v6.2 released

Fri Aug 02, 2013 2:18 pm

What's new in 6.2 (2013-Aug-02 10:37):

*) console - added "on-error" argument to ':do' command that is executed
if command raises error;
*) hotspot - fixed chap error after failed http-chap login (broken in v6.1);
*) console - added new ':return' command that interrupts execution of script
and passes argument as return value if script was called as function;
*) routerboot - fixed upgrade from RouterOS (could fail on some units);
*) userman - fixed payment gateway response notify processing;
*) console - resolved issue with 'from-pool' propery in '/ipv6 address';
*) console - array value syntax in expressions '{1;2;3;4}' now can
specify values with word keys as '{a=1;b=2}';
*) console - added 'verbose' argument to '/import' command that enables
line-by-line script import. By default import whole script at once
and don't print it, as it was in version 6.0;
*) console - ':global', ':local' and ':set' commands have new parameter 'do'
that allows assigning block of commands to the variable;
*) console - global variables now are common to all users and are
available to all users with at least "read,write,test,policy" policy;
*) console - fixed parameter passing to scripts. Script parameters can
be accessed without declaring them with ':local' and ':global' commands.
For backwards compatibility global variables are first looked up in
script parametrs;
*) console - '$var 1 2 a="a" b="b"' syntax for passing parameters to commands
stored in a variable. Parameters are accessed as '$1' '$2' '$a' '$b';
*) ipsec - fixed peer address matching;
*) ups - query smart ups capabilities before issuing any commands;
*) improved CCR responsiveness on other interfaces when one interface is under attack;
*) sms tool - added sim-pin setting;
*) dhcp server - framed routes are now also added to the server routing table;
*) dhcpv6 server - added binding-script option;
*) proxy - allow multiple src-address for ipv4 and ipv6;
*) eoip,gre tunnels could occasionally crash multicore router;
*) fixed bug - sometimes some types of interfaces would stop working;
*) ipsec sometimes could crash kernel on CCR;
*) connection tracking sometimes could crash kernel on CCR;
*) ppp,pptp,l2tp,sstp - added default-route-distance parameter;
*) scep - "/cert scep ra" merged into "/cert scep client" without saving ra config;
*) ipsec - fix phase1 autonegotiation on little endian platforms;
*) pppoe server - allow service with empty service-name to accept all pppoe clients;
*) lcd - current-screen option is replaced with "/lcd show" command
*) lcd - current-interface option is replaced with "/lcd interface display" command
*) graphing - make graphs stable on ppp & ovpn interfaces;
*) www, hotspot - fixed problem when www service stopped responding on high load;
*) winbox, webfig: allow to enter space in the text fields;
*) webfig - fixed configuration of VPLS & routing filters;
*) lcd - added option for enabling or disabling the touch screen;
*) lcd - added options for screen switching;
*) lcd - up to 10 non-physical interfaces can now be added to the lcd;
*) lcd - all interface graph screen can now be customized from /lcd interface page;
*) backup - changed default backup file name to <id>-<YY><MM><DD>-<HHMM>.backup
for file browsers to sort them properly;
*) webfig - it did not work in Opera;
*) webfig - made terminal work again;
*) winbox - added ability to fully set up traffic generator in winbox;
*) trafficgen - allow ranges for ip addrs and udp ports;
*) trafficgen - add tcp header support;
*) queue simple - fixed bug - actual queue order sometimes was wrong;
*) queue simple - queue is not invalid when at least one of target interfaces is up;
*) fixed crash when setting master-port on AR8327 switch chips;
*) fixed addresslist - dynamic entries sometimes would still
show up even afther being timed out;
*) added /ip settings allow-hw-fast-path setting to control AR8327N hardware ipv4 fast path;
*) vrrp - allow more than one vrrp on interface;

http://www.mikrotik.com/download

Campano · Fri Aug 02, 2013 3:01 pm

Tremendous Changelog!
go ahead for the best release and kill more bugs!

jajajaja

SoKaR · Fri Aug 02, 2013 3:43 pm

Oh hell... just updated and now same sh** as 6.1 ! What are u doing guys?

Again crashing, again kernel panic and again it seem's to be an IPSec problem?!

6.2.png

janisk · Fri Aug 02, 2013 3:45 pm

SoKaR need more information as all known IPSEC problems where resolved in 6.2

cheeze · Fri Aug 02, 2013 3:47 pm

Well done Mikrotik peeps...

Here's hoping that this is a good base to start from and continue onto version 6 of ROS.

If I may ask. Most of the releases/work since ROS 6 came out have been maintainence/bug fixes. That I think is good. I think that should continue.

However I am curious if more things are going to be added. Like echo functionality in BFD, or BGP labeled unicast, or FRR capabilities, or the myriads of other requests.

Not looking for an ETA or anything....just curious if the next release cycle from Mikrotik is leaning to be more bug fixes, or if it will be more features....

As always, yalls are awesome

Fri Aug 02, 2013 3:48 pm

new routing stuff is going into v7, which contains the long promised "new routing engine"

SoKaR · Fri Aug 02, 2013 3:50 pm

SoKaR need more information as all known IPSEC problems where resolved in 6.2

I just sent u an E-Mail (support@mikrotik.com) few seconds ago with two supout files.

Fri Aug 02, 2013 3:53 pm

SoKaR you have a wild mix of different packages, each from another v6.2 build. re-download them ALL and upload exactly the same set you have now, so that they don't make conflicts. Your file:

        gps               6.2      Jul/17/2013 15:40:41
             ntp               6.2      Jul/17/2013 15:40:41
             ppp               6.2      Aug/02/2013 10:37:21
             ups               6.2      Jul/17/2013 15:40:41
            dhcp               6.2      Aug/02/2013 10:37:21
            ipv6               6.2      Aug/02/2013 10:37:21
            mpls               6.2      Aug/02/2013 10:37:21
         hotspot               6.2      Aug/02/2013 10:37:21
        security               6.2      Aug/02/2013 10:37:21
  advanced-tools               6.2      Aug/02/2013 10:37:21
           calea               6.2      Jul/17/2013 15:40:41
   routeros-tile               6.2      Aug/02/2013 10:37:21
    user-manager               6.2      Jul/17/2013 15:40:41
          system               6.2      Aug/02/2013 10:37:21
        openflow               6.2      Jul/17/2013 15:40:41
        wireless               6.2      Aug/02/2013 10:37:21
       multicast               6.2      Jul/17/2013 15:40:41
         routing               6.2      Aug/02/2013 10:37:21

To ALL: please upgrade ALL packages, not just system, or just some of them, especially if you used a pre-release v6.2 build.

SoKaR · Fri Aug 02, 2013 3:56 pm

Okay will give a second try now :-S

Fri Aug 02, 2013 3:57 pm

Okay will give a second try now :-S

We also found an IPsec issue that could be related to your specific config, so problem still exists on our site too. We will make a special fix for you to test.

mhoungbo · Fri Aug 02, 2013 4:00 pm

What's new in 6.2 (2013-Aug-02 10:37):

http://www.mikrotik.com/download

no news about Qualcomm USB CDMA Modem support ?

[Ticket#2013061366000068]

Fri Aug 02, 2013 4:00 pm

I would also suggest to install only packages that you need, it is funny to see wireless package on CCR , or routing package on stand alone home router.
At very least keep them disabled.

SoKaR · Fri Aug 02, 2013 4:10 pm

Okay will give a second try now :-S
We also found an IPsec issue that could be related to your specific config, so problem still exists on our site too. We will make a special fix for you to test.

Yes all packages are the same release now and IPSec will still crash. Do you need again a supout file?

Numenori · Fri Aug 02, 2013 4:17 pm

What about problems with invalid ip from forum topics like this - http://forum.mikrotik.com/viewtopic.php?f=2&t=73591 or this - http://forum.mikrotik.com/viewtopic.php?f=1&t=74339
I don't see any information about solve this problem in the changelog.

Fri Aug 02, 2013 4:23 pm

What about problems with invalid ip from forum topics like this - http://forum.mikrotik.com/viewtopic.php?f=2&t=73591 or this - http://forum.mikrotik.com/viewtopic.php?f=1&t=74339
I don't see any information about solve this problem in the changelog.

you must contact support to report issues, not all topics of this community forum are cheked for reproducable issues. You must send supout.rif files and instructions to support@mikrotik.com

Numenori · Fri Aug 02, 2013 4:30 pm

What about problems with invalid ip from forum topics like this - http://forum.mikrotik.com/viewtopic.php?f=2&t=73591 or this - http://forum.mikrotik.com/viewtopic.php?f=1&t=74339
I don't see any information about solve this problem in the changelog.
you must contact support to report issues, not all topics of this community forum are cheked for reproducable issues. You must send supout.rif files and instructions to support@mikrotik.com

Yes, support know about this problem and one of the last thing i heard from he was

Hello,

We hope to introduce fix in the next RouterOS version (6.2).

Regards,
Sergejs

So, fixed or not?

stmx38 · Fri Aug 02, 2013 4:31 pm

I have early installer 6.2 pre.
And now I tried to upgrade the system via Winbox -
System --> Package --> Check-for-updates --> Check-for-updates --> System is already up to date

It compare only version number but not release date ?

fnkysknky · Fri Aug 02, 2013 4:33 pm

added /ip settings allow-hw-fast-path setting to control AR8327N hardware ipv4 fast path;

This sounds very interesting, can anyone elaborate on what is fastpath'd?

StubArea51 · Fri Aug 02, 2013 6:30 pm

Sounds similar to Cisco's CEF or the earlier fast switching. ASIC based hardware forwarding? It would be nice

skot · Fri Aug 02, 2013 8:56 pm

The new console / scripting changes are very nice! Some really good additions, especially the first one for error handling. Thank you MT.

ghi000 · Fri Aug 02, 2013 8:58 pm

IPSec now working between MIPSBE to MIPSBE and MIPSBE to PPC+viceversa. I will try tomorrow all platforms, except mipsle. Good job, Mikrotik. If i'll have an address, i'll send you some beer, Normis

Too bad i bought today a v5 license for x86

upgradeable to 7.x

just because IPSec was not working.

angboontiong · Fri Aug 02, 2013 10:36 pm

V6 have nothing to do with wireless ???

brointhemix · Fri Aug 02, 2013 10:43 pm

Hello!

Looks like you have broken the sniffer app with v6.2.

I have two Ethernet ports SPAN mirrored to a CCR1036. With v6.1 it sniffed just fine, after upgrading to v6.2 I get nothing (the sniff packets/connections/hosts/procotols tables are empty). Could you look into that?

Here is my sniffer setup:

/tool sniffer
set file-limit=0KiB filter-direction=rx memory-limit=10240KiB memory-scroll=no \
    only-headers=yes

BR

steen · Fri Aug 02, 2013 11:32 pm

Hello Folks!

I test installing some devices, first our PPC RB333 used as massive vpn router, and office landscape RB433 router and some RB2011UAS-2HnD devices.

Whole vpn network came up.
Office landscape came up.
And RB2011 device came up.

No visible problems.

If this fallout well and reports look nice in forum, we will go for more devices, also business critical IPSEC network devices and wireless infrastructure.

I will be back with more reports

Ulypka · Fri Aug 02, 2013 11:35 pm

[Ticket#2013070166000356]
the problem remains
memory leaks and PPTP session hang

theprism · Fri Aug 02, 2013 11:47 pm

Hi guys,

In v6.1 the following IPSEC problem existed:
Windows 7 <- L2TP/IPSEC -> Mikrotik 6.1 stopped working.
Passing over L2TP and when at IPSEC it compains in logging about payload and triming.

Was this fixed?

Thanks,
the_prism

wa4zlw · Sat Aug 03, 2013 2:31 am

HI Normis....

Disappointed that I didn't see any routing fixes for BGP and OSPF plus other items that really need to be fixed.

WHy are you now talking about v7 when v6 is so new? I think that routing fixes need to be added ASAP and better testing done. folks have bought the CCRs and really can't use them due to problems reported.

THoughts?

thanks leon

efaden · Sat Aug 03, 2013 2:40 am

Still have unknown switch type on switch #1 on the RB2011.

stmx38 · Sat Aug 03, 2013 8:11 am

IPSec:
Mikrotik - Mikrotik = ok
Mikrotik - ASA = ok
Mikrotik - NetScreen = ok

Thank you!

infused · Sat Aug 03, 2013 8:18 am

hmm, to try or not to try...

