.. But I do not see to be better in future

nothing changed from many releases.. the same power for firewall.. from v6

One company from Poland install some distrubution of Linux on CCR, and filtering 2mln/packets of TCP with 50% CPU.. So I thing there is a posibillity to do that the same on RouterOS..

przent · Fri May 17, 2013 3:17 pm

One company from Poland install some distrubution of Linux on CCR, and filtering 2mln/packets of TCP with 50% CPU.. So I thing there is a posibillity to do that the same on RouterOS..

Post a link/contact please. I`m very curious about this. Thanks.

hngjared · Fri May 17, 2013 8:46 pm

The TX rates in the host and active lists for the hotspot are not shown with 6.0rc11 - 14.

Any fixes available or is there a setting that has been added?

Thanks!

djras · Sat May 18, 2013 12:43 pm

After upgrade to v6.0rc14 on RB2011LS I have a problem w Graphing:
Tools/Graphing->ResouseGraphs types=unknown for CPU, HDD, Memory

djras · Sat May 18, 2013 12:52 pm

Downgrade from v6.0rc14 set default interfaces mac address. But I have specific MAC address. The Backup configuration in .rsc doesn't restore these data as was earlier in ROSv5.20
============
/interface ethernet
set 0 arp=enabled auto-negotiation=yes bandwidth=unlimited/unlimited \
disabled=no full-duplex=yes l2mtu=1524 mac-address=D4:CA:6D:4D:21:D5 \
master-port=none mtu=1500 name=ether1 speed=100Mbps
.................
==========

bohorquezf · Sun May 19, 2013 1:08 am

May check the availability of Mikrotik page?, We are training sessions and Examination for MTCNA, and you can enter the page Mikrotik.com

djras · Sun May 19, 2013 9:04 am

The v6.0rc14 version on CCR1036-12G-4S slowly opens the Queues page (SimpleQueue) in WinBox of 5-10 seconds for 50 or more records and in addition open columns. In prev versions everything opened instantly. In this version delays began after addition of additional columns UpAvgRate DownAvgRate UpAvgPack DownAvgPack.

shrek777 · Sun May 19, 2013 3:58 pm

Website is having problems. http://www.mikrotik.com
I want to sent supout for ccr.

We have following services.
2 CCRs same configuration on each with same client quantity and same traffic, there are in different part of country, but same problems.

EOIP Tunnel with 100 mbps traffic
PPP server
Radius client (queues are added dynamically).

Problems on both CCRs EOIP tunnels are disconnecting very often, 20-30 times per day.

one of them are restarting 3-4 times a day and second 1 time in two days.

before on this points with same configuration was RB1100 and it was working about 1 year without any reboot, eoip disconnect or something else.
Also on both point we have CCR1036-12G-4S and both of them have only one CPU with 85% load, others are below 10% and cpu TEMP is 55c

mark · Sun May 19, 2013 4:50 pm

Even with low load »bridge-local« keeps disappearing, resulting in my RouterBoard becoming unusable.

Furthermore, after that a »link local« doesn't get assigned to »bridge-local« even if I removed all link local addresses and rebooted the router. I believe it is the same bug that has been reported with v6rc12 here: http://forum.mikrotik.com/viewtopic.php ... wn#p362379

RouterOS-v6rc14-IPv6-total-failure.png

Bergante · Mon May 20, 2013 11:28 am

There's a regression problem.

On 5.25 I can add a simple queue and bind it to an interface. Example:

[admin@OmniTik-casa] /queue simple> add name=queue-guest max-limit=512K/128K interface=ether2-master-local

This doesn't work on 6.0rc14. I can't add the "interface" clause to a simple queue.

This is a simple rate limit applied to a bridging interface, nothing complex. And 6.0 rc14 doesn't work as the documentation states.

Has the simple queuing been changed on 6.0 or is this a bug?

Mon May 20, 2013 1:57 pm

There's a regression problem.

On 5.25 I can add a simple queue and bind it to an interface. Example:

[admin@OmniTik-casa] /queue simple> add name=queue-guest max-limit=512K/128K interface=ether2-master-local

This doesn't work on 6.0rc14. I can't add the "interface" clause to a simple queue.

This is a simple rate limit applied to a bridging interface, nothing complex. And 6.0 rc14 doesn't work as the documentation states.

Has the simple queuing been changed on 6.0 or is this a bug?

"target" option now can be used - specify interface there.

Bergante · Mon May 20, 2013 2:13 pm

"target" option now can be used - specify interface there.

Thank you

Documentation should be updated, then.

Also, there is something confusing in the documentation.

In case "target" is an interface, which direction is considered "upload" and which one is considered "download"? I would rather use "transmit" and "receive" provided, of course, "transmit" means "transmitted through the interface" and "received" means "received through the interface".

Or maybe "ingress" and "egress". But "upload" and "download" can be applied to the router or the network connected to the router interface.