Any specific steps to upgrade ccr from rc13?

Sat Aug 03, 2013 8:24 am

new routing stuff is going into v7, which contains the long promised "new routing engine"

Normunds,

We will hold a celebration when v7 with "new routing engine" is available

If you guys next need testers let me know, we are using most of the more advanced features, have a myriad of gear from other vendors to test against and seem to have a knack for finding bugs.

mymicha · Sat Aug 03, 2013 11:34 am

Hey guys,

I did update all of my equipment to 6.2 yesterday. UPS is now working again. Yeah! It's connected via USB to a RB751U-2HnD box. It did work after applying update 6.2 and upgrading the RouterBoard firmware to 3.09. In former releases the APC 1500 Smart UPS wasn't recognized. It wasn't possible to select the USB port within the configuration dialog.

Maybe I found one bug within the proxy module. After updating one RB450G from version 6.1 to 6.2 the local proxy was disabled and it's configuration was reseted. I tried to reconfigure the module again, but I failed. I'm using a parent proxy setting on this device. After entering the IP address from the parent proxy I receive the error: Couldn't change Web Proxy Settings - parent-proxy must have the same address family as src-address (6)

It would be nice to get this confirmed or any advice is welcome!

Thanks for your great work.
Michael

tromideh · Sat Aug 03, 2013 11:42 am

Hi
Web Proxy in Transparent Mode Error

mymicha · Sat Aug 03, 2013 12:11 pm

@tromideh: Thanks for confirming my experience

TKITFrank · Sat Aug 03, 2013 12:29 pm

Hi,

"ip proxy access" seems messed up.
As soon as you enter src-address to a rule it stops working.

This is from my setup. 48 allows my internal network and 49 drops the rest. But 48 never gets hit. Worked prior to upgrade to 6.2
I removed the dst-port also but to no avail. If i disable 49 everythings working.
I also tried to set rule 48 to be rule 1 but that did not help at all.

48 src-address=192.168.255.0/24 dst-port=80 action=allow hits=0
49 dst-port="" action=deny hits=472

So going to rollback to 6.1.....

Update
Rollback helped!!

48 src-address=192.168.255.0/24 dst-port=80 action=allow hits=572
49 dst-port="" action=deny hits=0

mymicha · Sat Aug 03, 2013 12:33 pm

Maybe it has something to do with this change:

*) proxy - allow multiple src-address for ipv4 and ipv6;

TKITFrank · Sat Aug 03, 2013 12:38 pm

Maybe it has something to do with this change:

*) proxy - allow multiple src-address for ipv4 and ipv6;

Yep sounds like the cause

TKITFrank · Sat Aug 03, 2013 1:31 pm

new routing stuff is going into v7, which contains the long promised "new routing engine"

I was hoping for v6 release of this.
So there is no hope for it to be a test release for v6? Or better a backport to v6?

Sat Aug 03, 2013 1:41 pm

new routing stuff is going into v7, which contains the long promised "new routing engine"
I was hoping for v6 release of this.
So there is no hope for it to be a test release for v6? Or better a backport to v6?

I think that point releases should only be bug fixes and minor changes.

Major stuff e.g. Wireless Controller, and "new routing" should go in to v7

npero · Sat Aug 03, 2013 2:14 pm

Proxy, access list if you define source address don't working any more.
I have couple rules based on source address with deny redirect I been very surprised

when some IP just passed without redirecting or deny.

MrYan · Sat Aug 03, 2013 2:41 pm

Looks like the added default-route-distance parameter doesn't set the distance for IPv6 default routes:

[admin@mikrotik] /lcd> /ip route print
Flags: X - disabled, A - active, D - dynamic, C - connect, S - static, r - rip, b - bgp, o - ospf, m - mme, B - blackhole, U - unreachable, P - prohibit
 #      DST-ADDRESS        PREF-SRC        GATEWAY            DISTANCE
 2  DS  0.0.0.0/0                          10.112.112.172           11
 3 ADC  10.112.112.172/32  10.64.64.124    ppp-out1                  0

[admin@mikrotik] /lcd> /ipv6 route print
Flags: X - disabled, A - active, D - dynamic, C - connect, S - static, r - rip, o - ospf, b - bgp, U - unreachable
 #      DST-ADDRESS              GATEWAY                  DISTANCE
 1  DS  ::/0                     ppp-out1                        1

arn80 · Sat Aug 03, 2013 3:46 pm

Hello,

just upgraded 6.1 to 6.2 on SXT 2nD r2. If I run Winbox (checked on version 2.2.18), connect and then click "Quick set", then the quick set window appears but just one second later winbox get closed completely. This wasn't present on 6.1. It is 100% reproducible on my PC.

best regards

Remek

specialist · Sat Aug 03, 2013 5:09 pm

dyndns script stop working on 6.2.. allways said no update needed....

theprism · Sat Aug 03, 2013 5:29 pm

dyndns script stop working on 6.2.. allways said no update needed....

Do we have any temp. patch for this? Anyone found a successful workaround?

Thanks,
The_prism

JanezFord · Sat Aug 03, 2013 6:08 pm

Maybe I found one bug within the proxy module. After updating one RB450G from version 6.1 to 6.2 the local proxy was disabled and it's configuration was reseted. I tried to reconfigure the module again, but I failed. I'm using a parent proxy setting on this device. After entering the IP address from the parent proxy I receive the error: Couldn't change Web Proxy Settings - parent-proxy must have the same address family as src-address (6)

It would be nice to get this confirmed or any advice is welcome!

Thanks for your great work.
Michael

Same here ... tested on RB450G and RB800 - same result ... proxy disabled by default, does not accept any IP for parent proxy ...

damn ...

JF

athlonxp78 · Sat Aug 03, 2013 9:02 pm

Still have unknown switch type on switch #1 on the RB2011.

Same here... this affect the stability or is a simple display mismatch?

ghi000 · Sat Aug 03, 2013 10:47 pm

IPSec is dead on CCR, unusable.

It works with some MIPSBE, but not with all, all routers with the same configuration.

It worked with some of them for about 5 minutes, then stopped. With some other didn't work at all, and with some, as i type, the connection is still UP, but i will not wait, i'll put back the 1100AHx2 with v5.25.

There's no problem with MIPSBE <-> MIPSBE, at least i did not noticed anything, but we need the CCR.

I waited for this day to test it, but it is the same dissapointment.

ghi000 · Sat Aug 03, 2013 10:50 pm

And some connections are working only they are initiated from the other network, and not the CCR's. It is simply ignoring the routes, or the requests, i don't know.

I'll send back the CCR and not buy it. It was for testing.

andaks · Sat Aug 03, 2013 11:07 pm

I just newly buy RB1100x2AH. When i upgrade routerboard to v6.2, i capture error through serial when rebooting as below:

Rebooting...
Stopping services...
Oops: Exception in kernel mode, sig: 5 [#1]
SMP NR_CPUS=2 RB1120
NIP: e1520be0 LR: e157e44c CTR: e157e3e4
REGS: dcfa1cb0 TRAP: 0700 Not tainted (3.3.5-smp)
MSR: 00029000 <CE,EE,ME> CR: 24082022 XER: 20000000
TASK = dd67dea0[782] 'shif' THREAD: dcfa0000 CPU: 1
GPR00: e159677c dcfa1d60 dd67dea0 00000001 e157e458 00000000 00000000 ffffb1e1
GPR08: e157e458 e1520000 e1521fb8 802ac5c8 24082022 10018c40 03715280 08000000
GPR16: c0000a00 00000014 3fffffff 03814350 00000015 de4fc40c 00000000 deff2000
GPR24: 00008914 deff23c0 00000002 00000002 deff2000 deff2b60 df8b7000 deff23c0
NIP [e1520be0] fast_path_ipv4_hw_register+0x74/0x8c [packet_hook@0xe1520000]
LR [e157e44c] ath8327_flush_hosts+0x1cc/0x410 [ath8327@0xe157a000]
Call Trace:
[dcfa1d60] [deff2b60] 0xdeff2b60 (unreliable)
[dcfa1d70] [e159677c] gfar_get_stats+0x3748/0x45e4 [gianfar@0xe1593000]
[dcfa1da0] [802248cc] __dev_close_many+0x9c/0xf0
[dcfa1dc0] [8022494c] __dev_close+0x2c/0x4c
[dcfa1de0] [802287dc] __dev_change_flags+0x134/0x198
[dcfa1e00] [802288f4] dev_change_flags+0x1c/0x68
[dcfa1e20] [80283214] devinet_ioctl+0x638/0x89c
[dcfa1e80] [8020fa14] sock_ioctl+0x68/0x2e4
[dcfa1ea0] [800c0f34] do_vfs_ioctl+0x9c/0x754
[dcfa1f10] [800c1628] sys_ioctl+0x3c/0x74
[dcfa1f40] [8000d0d8] ret_from_syscall+0x0/0x3c
--- Exception: c01 at 0x30069ae4
LR = 0x300af944
Instruction dump:
4dbe0020 4bfffd28 2f830001 419e0020 81291fb8 7c844a78 7c840034 5484d97e
68840001 0f040000 4e800020 81091fb8 <0f080000> 814a0004 90891fb8 2f8a0000
---[ end trace 00f119be836a6c5e ]---

Trace/breakpoint trap
panicSaver_ioctl, cmd: 0x1, arg: 0x0
panicSaver_ioctl, cmd: 0x2, arg: 0x100123f0
copy_to_user 3072 -> 0
panicSaver_ioctl, cmd: 0x3, arg: 0x0
panicSaver: terminating
Restarting system.

RouterBOOT booter 3.02

RouterBoard 1100AHx2

CPU frequency: 1066 MHz
Memory size: 2048 MiB
NAND size: 128 MiB

Press any key within 2 seconds to enter setup..

loading kernel from nand... OK
setting up elf image... OK
jumping to kernel code
Starting...
Starting services...
MikroTik 6.2

current-firmware is 3.02. Forgot to check before upgrade. Downgrade to ver 5.25 no error during rebooting.

Why this happen? Or still bug exist in ver 6.2?

ntwkedout · Sun Aug 04, 2013 12:35 am

IPSEC Mikrotik to Cisco ASA

Working fine on 5.25, Upgraded to 6.2, Became very excited thinking everything was working fine, Then Noticed SIP phones were down, Ran a torch on WAN interface to find that return traffic was being processed under the WAN IP and not returning network ip. Almost like the NAT accept rule was not working right.

Downgraded back to 5.25 everything is back up and working. Other than that it seemed from the IPSEC debug in the log that it was working fine.

ntwkedout · Sun Aug 04, 2013 12:44 am

IPSEC Mikrotik to Cisco ASA

Working fine on 5.25, Upgraded to 6.2, Became very excited thinking everything was working fine, Then Noticed SIP phones were down.

Downgraded back to 5.25 everything is back up and working. Other than that it seemed from the IPSEC debug in the log that it was working fine.

steen · Sun Aug 04, 2013 3:35 pm

Hello Folks!

So far everything is working, but one thing, switch1 type is unknown in RB2011UAS-2HnD, see picture.

I do not know if it affect something yet.

Does IPSEC work now again, that is a showstopper for us if not.
Also does 6.2 have any known problem with wireleas nv2 traffic, that is also a showstopper.

We currently have a majority of our devices on 5.22, we stopped due to lot of fuzz around higher versions, 5.22 is so far rock solid at our place.

ghi000 · Sun Aug 04, 2013 3:47 pm

same thing here

no more new routeros in prod environment

except this router, which will not work with v5.xx

IPSec working only between mipsbe devices. Cannot use the CCR with MIPSBE and PPC.

scooter · Sun Aug 04, 2013 4:09 pm

Hi,
i have RB2011UAS-2HnD-IN, firmware 3.08, after upgrade to ver. 6.2 my device/system losing firewall rule and dhcp server is deactiveting (avery reboot or 10 minutes). I did faktory reset and same result.
After downgrad to 6.1 is firewal and dhcp ok.

Some one with same experience?

scooter · Sun Aug 04, 2013 4:14 pm

Hi,
i have RB2011UAS-2HnD-IN, firmware 3.08, after upgrade to ver. 6.2 my device/system losing firewall rule and dhcp server is deactiveting (avery reboot or 10 minutes). I did faktory reset and same result.
After downgrad to 6.1 is firewal and dhcp ok.

Some one with same experience?

yottabit · Sun Aug 04, 2013 4:24 pm

Running RB751G on v6.1 with firmware 3.02. Upgrade to v6.2 works, but trying to upgrade attempt to firmware 3.09 causes the RB to hang on reboot. Ethernet activity lights flash but not the Wi-Fi activity light, can't login, and the Ethernet switch doesn't function. Hard power cycling recovers the RB but back to the 3.02 firmware.

No problems on the 911G upgrading to its latest firmware, 3.07.

Nighthawke · Sun Aug 04, 2013 6:11 pm

OK, we're back to a stable (no crashes) release, mine has not crashed after 24 hours running on a production router.

Now you guys just need to concentrate on getting the IPsec fixed up.

Oh, BTW, the LED's are working now!

v3locity · Sun Aug 04, 2013 10:55 pm

Hello Folks!

So far everything is working, but one thing, switch1 type is unknown in RB2011UAS-2HnD, see picture.

I do not know if it affect something yet.

Does IPSEC work now again, that is a showstopper for us if not.
Also does 6.2 have any known problem with wireleas nv2 traffic, that is also a showstopper.

We currently have a majority of our devices on 5.22, we stopped due to lot of fuzz around higher versions, 5.22 is so far rock solid at our place.

Hi,

Seeing the same problem (switch1 unknown) on RB951G and RB750G after upgrading from 6.1 to 6.2.

This affects the VLAN switching functionality. I have set ether1 as master port and ether2-5 as slaves and IP address on vlan1 on ether1. After upgrading to 6.2, IP address on vlan1 cannot be accessed but switch remains working.

Had to hard reset to downgrade back to 6.1 to restore full functionality.

Also, I think interface graphs are not working. I rebooted my router and the interface graphs were erased for the most recent time periods, same problem as with 6.1.

JanezFord · Mon Aug 05, 2013 12:04 am

OK, we're back to a stable (no crashes) release, mine has not crashed after 24 hours running on a production router.

Now you guys just need to concentrate on getting the IPsec fixed up.

Oh, BTW, the LED's are working now!

my rb800 crashed after 20h of operation ... I also lost all interface graphing data ....

edit: after looking at terminal window I see:

jan/02/1970 01:01:08 system,error,critical Out of memory condition was detected

this is a first on this unit... it usually consumes about 90megs for normal operation ...

JF

payday · Mon Aug 05, 2013 12:42 am

Also, I think interface graphs are not working. I rebooted my router and the interface graphs were erased for the most recent time periods, same problem as with 6.1.

My interface graphs are erased every time I reboot my RB2011UAS-2HnD.

What is more interesting: after reboot static IP address assigned to bridge-local changes its assignment to wlan1. Once I changed it back to bridge-local and after reboot it was again bound to wlan1.

ntwkedout · Mon Aug 05, 2013 1:33 am

IPSEC

Mikrotik RB493G to cisco ASA 5520

v5.25 everything is working Upgraded to 6.2 and SIP phones stopped registering but all other traffic seems to be working. Downgraded back to 5.25 everything started working again.

I have been troubleshooting after hours and cant see why the phones stopped working. Everything looks like it should be working on 6.2 but it does not. Stuck on this one. Going back 5.25 for now.

xds · Mon Aug 05, 2013 3:30 am

steen, ghi000,
In my RB2011UAS-2HnD-IN, all is well!

swich.jpg

-------------------------------------
payday,
My interface graphs are erased every time I reboot my RB2011UAS-2HnD...
-------------------------------------------------------

I confirm, on my RB2011UAS-2HnD-IN, the same happens,
interface graphs are erased every time I reboot!

sisal · Mon Aug 05, 2013 11:39 am

Hi

With update to v6.2, SSH key length has been limited to 1024bytes:

[admin@KN302250-RB2011L-954] > /file print file="ac_pubkey_xy.txt"
[admin@KN302250-RB2011L-954] > /file set "ac_pubkey_xy.txt" contents="ssh-dss AAAAB3NzaC..."
[admin@KN302250-RB2011L-954] > /user ssh-keys import public-key-file="ac_pubkey_xy.txt" user=admin
key size is limited by 1024 bytes (see FIPS186-2)!

On v6.1 (and 5.x) you supported keys with more than 1024 bytes (like FIPS186-3).

sis

ghi000 · Mon Aug 05, 2013 12:14 pm

well, congratulations. you can now use it as a simple home router

mymicha · Mon Aug 05, 2013 1:37 pm

well, congratulations. you can now use it as a simple home router

efaden · Mon Aug 05, 2013 2:13 pm

I can also confirm that on all three of my RB2011s the interface graphs are erased on each reboot.

xds · Mon Aug 05, 2013 3:07 pm

Hi Normis,
could you tell what is the reason of the script: https://www.changeip.com/mikrotik/ - which works great in ROS v6.1, now no longer works in ROS v6.2.
It has something to do with the changes described in the "changelog", or just the script to change something or add?

Mon Aug 05, 2013 3:10 pm

Hi Normis,
could you tell what is the reason of the script: https://www.changeip.com/mikrotik/ - which works great in ROS v6.1, now no longer works in ROS v6.2.
It has something to do with the changes described in the "changelog", or just the script to change something or add?

it should work. what fails for you ?

xds · Mon Aug 05, 2013 3:15 pm

Hi Normis,
could you tell what is the reason of the script: https://www.changeip.com/mikrotik/ - which works great in ROS v6.1, now no longer works in ROS v6.2.
It has something to do with the changes described in the "changelog", or just the script to change something or add?
it should work. what fails for you ?

Always says "DDNS: No changes necessary." and does not change the IP address!
(when turn the system on ROS v6.1 everything works well)

Mon Aug 05, 2013 3:17 pm

Type of uninitialized variable is "nil" not "nothing". Change it and it should work.

xds · Mon Aug 05, 2013 3:20 pm

Type of uninitialized variable is "nil" not "nothing". Change it and it should work.

Thank you!
Now I just try!

nkm · Mon Aug 05, 2013 3:26 pm

i have this problem too and access list to proxy does not work!

Hey guys,

I did update all of my equipment to 6.2 yesterday. UPS is now working again. Yeah! It's connected via USB to a RB751U-2HnD box. It did work after applying update 6.2 and upgrading the RouterBoard firmware to 3.09. In former releases the APC 1500 Smart UPS wasn't recognized. It wasn't possible to select the USB port within the configuration dialog.

Maybe I found one bug within the proxy module. After updating one RB450G from version 6.1 to 6.2 the local proxy was disabled and it's configuration was reseted. I tried to reconfigure the module again, but I failed. I'm using a parent proxy setting on this device. After entering the IP address from the parent proxy I receive the error: Couldn't change Web Proxy Settings - parent-proxy must have the same address family as src-address (6)

It would be nice to get this confirmed or any advice is welcome!

Thanks for your great work.
Michael

xds · Mon Aug 05, 2013 3:39 pm

Type of uninitialized variable is "nil" not "nothing". Change it and it should work.
Thank you!
Now I just try!

Ohh great, everything works just fine!

Thank you very much!

Normis, mrz,
Could you add a script with "nil" and/or change it to "nil", script: http://wiki.mikrotik.com/wiki/Dynamic_D ... angeIP.com
and add another "The script below is RouterOS 6.2 Tested!"
For some it will be useful, especially since it is "wiki mikrotik"?!

Thank you again mrz!

mcskiller · Mon Aug 05, 2013 4:16 pm

Hi ppl, i have recently upgraded my rb750 from v 5.21 to v6.2 and i noted more ram consumption at same config. Its a bug? Or is properly right in v 6.2

Chupaka · Mon Aug 05, 2013 5:09 pm

Type of uninitialized variable is "nil" not "nothing". Change it and it should work.

wow... could you please explain this?..

[admin@TestPlace] > {:global uuu22; :put [ :typeof $uuu22 ]; }
nil
[admin@TestPlace] > {:local uuu22; :put [ :typeof $uuu22 ]; }      
nothing
[admin@TestPlace] > {:local uuu224; :put [ :typeof $uuu224 ]; }
nothing

UPD: while in v5.14 it shows this:

[admin@MikroTik] > {:global uuu22; :put [ :typeof $uuu22 ]; }
nothing
[admin@MikroTik] > {:local uuu22; :put [ :typeof $uuu22 ]; }
nothing
[admin@MikroTik] > {:local uuu224; :put [ :typeof $uuu224 ]; }
nothing

specialist · Mon Aug 05, 2013 6:12 pm

I just newly buy RB1100x2AH. When i upgrade routerboard to v6.2, i capture error through serial when rebooting as below:
Rebooting...
Stopping services...
Oops: Exception in kernel mode, sig: 5 [#1]
SMP NR_CPUS=2 RB1120
NIP: e1520be0 LR: e157e44c CTR: e157e3e4
REGS: dcfa1cb0 TRAP: 0700 Not tainted (3.3.5-smp)
MSR: 00029000 <CE,EE,ME> CR: 24082022 XER: 20000000
TASK = dd67dea0[782] 'shif' THREAD: dcfa0000 CPU: 1
GPR00: e159677c dcfa1d60 dd67dea0 00000001 e157e458 00000000 00000000 ffffb1e1
GPR08: e157e458 e1520000 e1521fb8 802ac5c8 24082022 10018c40 03715280 08000000
GPR16: c0000a00 00000014 3fffffff 03814350 00000015 de4fc40c 00000000 deff2000
GPR24: 00008914 deff23c0 00000002 00000002 deff2000 deff2b60 df8b7000 deff23c0
NIP [e1520be0] fast_path_ipv4_hw_register+0x74/0x8c [packet_hook@0xe1520000]
LR [e157e44c] ath8327_flush_hosts+0x1cc/0x410 [ath8327@0xe157a000]
Call Trace:
[dcfa1d60] [deff2b60] 0xdeff2b60 (unreliable)
[dcfa1d70] [e159677c] gfar_get_stats+0x3748/0x45e4 [gianfar@0xe1593000]
[dcfa1da0] [802248cc] __dev_close_many+0x9c/0xf0
[dcfa1dc0] [8022494c] __dev_close+0x2c/0x4c
[dcfa1de0] [802287dc] __dev_change_flags+0x134/0x198
[dcfa1e00] [802288f4] dev_change_flags+0x1c/0x68
[dcfa1e20] [80283214] devinet_ioctl+0x638/0x89c
[dcfa1e80] [8020fa14] sock_ioctl+0x68/0x2e4
[dcfa1ea0] [800c0f34] do_vfs_ioctl+0x9c/0x754
[dcfa1f10] [800c1628] sys_ioctl+0x3c/0x74
[dcfa1f40] [8000d0d8] ret_from_syscall+0x0/0x3c
--- Exception: c01 at 0x30069ae4
LR = 0x300af944
Instruction dump:
4dbe0020 4bfffd28 2f830001 419e0020 81291fb8 7c844a78 7c840034 5484d97e
68840001 0f040000 4e800020 81091fb8 <0f080000> 814a0004 90891fb8 2f8a0000
---[ end trace 00f119be836a6c5e ]---

Trace/breakpoint trap
panicSaver_ioctl, cmd: 0x1, arg: 0x0
panicSaver_ioctl, cmd: 0x2, arg: 0x100123f0
copy_to_user 3072 -> 0
panicSaver_ioctl, cmd: 0x3, arg: 0x0
panicSaver: terminating
Restarting system.

RouterBOOT booter 3.02

RouterBoard 1100AHx2

CPU frequency: 1066 MHz
Memory size: 2048 MiB
NAND size: 128 MiB

Press any key within 2 seconds to enter setup..

loading kernel from nand... OK
setting up elf image... OK
jumping to kernel code
Starting...
Starting services...
MikroTik 6.2
current-firmware is 3.02. Forgot to check before upgrade. Downgrade to ver 5.25 no error during rebooting.