Chupaka · Mon May 20, 2013 2:37 pm

it's actually 'upload' of a 'target', i.e. traffic received on the interface

rower · Mon May 20, 2013 2:56 pm

it's actually 'upload' of a 'target', i.e. traffic received on the interface

does not feel clear nor simple.

Mon May 20, 2013 3:06 pm

it's actually 'upload' of a 'target', i.e. traffic received on the interface
does not feel clear nor simple.

It is "Target upload" not "router upload". Therefore, these options are from perspective of the target. If you want to limit your users's upload capability, set "target upload".

honzam · Mon May 20, 2013 5:54 pm

6.0 final to download...
http://www.mikrotik.com/download

GhostWorks · Mon May 20, 2013 6:19 pm

6.0 just been released ... Fantastic as I just ordered a CCR1036

Bergante · Mon May 20, 2013 7:01 pm

it's actually 'upload' of a 'target', i.e. traffic received on the interface
does not feel clear nor simple.
It is "Target upload" not "router upload". Therefore, these options are from perspective of the target. If you want to limit your users's upload capability, set "target upload".

I'm still confused about "target" and "dst".

And it seems that using "target" I end up with a symmetric queue (I was trying something like 1M/128k) while using "dst" (again, with an interface as an argument) now it works.

So, this configuration allows me to limit output bandwith (on the interface) to around 1 Mbps, input to around 128 kbps.

# may/20/2013 17:59:08 by RouterOS 6.0rc14
# software id = PAXV-BRR1
#
/queue simple
add burst-limit=2M/192k burst-time=3s/3s dst=ether2-master-local limit-at=1M/128k max-limit=1M/128k name=guest-limit target=""

Still, upload and download are poor names. If you put them on an upstream interface, they work. If you apply the queue to a downstream interface it's the opposite, which is counter-intuitive. Or at least in this particular case :/

keraia · Mon May 20, 2013 7:32 pm

I just installed version 6.0.
First impressions compared to 5.24: (mipsbe on RB435G)
Memory usage +100%
Disk usage -50%
CPU usage about +30% for the same bandwidth (this measurement is approximate as I cannot replicate exact same bandwidth/number of connections)

I would go back to 5.24 but since we are going to move to 6.x eventually, I will give it a try for some time.

onnoossendrijver · Mon May 20, 2013 10:07 pm

I just installed version 6.0.
First impressions compared to 5.24: (mipsbe on RB435G)
Memory usage +100%
Disk usage -50%
CPU usage about +30% for the same bandwidth (this measurement is approximate as I cannot replicate exact same bandwidth/number of connections)

I would go back to 5.24 but since we are going to move to 6.x eventually, I will give it a try for some time.

Hm.. I have a different experience..
Got 2 RB600A, same configuration, same test..
6.0 vs 5.24:
memory: about the same
Disk: -50% (like you)
CPU: -30% (opposite of you)

I think I like 6.0

gnic · Tue May 21, 2013 5:40 am

Have problems when I try to run the torch in a Simple Queue

*RB1100AHx2

gnic · Tue May 21, 2013 6:34 am

All Simple Queues are gone, closed the winbox and opened again several times and still not displaying no queue.

Tue May 21, 2013 8:23 am

All Simple Queues are gone, closed the winbox and opened again several times and still not displaying no queue.

what version did you have before you upgraded?

gnic · Tue May 21, 2013 8:41 am

All Simple Queues are gone, closed the winbox and opened again several times and still not displaying no queue.

what version did you have before you upgraded?

upgrade v5.21 to v6.0 , Had this same problem reported after the queues are gone --> fixed queues - router could become unresponsive when configuring queues;)

I returned to 5.21 everything normal now

Tue May 21, 2013 8:44 am

Always make backups, and if possible, upgrade to v5.25 first, then to v6. This is not required, but could avoid some unknown issues.

afindlay · Wed May 22, 2013 5:52 pm

I had just got IPv6 router advertisement working properly on 5.17 and I upgraded to 6.0rc14.
No router advertisements any more. In fact it had lost all the config under "/ipv6 nd" when seen from the command-line. The config could be seen from the web interface, but it didn't work. I re-entered the config from the command-line but still nothing.

Downgraded to 5.25 and all is OK again...

Andrew

wa4zlw · Sat Jul 20, 2013 9:44 pm

does anyone have any more informaiton on linux on CCR? who got it going, etc> any links etc?

thanks leon

Ivoshiee · Sat Jul 20, 2013 10:25 pm

does anyone have any more informaiton on linux on CCR? who got it going, etc> any links etc?

thanks leon

It sounds something for the OpenWRT. Maybe they are developing it.

IAmStaka · Thu Aug 08, 2013 1:45 am

meanwhile i can't handle SSYN floods

help <3

ros code

ros code

Who is online