Why this happen? Or still bug exist in ver 6.2?

http://routerboard.com/fwf/158/ar9344-3.08.fwf

3.08 is the last firmware for routerboot, try update it.

nagaju · Mon Aug 05, 2013 7:12 pm

Hello,

with ros 6.2, mangling web proxy packets with dscp doesn't work.

ros code

/ip firewall mangle print

0   ;;; proxy hit
     chain=output action=mark-packet new-packet-mark=proxy-hit passthrough=no 
     out-interface=br-lan dscp=4

after downgrading to 6.1 it works again.

xds · Mon Aug 05, 2013 8:15 pm

-------------------------------------
payday,
My interface graphs are erased every time I reboot my RB2011UAS-2HnD...
-------------------------------------------------------

I confirm, on my RB2011UAS-2HnD-IN, the same happens,
interface graphs are erased every time I reboot!

I think another bug,

In Firewall, NAT, Mangle... (Column "Bytes", "Packets") erased every time, when I reboot the router!

firewall.jpg

Someone is it?

Basdno · Mon Aug 05, 2013 10:23 pm

Maybe a stupid question, but just making sure I have understood it.

I am simply going to use the CCR as an ordinary Gateway between me and my BackboneISP, with normal firewalling rules and stuff. Nothing very advanced yet.

Then I wonder a bit about all these IPSEC troubles and stuff, they are only happening when you use the CCR as one of the endpoints right??!

All traffic just "flowing throu" isnt affected right? Even if customers should run their own tunnels with IPSEC or similar throu between f.ex. different Office branches?

Thx in advance for any feedback!

efaden · Mon Aug 05, 2013 10:28 pm

Maybe a stupid question, but just making sure I have understood it.

I am simply going to use the CCR as an ordinary Gateway between me and my BackboneISP, with normal firewalling rules and stuff. Nothing very advanced yet.

Then I wonder a bit about all these IPSEC troubles and stuff, they are only happening when you use the CCR as one of the endpoints right??!

All traffic just "flowing throu" isnt affected right? Even if customers should run their own tunnels with IPSEC or similar throu between f.ex. different Office branches?

Thx in advance for any feedback!

I just downgraded to 6.1.

In 6.2 Quick Settings was slow as hell on the web interface and crashed winbox, the CPU usage was really high, erased interface graphs on reboot, and some scripts broke... oh... and my switch #1 was unknown type. ... Seems much more stable on 6.1. Hopefully you can workout all of these bugs.

Basdno · Mon Aug 05, 2013 11:56 pm

Maybe a stupid question, but just making sure I have understood it.

I am simply going to use the CCR as an ordinary Gateway between me and my BackboneISP, with normal firewalling rules and stuff. Nothing very advanced yet.

Then I wonder a bit about all these IPSEC troubles and stuff, they are only happening when you use the CCR as one of the endpoints right??!

All traffic just "flowing throu" isnt affected right? Even if customers should run their own tunnels with IPSEC or similar throu between f.ex. different Office branches?

Thx in advance for any feedback!
I just downgraded to 6.1.

In 6.2 Quick Settings was slow as hell on the web interface and crashed winbox, the CPU usage was really high, erased interface graphs on reboot, and some scripts broke... oh... and my switch #1 was unknown type. ... Seems much more stable on 6.1. Hopefully you can workout all of these bugs.

So, v6.1 is safe to use in a production environment even the long changelog on 6.2? Or do I really have to wait for v6.3 before I finally can start using my CCR-1036 that has been sitting on my shelf already for over half a year waiting?

efaden · Tue Aug 06, 2013 12:02 am

Maybe a stupid question, but just making sure I have understood it.

I am simply going to use the CCR as an ordinary Gateway between me and my BackboneISP, with normal firewalling rules and stuff. Nothing very advanced yet.

Then I wonder a bit about all these IPSEC troubles and stuff, they are only happening when you use the CCR as one of the endpoints right??!

All traffic just "flowing throu" isnt affected right? Even if customers should run their own tunnels with IPSEC or similar throu between f.ex. different Office branches?

Thx in advance for any feedback!
I just downgraded to 6.1.

In 6.2 Quick Settings was slow as hell on the web interface and crashed winbox, the CPU usage was really high, erased interface graphs on reboot, and some scripts broke... oh... and my switch #1 was unknown type. ... Seems much more stable on 6.1. Hopefully you can workout all of these bugs.

So, v6.1 is safe to use in a production environment even the long changelog on 6.2? Or do I really have to wait for v6.3 before I finally can start using my CCR-1036 that has been sitting on my shelf already for over half a year waiting?

6.1 has been good for me. But 6.2 was a mess.

Sent from my SCH-I545 using Tapatalk 2

JanezFord · Tue Aug 06, 2013 12:06 am

I can also confirm that on all three of my RB2011s the interface graphs are erased on each reboot.

You also can add rb800, rb711, rb450g, rb751g, rb751u to the list ... all of these loose interface graphs with every reboot...

Chupaka · Tue Aug 06, 2013 12:18 am

I think another bug,
In Firewall, NAT, Mangle... (Column "Bytes", "Packets") erased every time, when I reboot the router!
Someone is it?

fat troll

xds · Tue Aug 06, 2013 12:48 am

I think another bug,
In Firewall, NAT, Mangle... (Column "Bytes", "Packets") erased every time, when I reboot the router!
Someone is it?
fat troll

Do not quite understand what you mean?

payday · Tue Aug 06, 2013 12:53 am

I guess there is something wrong with variables in v6.2.

One of my scripts stopped working after upgrade from v6.1 to v6.2 so I started to search for the reason and I found this strange behavior:

[admin@MT] /ip route>  {get 1 gateway;  :local test;  :put $test}

[admin@MT] /ip route>  {get [find comment=Main] gateway;  :global test;  :put $test}

[admin@MT] /ip route>  {get [find comment=Main] gateway;  :local test;  :put $test}
192.168.0.0/24
[admin@MT] /ip route>  {get [find comment=Main] gateway;  :local test;  :put [:typeof $test]} 
ip-prefix

Result of first three commands should be empty line. Example above shows that when uninitialized local variable is created after using find, this variable contains rubbish - it is not empty. This problem exists also in other menu levels (not only in ip route). If I create more local unitialized variables after find then every of them contains rubbish:

[admin@MT] /ip route>
{get [find comment=Main] gateway
:local test1
:local test2
:local test3
:put $test1
:put $test2
:put $test3} 

192.168.0.0/24
192.168.0.1
bridge-local

duvi · Tue Aug 06, 2013 12:56 am

I think another bug,
In Firewall, NAT, Mangle... (Column "Bytes", "Packets") erased every time, when I reboot the router!
Someone is it?
fat troll
Do not quite understand what you mean?

He means that this has been the behaviour of RouterOs like forever (:

jose · Tue Aug 06, 2013 1:07 am

Hey guys,

I did update all of my equipment to 6.2 yesterday. UPS is now working again. Yeah! It's connected via USB to a RB751U-2HnD box. It did work after applying update 6.2 and upgrading the RouterBoard firmware to 3.09. In former releases the APC 1500 Smart UPS wasn't recognized. It wasn't possible to select the USB port within the configuration dialog.

Maybe I found one bug within the proxy module. After updating one RB450G from version 6.1 to 6.2 the local proxy was disabled and it's configuration was reseted. I tried to reconfigure the module again, but I failed. I'm using a parent proxy setting on this device. After entering the IP address from the parent proxy I receive the error: Couldn't change Web Proxy Settings - parent-proxy must have the same address family as src-address (6)

It would be nice to get this confirmed or any advice is welcome!

Thanks for your great work.
Michael

Michael

Exactly the same here.
Have you found a solution yet?

My router is CCR1016.

Thanks

xds · Tue Aug 06, 2013 1:30 am

fat troll
Do not quite understand what you mean?
He means that this has been the behaviour of RouterOs like forever (:

duvi
Thank you for your explanation, but sometimes you need something that will give you a little smile, and see how many people on the forum will turn my attention to what I am describing, until now it is only Chupaka!
Thank you Chupaka!

___
(sorry for some extra post)

mymicha · Tue Aug 06, 2013 7:17 am

Michael

Exactly the same here.
Have you found a solution yet?

My router is CCR1016.

Thanks

@Jose: No, I did not. I downgraded every device (to 6.1) which needs this parent proxy functionality.

Tue Aug 06, 2013 10:55 am

Type of uninitialized variable is "nil" not "nothing". Change it and it should work.
wow... could you please explain this?..
Code: Select all
[admin@TestPlace] > {:global uuu22; :put [ :typeof $uuu22 ]; }
nil
[admin@TestPlace] > {:local uuu22; :put [ :typeof $uuu22 ]; }      
nothing
[admin@TestPlace] > {:local uuu224; :put [ :typeof $uuu224 ]; }
nothing

global variables are nil, this is due to new changes in scripting. In next version we will make so that all of them will return nothing to avoid confusion.

Tue Aug 06, 2013 12:23 pm

I guess there is something wrong with variables in v6.2.

Thank you for the report. Will be fixed in next release.

Chupaka · Tue Aug 06, 2013 2:38 pm

In next version we will make so that all of them will return nothing

...and everything will be like for years before

jfdumont · Tue Aug 06, 2013 4:36 pm

Hello,

on Model 1100Hx2 both switch are Type unknown

regards

JF

yarr · Tue Aug 06, 2013 7:15 pm

It looks like global variables in scripts are not working.
They are not created and cannot be read in scripts.

armandfumal · Tue Aug 06, 2013 9:36 pm

no problem here

sw.JPG

Russian · Wed Aug 07, 2013 1:19 am

Hi,

Thank you for removing the explicit output on importing scripts.

I have though found 2 more problems with V6.2 regarding the scripts I use.

1.) When you have a variable and want to add to it with itself in brackets then then it only takes the last entered text.
Ex:
:local Tester
:set Tester "Hi there";
:set Tester ([$Tester].", I am having problems with ROS.");
:set Tester ([$Tester]." Can you maybe ");
:set Tester ([$Tester]." Assist me ");

:put $Tester

Result : Assist me
instead of : Hi there, I am having problems with ROS. Can you maybe Assist me

2.) I am trying to download a page from an http server (This worked in all versions before 6.2 since fetch was introduced)
Ex:
/tool fetch url="http://172.16.1.201/mikrotik/config/scripts.php" dst-path=WugUpdateScript.rsc keep-result=yes << This works but, now I use that same url in a variable and it fails
/tool fetch url=$updateStr dst-path=WugUpdateScript.rsc keep-result=yes
The error message: failure: invalid URL protocol

Could you please assist with this

Keep up the good work, and it was good to see your guys here in SA at MUM SA 2013, can not wait for the next event!!

Russian

Chupaka · Wed Aug 07, 2013 3:06 am

1.) When you have a variable and want to add to it with itself in brackets then then it only takes the last entered text.
Ex:
:local Tester
:set Tester "Hi there";
:set Tester ([$Tester].", I am having problems with ROS.");
:set Tester ([$Tester]." Can you maybe ");
:set Tester ([$Tester]." Assist me ");

:put $Tester

Result : Assist me
instead of : Hi there, I am having problems with ROS. Can you maybe Assist me

yeah, it worked before, but I don't think you should use square brackets at all. just remove them around the variable - and it will work as earlier

2.) I am trying to download a page from an http server (This worked in all versions before 6.2 since fetch was introduced)
Ex:
/tool fetch url="http://172.16.1.201/mikrotik/config/scripts.php" dst-path=WugUpdateScript.rsc keep-result=yes << This works but, now I use that same url in a variable and it fails
/tool fetch url=$updateStr dst-path=WugUpdateScript.rsc keep-result=yes
The error message: failure: invalid URL protocol

seems like it works fine for me:

[admin@TestPlace] > {:local myurl "http://infolan.by/"; /tool fetch url=$myurl keep-result=no ; }                 
  status: finished

[admin@TestPlace] >

c0d3rSh3ll · Wed Aug 07, 2013 3:24 am

6.2 need more ram memory? like a windows OS.
someone with the same problem?

efaden · Wed Aug 07, 2013 4:28 am

To update on the graphs with disappearing data. ... When I downgraded to 6.1 all my old data was still there... my graph just has a big hole in it for the time I was running 6.2. See the image.

wispwest · Wed Aug 07, 2013 5:09 am

I was hoping for wireless improvements as well. I'm struggling lately with NV2 on my PtP links. Data rates are never stable, latency / jitter is all over the place. I miss Ntreme 1ms pings, I am fine with 2-3ms with NV2, if it was stable, but its not. I'll have 10-14ms spikes about ever 20 pings or so.

Also throughput seems to have declined. Back on 5.XX I used to have 300/300 rates all day. Now their struggling by 270/243, sometimes even 240/240. With 300/300 I could push ~180-200mb speedtest UDP. Now I'm hitting 150-160 with 6.1

garlicbulb · Wed Aug 07, 2013 9:08 am

/ping 2620

e000::8001 works

/tool ping 2620

e000::8001 says "invalid value for argument address"

janisk · Wed Aug 07, 2013 9:19 am

'/tool ping-speed' does not support ipv6 addresses

in CLI and API ping "lives" in the root menu.

infused · Wed Aug 07, 2013 9:30 am

Is there a beta for 6.3 for testing with some of these issues fixed? (queue graphs etc)

janisk · Wed Aug 07, 2013 9:41 am

it is not yet available for public.

leonset · Wed Aug 07, 2013 10:35 am

Hi,

VRRP AH auth seems to be broken in v6.2 as it was in v6.0. It worked ok with v6.1, but I can't use that release due to IPSec problems solved in v6.2. I've had to switch to "simple" authentication and/or vrrp v3 protocol.

Please, check vrrp again for next release.
Thanks!

Russian · Wed Aug 07, 2013 11:11 am

yeah, it worked before, but I don't think you should use square brackets at all. just remove them around the variable - and it will work as earlier

Hi, Yes I have removed the brackets, not sure why I had it like that in the first place, but thought that I would just mention this

2.) I am trying to download a page from an http server (This worked in all versions before 6.2 since fetch was introduced)
Ex:
/tool fetch url="http://172.16.1.201/mikrotik/config/scripts.php" dst-path=WugUpdateScript.rsc keep-result=yes << This works but, now I use that same url in a variable and it fails
/tool fetch url=$updateStr dst-path=WugUpdateScript.rsc keep-result=yes
The error message: failure: invalid URL protocol
seems like it works fine for me:
Code: Select all
[admin@TestPlace] > {:local myurl "http://infolan.by/"; /tool fetch url=$myurl keep-result=no ; }                 
  status: finished

[admin@TestPlace] > 

Hi,

Sorry, It seems I did not do proper testing on that one, it was not the url itself, but for some reason my username is being added infront of the url. This is a very strange problem.This script works on 6.1 and below copied and pasted into script and run, but on 6.2 it adds my username infront of the url.

:local updateserver "172.16.x.x";
:local WugScriptVersions;

##1#		
:if ([/sys scr find name=WugScriptVersions]!="") do={
}
###

:local autoUpdateStr ("http://$updateserver/scripts.php");
:put $autoUpdateStr;
:local updateStr
##2#
:for i from=0 to=([:len $autoUpdateStr]-1) do={
    :local tmp [:pick $autoUpdateStr $i];
    :if ($tmp=" ") do={
	:set updateStr ($updateStr."%20");
    } else={
        :set updateStr ($updateStr.$tmp);
    }	
}
###
:put $updateStr;
:global Update
:set Update $updateStr;

/tool fetch url=$updateStr dst-path=WugUpdateScript.rsc keep-result=yes

Result:

[Russian@GC3_Main] > sys scr run script1 
http://172.16.x.x/scripts.php
Russianhttp://172.16.x.x/scripts.php
failure: invalid URL protocol

The problem is even stranger, if I remove the ##1# block, that does nothing, it works. Same If I remove the ##2# block. I can not see that there should be any problem or conflicts though. I have tested this on a few boards now and get the same result.

Any idee what is going on here?

Russian · Wed Aug 07, 2013 12:04 pm

Hi,

Sorry, It seems I did not do proper testing on that one, it was not the url itself, but for some reason my username is being added infront of the url. This is a very strange problem.This script works on 6.1 and below copied and pasted into script and run, but on 6.2 it adds my username infront of the url.
Code: Select all
:local updateserver "172.16.x.x";
:local WugScriptVersions;

##1#		
:if ([/sys scr find name=WugScriptVersions]!="") do={
}
###

:local autoUpdateStr ("http://$updateserver/scripts.php");
:put $autoUpdateStr;
:local updateStr
##2#
:for i from=0 to=([:len $autoUpdateStr]-1) do={
    :local tmp [:pick $autoUpdateStr $i];
    :if ($tmp=" ") do={
	:set updateStr ($updateStr."%20");
    } else={
        :set updateStr ($updateStr.$tmp);
    }	
}
###
:put $updateStr;
:global Update
:set Update $updateStr;

/tool fetch url=$updateStr dst-path=WugUpdateScript.rsc keep-result=yes
Result:
Code: Select all
[Russian@GC3_Main] > sys scr run script1 
http://172.16.x.x/scripts.php
Russianhttp://172.16.x.x/scripts.php
failure: invalid URL protocol
The problem is even stranger, if I remove the ##1# block, that does nothing, it works. Same If I remove the ##2# block. I can not see that there should be any problem or conflicts though. I have tested this on a few boards now and get the same result.

Any idee what is going on here?

Ok so I still have no idee why this is happening but made a work around that will sort this for now, just extra code to process though.

:set updateStr [:pick $updateStr [:find $updateStr "http"] [:len $updateStr]]

Added this line before every fetch command where I used $updateStr, and now it works.

O, just to add the global var I used was just so that I could see the url outcome by running it as an schedule, It has been removed

Russian · Wed Aug 07, 2013 1:45 pm

Hi,

Ok I found another problem, just when I though I had it all sorted.

:foreach addy in=[/ip ad find] do={[:put $addy]}

The above code should give you the position result of what it finds, but on 6.2 the result is blank/nil/nothing < one of those

I also see that with long scripts, as from 6.1 I am not able to edit the scripts from the winbox window after opening them, it does not allow any input. Will this be fixed or will this stay this way?

dada · Wed Aug 07, 2013 2:47 pm

It looks like global variables in scripts are not working.
They are not created and cannot be read in scripts.

you probably hit the access policy problem. The scripts must have read,write,test,policy rights to be able to access-modify global vars - when started from scheduler or via winbox. When the script is started from a CLI it looks like it has more rights than shows policy in script definition.

payday · Wed Aug 07, 2013 3:37 pm

The problem is even stranger, if I remove the ##1# block, that does nothing, it works. Same If I remove the ##2# block. I can not see that there should be any problem or conflicts though. I have tested this on a few boards now and get the same result.
Any idee what is going on here?

Read this: http://forum.mikrotik.com/viewtopic.php ... 50#p380610
and this: http://forum.mikrotik.com/viewtopic.php ... 50#p380689

Temporary solution is to use global variables and remove them after use.

dzieva · Wed Aug 07, 2013 4:56 pm

I was hoping for wireless improvements as well. I'm struggling lately with NV2 on my PtP links. Data rates are never stable, latency / jitter is all over the place. I miss Ntreme 1ms pings, I am fine with 2-3ms with NV2, if it was stable, but its not. I'll have 10-14ms spikes about ever 20 pings or so.

Also throughput seems to have declined. Back on 5.XX I used to have 300/300 rates all day. Now their struggling by 270/243, sometimes even 240/240. With 300/300 I could push ~180-200mb speedtest UDP. Now I'm hitting 150-160 with 6.1

This also happens to me, I've tried everything, changed rb, antenna configuration, but not resolved.
Mikrotik should review nv2, believe they will lose their wireless system, because it is getting far behind the competition.

Att.
Dzieva.

friction · Wed Aug 07, 2013 5:05 pm

I can confirm global variables are somewhat broken, even with all policies assigned:

Scripts:

 2   name="add-global" owner="admin" policy=ftp,reboot,read,write,policy,test,winbox,password,sniff,sensitive,api last-started=aug/07/2013 15:07:38 
     run-count=1 source=
       :global arandomglobal "This is the value of the global variable"
       :put $arandomglobal

 3   name="read-global" owner="admin" policy=ftp,reboot,read,write,policy,test,winbox,password,sniff,sensitive,api last-started=aug/07/2013 15:10:26 
     run-count=1 source=
       :global arandomglobal
       :put [:typeof $arandomglobal]
       :put "READ: $arandomglobal"

[admin@nxxs-cpe311we01] /system script> run add-global 
This is the value of the global variable
[admin@nxxs-cpe311we01] /system script> /environment print 
arandomglobal="This is the value of the global variable"
[admin@nxxs-cpe311we01] /system script> run read-global  
nil
READ:

In this case the line ":global arandomglobal" re-sets the global variable to a null/nil-value

adaptation:

 3   name="read-global" owner="admin" policy=ftp,reboot,read,write,policy,test,winbox,password,sniff,sensitive,api last-started=aug/07/2013 15:15:59 
     run-count=3 source=
       #:global arandomglobal
       :put [:typeof $arandomglobal]
       :put "READ: $arandomglobal"

[admin@nxxs-cpe311we01] /system script> run read-global 
nothing
READ:

In this case the global variable does not exist for the script...

In command line it all works well:

[admin@nxxs-cpe311we01] /system script> run add-global          
This is the value of the global variable
[admin@nxxs-cpe311we01] /system script> /environment print 
arandomglobal="This is the value of the global variable"
[admin@nxxs-cpe311we01] /system script> :put $arandomglobal
This is the value of the global variable
[admin@nxxs-cpe311we01] /system script>

Wed Aug 07, 2013 5:10 pm

Works for me. Are you running final build (not pre-release)? Because this problem was fixed in one of the last builds

lyma · Wed Aug 07, 2013 6:38 pm

new routing stuff is going into v7, which contains the long promised "new routing engine"
Normunds,

We will hold a celebration when v7 with "new routing engine" is available

If you guys next need testers let me know, we are using most of the more advanced features, have a myriad of gear from other vendors to test against and seem to have a knack for finding bugs.

+1

friction · Wed Aug 07, 2013 6:55 pm

Darn, you're right, it was not the final 6.2... Works perfectly now. +1 to you sir. (hours lost...)

Works for me. Are you running final build (not pre-release)? Because this problem was fixed in one of the last builds

armandfumal · Wed Aug 07, 2013 8:31 pm

to avoid this type of problem, I think MKT should not naming pre release with the same name as final build
the only way is to check the build date...no build number exist...

Wed Aug 07, 2013 10:45 pm

to avoid this type of problem, I think MKT should not naming pre release with the same name as final build
the only way is to check the build date...no build number exist...

Well, in fairness, MT never "announced" 6.2. They simply had it there for "bleeding edge enthusiasts" to test (OR if prompted by support) - and if you fit into that category, you BETTER upgrade to the final as soon as it's announced.

honzam · Wed Aug 07, 2013 11:27 pm

to avoid this type of problem, I think MKT should not naming pre release with the same name as final build
the only way is to check the build date...no build number exist...

+1

PhilipLykov · Thu Aug 08, 2013 1:10 am

After upgrade to 6.2 on my RB751U-2HnD all WiFi troubles come back. I have two macbook pro, two iPads and two iPhones. At least two iPads experience the problem: every few minutes WiFi is freezing on it, it appears as connected but no replies from MKTK goes back to iPad. Preamble is Long and Cipher is AES CCM as was before.

yarr · Thu Aug 08, 2013 7:11 am

I've reinstalled 6.2 but I'm still having issues with global variables.
For example http://wiki.mikrotik.com/wiki/Dynamic_D ... _No-IP_DNS is not working even after global variable manual creation, it looks like it can work with global variables at all.
Please advice.

infused · Thu Aug 08, 2013 7:41 am

So, we have a CCR with RC13 still, as it's generally stable for us. However, lost the queues in winbox recently. Needs a reboot.

Would you lot recommend updating it to 6.2?

We have:
40 odd filters and 40 odd nats
15 queues

That's all we use it for, but it is a production router.

raz · Thu Aug 08, 2013 10:08 am

So, we have a CCR with RC13 still, as it's generally stable for us. However, lost the queues in winbox recently. Needs a reboot.

Would you lot recommend updating it to 6.2?

We have:
40 odd filters and 40 odd nats
15 queues

That's all we use it for, but it is a production router.

Here a RB1100 with 60 Firewall Rules, 10 Queues, including VPN, WLAN AP (3rd Party HW), works perfect.

On my CCR i try the 6.2 this Weekend, including BGP Fullfeed.

Thu Aug 08, 2013 2:09 pm

After upgrade to 6.2 on my RB751U-2HnD all WiFi troubles come back. I have two macbook pro, two iPads and two iPhones. At least two iPads experience the problem: every few minutes WiFi is freezing on it, it appears as connected but no replies from MKTK goes back to iPad. Preamble is Long and Cipher is AES CCM as was before.

What RouterOS version you used before v6.2?

cbrown · Thu Aug 08, 2013 3:25 pm

Also throughput seems to have declined. Back on 5.XX I used to have 300/300 rates all day. Now their struggling by 270/243, sometimes even 240/240. With 300/300 I could push ~180-200mb speedtest UDP. Now I'm hitting 150-160 with 6.1

I have also experienced this.

Thu Aug 08, 2013 3:42 pm

Also throughput seems to have declined. Back on 5.XX I used to have 300/300 rates all day. Now their struggling by 270/243, sometimes even 240/240. With 300/300 I could push ~180-200mb speedtest UDP. Now I'm hitting 150-160 with 6.1
I have also experienced this.

Which exact RouterOS versions you tried?

cbrown · Thu Aug 08, 2013 4:04 pm

I have a few test links that I play with regularly and have had about every v6 release candidate, 6.0, 6.1, a bunch of pre-release 6.2 builds, and the current 6.2 build.

The current 6.2 build does seem to be slightly better, but still not quite where it use to be.

Rudios · Thu Aug 08, 2013 4:31 pm

I think I found a typo on the login.html file of the hotspot files.
The file starts with html, head and title and inside the head there are 4 meta assignments.
I believe that the last meta entry is ended with a faulty double quote character, just before the closing arrow '>'
If the hotspot login page is shown the style piece is shown on the browser.
If I change the last double quote to the one used on all the other places in the file, the style showing in browser is gone.
I hope my explanation is clear.

PhilipLykov · Thu Aug 08, 2013 7:45 pm

After upgrade to 6.2 on my RB751U-2HnD all WiFi troubles come back. I have two macbook pro, two iPads and two iPhones. At least two iPads experience the problem: every few minutes WiFi is freezing on it, it appears as connected but no replies from MKTK goes back to iPad. Preamble is Long and Cipher is AES CCM as was before.
What RouterOS version you used before v6.2?

As far as I remember it was 6.1, but possible one of RC.

peo · Thu Aug 08, 2013 9:01 pm

I want to rapport bug.
Certificates>Ca Certificates
There is no way to remove ca certificate on ROS ver. 6.2. Tested on RB1100AH and RB493G

chemil · Fri Aug 09, 2013 10:28 am

It seems that this function is no longer supported in v6 and CCR:

/tool graphing set page-refresh=500

friction · Fri Aug 09, 2013 9:13 pm

Anyone else experiencing this on a RB2011UAS-2HnD or similar? It keeps crashing continuously since last firmware upgrade towards ROS 6.2 (two days ago):

(28 messages not shown)
aug/09/2013 15:54:39 system,error,critical router was rebooted without proper shutdown by watchdog timer
aug/09/2013 16:39:45 system,error,critical router was rebooted without proper shutdown by watchdog timer
aug/09/2013 16:54:38 system,error,critical router was rebooted without proper shutdown by watchdog timer
aug/09/2013 17:24:43 system,error,critical router was rebooted without proper shutdown by watchdog timer
aug/09/2013 17:39:43 system,error,critical router was rebooted without proper shutdown by watchdog timer
aug/09/2013 17:54:42 system,error,critical router was rebooted without proper shutdown by watchdog timer
aug/09/2013 18:24:43 system,error,critical router was rebooted without proper shutdown by watchdog timer
aug/09/2013 18:54:44 system,error,critical router was rebooted without proper shutdown by watchdog timer

/system watchdog> print
     watch-address: none
    watchdog-timer: yes
     no-ping-delay: 5m
  automatic-supout: yes
  auto-send-supout: no

Rudios · Fri Aug 09, 2013 9:18 pm

Anyone else experiencing this on a RB2011UAS-2HnD or similar? It keeps crashing continuously since last firmware upgrade towards ROS 6.2 (two days ago):

(28 messages not shown)
aug/09/2013 15:54:39 system,error,critical router was rebooted without proper shutdown by watchdog timer
aug/09/2013 16:39:45 system,error,critical router was rebooted without proper shutdown by watchdog timer
aug/09/2013 16:54:38 system,error,critical router was rebooted without proper shutdown by watchdog timer
aug/09/2013 17:24:43 system,error,critical router was rebooted without proper shutdown by watchdog timer
aug/09/2013 17:39:43 system,error,critical router was rebooted without proper shutdown by watchdog timer
aug/09/2013 17:54:42 system,error,critical router was rebooted without proper shutdown by watchdog timer
aug/09/2013 18:24:43 system,error,critical router was rebooted without proper shutdown by watchdog timer
aug/09/2013 18:54:44 system,error,critical router was rebooted without proper shutdown by watchdog timer

/system watchdog> print
     watch-address: none
    watchdog-timer: yes
     no-ping-delay: 5m
  automatic-supout: yes
  auto-send-supout: no

Have updated two RB2011's within days of the 6.2 release and neither of them is showing this.

Chupaka · Fri Aug 09, 2013 10:08 pm

It seems that this function is no longer supported in v6 and CCR:
Code: Select all
/tool graphing set page-refresh=500

what do you mean?

[admin@TestPlace] /tool graphing> set page-refresh=500
[admin@TestPlace] /tool graphing> print 
   store-every: hour
  page-refresh: 500
[admin@TestPlace] /tool graphing>

Attila · Fri Aug 09, 2013 11:36 pm

CCR1036 interface graphs are lost after reboot. Reset Configurtion and NetInstall with nand format didn't solve the issue.

Fri Aug 09, 2013 11:47 pm

Graphs loosing, scripts problems and not getting the voltage on RB2011UAS-2HND made me to downgrade from 6.2 to 5.25, where everything works.
Seems like version 6 is untested crap with only one reason: to make all so far content and happy administrators mad. But why? For what?

ntwkedout · Sat Aug 10, 2013 2:42 am

IPSEC

Mikrotik RB493G to cisco ASA 5520

v5.25 everything is working Upgraded to 6.2 and SIP phones stopped registering but all other traffic seems to be working. Downgraded back to 5.25 everything started working again.

I have been troubleshooting after hours and cant see why the phones stopped working. Everything looks like it should be working on 6.2 but it does not. Stuck on this one. Going back 5.25 for now.

After many hours of troubleshooting I found the issue to be service ports under firewall, I disabled sip, tftp and cleared connections. SIP phones all came back online and started working. Wish I had found it sooner would have saved me some time. In 5.25 I never had to disable them.

Tested on RB493G and CCR1036-12G-4S v6.2, both connecting to a ASA5520. Only thing is I am still having to flush SA every now and then because of dead tunnels. Using a script to keep pings going accross tunnels.

** Thanks for the hard work mikrotik staff cant wait to see v7

jandafields · Sat Aug 10, 2013 2:45 am

SIP connections don't work in 6.0 and 6.1 and 6.2

They work good in 5.25

ntwkedout · Sat Aug 10, 2013 4:58 am

SIP connections don't work in 6.0 and 6.1 and 6.2

They work good in 5.25

Did you do what I said worked for me in my last post? I am guessing you did not. I spent the last 2 days troubleshooting why sip would not work on 6.2 and it was the service ports. Disable them.

jandafields · Sat Aug 10, 2013 5:42 am

SIP connections don't work in 6.0 and 6.1 and 6.2

They work good in 5.25
Did you do what I said worked for me in my last post? I am guessing you did not. I spent the last 2 days troubleshooting why sip would not work on 6.2 and it was the service ports. Disable them.

Odd, I saw the first part of your post, but I didn't even notice your solution below it! Thanks, I'll give it a try.

chemil · Sat Aug 10, 2013 8:50 pm

It seems that this function is no longer supported in v6 and CCR:
Code: Select all
/tool graphing set page-refresh=500
what do you mean?
Code: Select all
[admin@TestPlace] /tool graphing> set page-refresh=500
[admin@TestPlace] /tool graphing> print 
   store-every: hour
  page-refresh: 500
[admin@TestPlace] /tool graphing> 

Hi, may be that I didn't fully understand this feature but in my case this setting didn't add meta refresh into html code.
http://forum.mikrotik.com/viewtopic.php?f=2&t=32569

efaden · Sun Aug 11, 2013 12:07 am

Also... I now have a graph showing the increased CPU usage on 6.2. This is from a box running minimal stuff. If you look at mid monday you can see a distinct drop from an average of 8% to an average of 3% or so. This was when the box was downgraded to 6.1. There is basically no load on this box and it is just idle. Not really sure why the CPU usage is higher in 6.2... anyone else seen this?

Normis?

aragon · Sun Aug 11, 2013 6:03 pm

Thanks for the scripting fixes since 6.1, but unfortunately we're hitting a new scripting bug.

:local q
:local msg
:set msg "hello world"

:set q [:parse ("/put message=\"the message is: " . $msg . "\"")]
$q
:put $q

On 6.2:

[admin@dev] /system script> run test
(eval /putmessage=the message is: hello world)

On 5.24:

[aragon@rb1.kush] /system script> run test
the message is: hello world
the message is: hello world

Reported to support@ - #2013081166000228

wirelesswaves · Mon Aug 12, 2013 1:12 am

I have several omnitiks and a few sextants on 6.1 and my phone customers are not complaining.. I put 6.1 on a couple of months ago to cure some ospf irregularities in that part of the network.

I didn't "uncheck" any service settings and all my Voip ATA's, registration and quality are good!

ekkas · Mon Aug 12, 2013 1:42 am

Simple suggestion for ROS6 which might have huge impact.
Very nice to see from-dscp-high-3-bits and from-priority-to-high-3-bits.
Thank you for that! I will make a big difference.

Now what about the other 3+2 bits in that little Byte.

I think I would be usefull if ROS would be able to set filter/mangle rule on condition dscp-low-3-bits and have actions set-dscp-high-3-bits and set-dscp-low-3-bits

This would mean, I can use high 3 bits for QoS and low 3 bits for e.g. differentiating capped and uncapped users, or to give QoS sub-classes/queues, or to classify routing info (we have different gateways on large network, so 'cheap' clients can be routed differently, etc. etc. etc.)
The point is I think DSCP is fantastic as it is the only way to transmit information between routers on a packet by packet bases.

It would be great if MT can let us harness the untapped great potential of these few bits.
I think if it is available, it would inspire people to think differently about their networks and it can save lots of CPU power downstream as some info can be transmitted like a simple network-wide connection mark, and certain processes can be centralized & transmitted.

If you really want to go full out, in Winbox you can perhaps (in mangle condition), choose to test between DSCP (6 bits) or high-3 or low-3 bits and/or have the last 2 EC bits there maybe? (or some other way?)
Then in action, choose to set dscp-6-bits, dscp-high-3-bits or dscp-high-3-bits. When setting dscp-high-3-bits, the lower 3 must be unchanged, and vice versa.

Am I just looking at it wrong, or does someone agree?

Regards
Ekkas

paranoid · Mon Aug 12, 2013 12:45 pm

Hi

seems that proxy arp with sstp tunnels is not working
any other got that problem? I've seen that problem since ros 6.0rc14

kind regards

Mon Aug 12, 2013 1:10 pm

Graphs loosing, scripts problems and not getting the voltage on RB2011UAS-2HND made me to downgrade from 6.2 to 5.25, where everything works.
Seems like version 6 is untested crap with only one reason: to make all so far content and happy administrators mad. But why? For what?

Please email to support@mikrotik.com with these problems so if we could check and fix them.
How are you reading the voltage - from snmp?

Mon Aug 12, 2013 1:28 pm

I think I found a typo on the login.html file of the hotspot files.
The file starts with html, head and title and inside the head there are 4 meta assignments.
I believe that the last meta entry is ended with a faulty double quote character, just before the closing arrow '>'
If the hotspot login page is shown the style piece is shown on the browser.
If I change the last double quote to the one used on all the other places in the file, the style showing in browser is gone.
I hope my explanation is clear.

Are you talking about this? What is so faulty about it?

Screen Shot 2013-08-12 at 13.28.11.jpg

AlArenal · Mon Aug 12, 2013 1:30 pm

Hm.. he said he had four meta entries and the faulty is the last but your screenie only shows three. Maybe he could paste / screenshot his version.

Mon Aug 12, 2013 1:34 pm

Also... I now have a graph showing the increased CPU usage on 6.2. This is from a box running minimal stuff. If you look at mid monday you can see a distinct drop from an average of 8% to an average of 3% or so. This was when the box was downgraded to 6.1. There is basically no load on this box and it is just idle. Not really sure why the CPU usage is higher in 6.2... anyone else seen this?

Normis?

Please try to use 'tool profile' and check what process eats so much CPU and report back with the results.

Mon Aug 12, 2013 1:37 pm

Hm.. he said he had four meta entries and the faulty is the last but your screenie only shows three. Maybe he could paste / screenshot his version.

So probably he has custom html and introduced the mistake himself, because this is default header from latest RouterOS.

TKITFrank · Mon Aug 12, 2013 2:29 pm

Hi Normis,

Any news on when a prebuild of 6.3 with the ip proxy access is fixed?

/Frank

Mon Aug 12, 2013 2:34 pm

Hi Normis,

Any news on when a prebuild of 6.3 with the ip proxy access is fixed?

/Frank

when we will fix it, pre-release will be made available

Rudios · Mon Aug 12, 2013 2:41 pm

Hm.. he said he had four meta entries and the faulty is the last but your screenie only shows three. Maybe he could paste / screenshot his version.
So probably he has custom html and introduced the mistake himself, because this is default header from latest RouterOS.

Normis,

I did not create a login page myself.
I've attached a screenshot of the page.
I've also seen that there are two different login pages, one in the root, and one in the lv directory.
The last one does not show the line which seems to be faulty in the other one.

Mon Aug 12, 2013 2:43 pm

Ok, looks like ours after all, thanks for report

AlArenal · Mon Aug 12, 2013 2:46 pm

There's a slash missing, but it will get parsed without problems in any browser. It's more of cosmetic thing and not 100% valid according to the doctype (xhtml requires correct nesting and opening / closing of tags), but then again - which website is?

Mon Aug 12, 2013 2:49 pm

Graphs loosing, scripts problems and not getting the voltage on RB2011UAS-2HND made me to downgrade from 6.2 to 5.25, where everything works.
Seems like version 6 is untested crap with only one reason: to make all so far content and happy administrators mad. But why? For what?
Please email to support@mikrotik.com with these problems so if we could check and fix them.
How are you reading the voltage - from snmp?

I alerady did (nr. 2013080466000205 - about loosing graphs). Today I received advice to netinstall v6.2. I am not sure that it will help, but I will try to.

Regarding the voltage I use this script:

# Define global variables
:global TempLast 
:global VoltLast 

# Define and fill actual values
:local TempAct [/system health get temperature]
:local VoltRAW [/system health get voltage]

# Divide voltage by 10 as it is 10 times multiplicated
:local VoltAct [ ($VoltRAW / 10) ]

# Calculate differences between last and actual values
:local TempDiff [ ($TempAct - $TempLast) ]
:local VoltDiff [ ($VoltAct - $VoltLast) ]
:tonum TempDiff
:tonum VoltDiff

# Calculate absolute values of differences
:if ( $TempDiff < 0 ) do={ :set TempDiff [ ( 0 - $TempDiff ) ] } 
:if ( $VoltDiff < 0 ) do={ :set VoltDiff [ ( 0 - $VoltDiff ) ] } 

# Do the reporting
:if ( $TempDiff >= 2 ) do={ :log info ("Environment: Last Temperature was $TempLast C. Actual Temperature is $TempAct C.") } 
:if ( $VoltDiff >= 2 ) do={ :log info ("Environment: Last Voltage was $VoltLast V. Actual Voltage is $VoltAct V.") } 

:if ( $TempDiff >= 2 ) do={ :set TempLast $TempAct } 
:if ( $VoltDiff >= 2 ) do={ :set VoltLast $VoltAct } 

# Debugging info below....
#:log info $TempAct
#:log info $VoltAct
#:log info $TempDiff
#:log info $VoltDiff

#:log info ("Environment: Actual Temperature is $TempAct C. Voltage is $VoltAct V.")

It gets voltage and temperature, compares it with last values and if vlaues change, it makes the info log entry.

This script works in v5.25 but does not work in v6.2. In v6.2 also DnsExit script does not work (do not know why):

# Set needed variables
:local username "some-name"
:local password "some-pass"
:local hostname "some-url.linkpc.net"

:global dyndnsForce
:global previousIP
#:global currentIP

# print some debug info
#:log info ("UpdateDnsExit: username = $username")
#:log info ("UpdateDnsExit: password = $password")
#:log info ("UpdateDnsExit: hostname = $hostname")
#:log info ("UpdateDnsExit: previousIP = $previousIP")

# get the current IP address from the internet (in case of double-nat)
/tool fetch mode=http address="ip.dnsexit.com" src-path="/" dst-path="usb1/dnsexit.tmp" port=80 host=ip.dnsexit.com
:local result [/file get [/file find name=usb1/dnsexit.tmp] contents]

# parse the current IP result
:local resultLen [:len $result]
:local currentIP [:pick $result 0 ($resultLen-1)]
#:log info "UpdateDnsExit: currentIP = $currentIP"

# Remove the # on next line to force an update every single time - useful for debugging,
# but you could end up getting blacklisted by DnsExit!

#:set dyndnsForce true

# Determine if dyndns update is needed
# more dyndns updater request details on DnsExit.com 

:if (($currentIP != $previousIP) || ($dyndnsForce = true)) do={
:set dyndnsForce false
:log info "UpdateDnsExit: Current IP $currentIP"
:log info "UpdateDnsExit: Previous IP $previousIP"
:set previousIP $currentIP
/tool fetch mode=http address="www.dnsexit.com" \
src-path="RemoteUpdate.sv?login=$username&password=$password&host=$hostname&myip=$currentIP" \
dst-path="usb1/dnsexit.txt" port=80 host=www.dnsexit.com
:local result [/file get usb1/dnsexit.txt contents]
#:log info "UpdateDnsExit: previousIP = $previousIP"
:log info ("DnsExit: DnsExit update needed")
:log info ("UpdateDnsExit: DnsExit Update Result: ".$result)
:log info ("DnsExit Update Result: ".$result)
} else={
:log info ("UpdateDnsExit: No DnsExit update needed")
}

Mon Aug 12, 2013 3:11 pm

jarda, please contact support@mikrotik.com we will provide you with test release of v6.3 which will have a fix for this problem.

Rudios · Mon Aug 12, 2013 3:48 pm

There's a slash missing, but it will get parsed without problems in any browser. It's more of cosmetic thing and not 100% valid according to the doctype (xhtml requires correct nesting and opening / closing of tags), but then again - which website is?

No it is not the slash.
It is a different double quote char. Look closely and you will see difference between the start and end of the line.

AlArenal · Mon Aug 12, 2013 3:57 pm

Ah! You're right, thanks for pointing it out. It's something that's easily overlooked.

P.S.:
But the slash is missing, nevertheless

javcia · Mon Aug 12, 2013 3:57 pm

Does any one have any problems with sfp? At this moment i`m using ether6 as uplink (i use media converter from fibre to copper)and it`s working fine, but when i change to SFP1 i lose connection. I use mikrotik S-31DLC20D sfp module and CCR recognize it.
Current ROS is 6.2 and firmware version is 3.09.

Chupaka · Mon Aug 12, 2013 4:46 pm

Hi, may be that I didn't fully understand this feature but in my case this setting didn't add meta refresh into html code.
http://forum.mikrotik.com/viewtopic.php?f=2&t=32569

haha, I can't see it even in v5.14 - seems like the problem is older than you think...

Lakis · Mon Aug 12, 2013 7:20 pm

ROS 6.2

/ip address> add
address: 9.9.9.8/24
interface: ether2
failure: netmask cannot be /0

/ip address> add
address: 192.168.88.55/24
interface: ether2
Script Error: action cancelled

??????

Mon Aug 12, 2013 8:13 pm

Both Winbox and WebFig are incorrectly reporting IPsec SA expiration time in 6.2.

RouterOS-6_2-sa-exp-time.jpg

No problems in console:

[andrew@MikroTik] > /ip ipsec installed-sa print
Flags: A - AH, E - ESP, P - pfs 
 0 E  spi=0x30BB8DB src-address=x.x.x.x dst-address=x.x.x.x auth-algorithm=sha1 enc-algorithm=3des replay=4 
      state=mature auth-key="d117cc9bab41bcfe966454cc461f0252882c998f" enc-key="38308591a3c5858c90bca35a66ef46a44e8f61347e428413
      addtime=aug/12/2013 19:46:38 expires-in=40m43s add-lifetime=48m/1h current-bytes=253972

bosanac · Tue Aug 13, 2013 1:58 pm

Both Winbox and WebFig are incorrectly reporting IPsec SA expiration time in 6.2.

RouterOS-6_2-sa-exp-time.jpg

No problems in console:

[andrew@MikroTik] > /ip ipsec installed-sa print
Flags: A - AH, E - ESP, P - pfs 
 0 E  spi=0x30BB8DB src-address=x.x.x.x dst-address=x.x.x.x auth-algorithm=sha1 enc-algorithm=3des replay=4 
      state=mature auth-key="d117cc9bab41bcfe966454cc461f0252882c998f" enc-key="38308591a3c5858c90bca35a66ef46a44e8f61347e428413
      addtime=aug/12/2013 19:46:38 expires-in=40m43s add-lifetime=48m/1h current-bytes=253972

and also in V6.1.

Toiletbowl · Tue Aug 13, 2013 2:06 pm

I think I found a typo on the login.html file of the hotspot files.
The file starts with html, head and title and inside the head there are 4 meta assignments.
I believe that the last meta entry is ended with a faulty double quote character, just before the closing arrow '>'
If the hotspot login page is shown the style piece is shown on the browser.
If I change the last double quote to the one used on all the other places in the file, the style showing in browser is gone.
I hope my explanation is clear.
Are you talking about this? What is so faulty about it?

Screen Shot 2013-08-12 at 13.28.11.jpg

Hi rudios

are you sure about this?

if you mind can i have your html already fix so i can upload easily without modifying the page TBH i have no experience on web editing lol

thanks

bosanac · Tue Aug 13, 2013 2:11 pm

Both Winbox and WebFig are incorrectly reporting IPsec SA expiration time in 6.2. And also in 6.1

Rudios · Tue Aug 13, 2013 3:02 pm

I think I found a typo on the login.html file of the hotspot files.
The file starts with html, head and title and inside the head there are 4 meta assignments.
I believe that the last meta entry is ended with a faulty double quote character, just before the closing arrow '>'
If the hotspot login page is shown the style piece is shown on the browser.
If I change the last double quote to the one used on all the other places in the file, the style showing in browser is gone.
I hope my explanation is clear.
Are you talking about this? What is so faulty about it?

Screen Shot 2013-08-12 at 13.28.11.jpg
Hi rudios

are you sure about this? if you mind can i have your html already fix so i can upload easily without modifying the page TBH i have no experience on web editing lol

thanks

I did not change the file.
You can change it very easily.
Copy the login.html from the files via winbox onto your PC.
Open it in e.g. notepad and remove the double quote char I marked.
Put a normal double quote char (used anywhere else in the file)
You could also remove the complete line, since the Latvian translated login.html file does not even have that 4th meta line.
Shouldn't be a problem deleting it completely.
When you have changed the file, copy it back to the mikrotik

ekkas · Wed Aug 14, 2013 3:31 am

Packet sniffer does not seem to work in 6.2. (No packets results show)
6.1 works fine.
(RB433AH)

janisk · Wed Aug 14, 2013 1:29 pm

Packet sniffer does not seem to work in 6.2. (No packets results show)
6.1 works fine.
(RB433AH)

will be fixed in 6.3

Wed Aug 14, 2013 6:46 pm

jarda, please contact support@mikrotik.com we will provide you with test release of v6.3 which will have a fix for this problem.

Unfortunatelly it did not help with the graphs. I have not checked the scripts, because when first problem (and I think the easier) is not solved than there is no reason to continue wasting the time by testing on devices that I need to run.

I have installed v6.3rc1 into one of my devices. When it is starting I can see the old graphs as they was lastly recorded by v5.25 with the last time when v5.25 lastly run(yesterday 9:53)

After a few seconds the graphs disappear (probably due to new time setup)

Meanwhile (yesterday 9:53 till today approx 17:00) I run v.6.2. No recordingw were made. After that, today around 17:00 I installed v6.3rc and let it run some time (half an hour) to record the data into graphs. After restart, no recording was done.

Only the way to continue data recording for graphs is to downgrade again back to v5.25...

TerAnYu · Thu Aug 15, 2013 5:48 am

Hello!
After installing the 6.2 noticed that the total does not exceed the port speed 70Mb/s, using p2p very noticeable restriction, although statistics indicate the port's current speed as 80-100 Mb/s.
Routerboard RB450G

karmag · Thu Aug 15, 2013 8:02 am

Discarding packet: locally originated, problem 6.2

tell me how to solve the problem

armandfumal · Thu Aug 15, 2013 9:41 pm

BUG report

V 6.2 release

editing hotspot user profile, setting incoming-filter and outgoing-filter in winbox does save it correctly.

console ok

ferrycupu · Fri Aug 16, 2013 4:49 am

let's try.....
i hope no bugs....

zamani · Fri Aug 16, 2013 10:41 am

hello everyone

I have a problem with user manager 6.2

when I create a profile with 2 limitations in user manager with different time,everything works wrong.
and users can pass my limitations.
(with one limitation everything works fine)

16-08-2013 11-44-33.jpg

for example with this setting when a user connect in 9 am everything works fine until 10:56am. after 10:56am second limitation not working,and user can pass 2M limitation.unless user disconnect in 10:55 then connect again in 10:56!!

please test.

thanks

morf · Fri Aug 16, 2013 11:06 am

It's fun!
ROS 6.2 shows not the correct values of the traffic.
How is this possible?

armandfumal · Fri Aug 16, 2013 10:10 pm

Feature Request for 6.3

adding the possibility to execute script with parameters
/system script run script1 $1 $2 $3 $4

thanks

Armand

raz · Sat Aug 17, 2013 4:53 pm

Packet sniffer does not seem to work in 6.2. (No packets results show)
6.1 works fine.
(RB433AH)

cant confirm that, tested on ccr1016 and rb1100.

Diamont · Sun Aug 18, 2013 7:17 pm

Can PPPoE server accept any username and password to accept all pppoe clients?

armandfumal · Sun Aug 18, 2013 8:48 pm

Packet sniffer does not seem to work in 6.2. (No packets results show)
6.1 works fine.
(RB433AH)

works ok on RB951G with 6.2

armandfumal · Mon Aug 19, 2013 1:27 pm

6.2 BUG REPORT

plateforme RG951G

tool graphs:
Flags: X - disabled
# INTERFACE ALLOW-ADDRESS STORE-ON-DISK
0 ether1-gat... 0.0.0.0/0 yes

not store on disk.
only in memory.
When I reboot, I lost all graphs...

Armand

AlArenal · Mon Aug 19, 2013 1:28 pm

not store on disk.
only in memory.
When I reboot, I lost all graphs...

Same here on RB2011UAS-RM.

armandfumal · Mon Aug 19, 2013 4:18 pm

6.2 BUG REPORT

plateforme RB951G

Reset counters on ethernet interface does not works
winbox and cli

bysard · Mon Aug 19, 2013 4:32 pm

Jarda "I alerady did (nr. 2013080466000205 - about loosing graphs). Today I received advice to netinstall v6.2. I am not sure that it will help, but I will try to."

This is interesting. MT support told me there is no need to use netinstall. They also told me because I used it (RoS 6.2 with netinstall 5.25) I lost 2 routers. Ticket no. 2013081366000304. So i guess there is different answer for every MT admin about the same thing. Depends which answer suits Mikrotik best? Am I wrong?

Toiletbowl · Tue Aug 20, 2013 2:03 am

Hi mikrotik

When will be the next release?

Thanks

Russian · Tue Aug 20, 2013 2:35 am

Hi mikrotik

When will be the next release?

Thanks

I hope the new version gets tested better before it is called "stable"

If you need a beta tester, you can send me the file/link, as we run quite some scripts and routing on our networks here. I will know soon enough if it works or not, or what might have gone wrong within the scope we use.

Tx

efaden · Tue Aug 20, 2013 3:25 am

Hi mikrotik

When will be the next release?

Thanks
I hope the new version gets tester better before it is called "stable"

If you need a beta tester, you can send me the file/link, as we run quite some scripts and routing on our networks here. i will know soon enough if it works or not, or what might have gone wrong within the scope we use.

Tx

+1 for stability. 6.2 was kind of a mess.

Sent from my SCH-I545 using Tapatalk 2

bwaters · Tue Aug 20, 2013 6:44 pm

I mentioned this before on 6.1 -- and I put in a bug request -- is that mantis installation still being used or is the forum the way to go?

Anyway the bug is here http://bugs.mikrotik-routeros.com/view.php?id=50

in 6.0RC9 and below tool fetch lets you specify a ip value in the address and a host name in the host parameter. It would call out the address with the host header you specified in host.

With 6.2 it requires the address parameter but completely ignores it if the host is present and instead does a dns lookup on the host name and then routes to it. Being able to specify the address independently from the host name is really handy for scripting monitoring or failover on webservers that have virtual hosts and multiple ips per domain name. it is also very userful for monitoring/failover of proxies.

honzam · Tue Aug 20, 2013 7:01 pm

Hi mikrotik

When will be the next release?

Thanks
I hope the new version gets tested better before it is called "stable"

If you need a beta tester, you can send me the file/link, as we run quite some scripts and routing on our networks here. I will know soon enough if it works or not, or what might have gone wrong within the scope we use.
Tx

http://www.mikrotik.com/download/share/ ... 6.3rc1.npk

Chupaka · Wed Aug 21, 2013 12:13 am

It's fun!
ROS 6.2 shows not the correct values of the traffic.
How is this possible?

Tools -> Torch shows only traffic on the selected interface. in Ethernet stats you see also traffic of subinterfaces (VLANs), which is not shown in Torch

pacman · Wed Aug 21, 2013 1:13 am

Will 6.3 include any fixes for the S**t reliability issues with wireless links we have seen with every version 6.x? I am getting pretty p***ed off with having to reboot APs just because they randomly decide that they cannot be bothered to pass traffic anymore (even though they are still registered with the client).

Why not try testing your software properly for a change before releasing it?????????

efaden · Wed Aug 21, 2013 1:23 am

It's fun!
ROS 6.2 shows not the correct values of the traffic.
How is this possible?
Tools -> Torch shows only traffic on the selected interface. in Ethernet stats you see also traffic of subinterfaces (VLANs), which is not shown in Torch

I didn't realize that the interfaces also showed VLAN traffic. Is this always the case or is it a bug in 6.2? Is there a way to get it to show only traffic not "claimed" by a vlan?

-Eric

ekkas · Wed Aug 21, 2013 3:34 am

Feature Request for 6.3
adding the possibility to execute script with parameters
/system script run script1 $1 $2 $3 $4
thanks
Armand

+1

and to have a firewall filter/mangle:
action=run-script
script-str=MyScript $src-address $out-interface...

That would be ridiculous!

mxmxmxmxmx · Wed Aug 21, 2013 3:16 pm

Will 6.3 include any fixes for the S**t reliability issues with wireless links we have seen with every version 6.x? I am getting pretty p***ed off with having to reboot APs just because they randomly decide that they cannot be bothered to pass traffic anymore (even though they are still registered with the client).

Why not try testing your software properly for a change before releasing it?????????

I've got the same problem

bysard · Wed Aug 21, 2013 4:03 pm

Will 6.3 include any fixes for the S**t reliability issues with wireless links we have seen with every version 6.x? I am getting pretty p***ed off with having to reboot APs just because they randomly decide that they cannot be bothered to pass traffic anymore (even though they are still registered with the client).

Why not try testing your software properly for a change before releasing it?????????
I've got the same problem

+1

javcia · Wed Aug 21, 2013 5:15 pm

IP route is not working with sfp interface.

armandfumal · Thu Aug 22, 2013 12:05 am

IP route is not working with sfp interface.

be more precise. plateforme, Ros version, etc...

On CCR-1036 with 6.2, It is working for me.
regards

javcia · Thu Aug 22, 2013 1:25 pm

IP route is not working with sfp interface.
be more precise. plateforme, Ros version, etc...

On CCR-1036 with 6.2, It is working for me.
regards

armandfumal thank you for responce. I have CCR1036 with 6.2 and 3.09 firmware. At this moment i`m using ether6 as a main uplink port. But when i change from ether6 to sfp1 my gateway is unreachable which is x.x.172.77. I`m using Mikrotik S-31DLC20D sfp.
My config:
/ip route> print
# DST-ADDRESS PREF-SRC GATEWAY DISTANCE
0 A S 0.0.0.0/0 x.x.172.78 ether6 1
1 ADC 10.0.1.0/24 10.0.1.1 VLAN - 101 0
2 ADC x.x.172.76/30 x.x.172.78 ether6 0
3 ADC x.x.232.0/24 x.x.232.1 VLAN - 100 0
4 ADC 172.16.1.0/24 172.16.1.1 VLAN - 102 0
5 ADC 172.16.2.0/24 172.16.2.1 ether1 0
6 ADC 192.168.2.0/24 192.168.2.1 VLAN - 2 0

ip address print
Flags: X - disabled, I - invalid, D - dynamic
# ADDRESS NETWORK INTERFACE
0 172.16.2.1/24 172.16.2.0 ether1
1 192.168.2.1/24 192.168.2.0 VLAN - 2
2 x.x.232.1/24 x.x.232.0 VLAN - 100
3 10.0.1.1/24 10.0.1.0 VLAN - 101
4 172.16.1.1/24 172.16.1.0 VLAN - 102
5 x.x.172.78/30 x.x.172.76 ether6

ip firewall nat print
Flags: X - disabled, I - invalid, D - dynamic
0 chain=srcnat action=src-nat to-addresses=x.x.232.254
src-address=192.168.2.0/24 out-interface=ether6

1 chain=srcnat action=src-nat to-addresses=x.x.232.254
src-address=172.16.2.0/24 out-interface=ether6

2 chain=srcnat action=src-nat to-addresses=x.x.232.254
src-address=172.16.1.0/24 out-interface=ether6

3 chain=srcnat action=src-nat to-addresses=x.x.232.254
src-address=10.0.1.0/24 out-interface=ether6

Rudios · Thu Aug 22, 2013 4:32 pm

IP route is not working with sfp interface.
be more precise. plateforme, Ros version, etc...

On CCR-1036 with 6.2, It is working for me.
regards
armandfumal thank you for responce. I have CCR1036 with 6.2 and 3.09 firmware. At this moment i`m using ether6 as a main uplink port. But when i change from ether6 to sfp1 my gateway is unreachable which is x.x.172.77. I`m using Mikrotik S-31DLC20D sfp.
My config:
/ip route> print
# DST-ADDRESS PREF-SRC GATEWAY DISTANCE
0 A S 0.0.0.0/0 x.x.172.78 ether6 1
1 ADC 10.0.1.0/24 10.0.1.1 VLAN - 101 0
2 ADC x.x.172.76/30 x.x.172.78 ether6 0
3 ADC x.x.232.0/24 x.x.232.1 VLAN - 100 0
4 ADC 172.16.1.0/24 172.16.1.1 VLAN - 102 0
5 ADC 172.16.2.0/24 172.16.2.1 ether1 0
6 ADC 192.168.2.0/24 192.168.2.1 VLAN - 2 0

ip address print
Flags: X - disabled, I - invalid, D - dynamic
# ADDRESS NETWORK INTERFACE
0 172.16.2.1/24 172.16.2.0 ether1
1 192.168.2.1/24 192.168.2.0 VLAN - 2
2 x.x.232.1/24 x.x.232.0 VLAN - 100
3 10.0.1.1/24 10.0.1.0 VLAN - 101
4 172.16.1.1/24 172.16.1.0 VLAN - 102
5 x.x.172.78/30 x.x.172.76 ether6

ip firewall nat print
Flags: X - disabled, I - invalid, D - dynamic
0 chain=srcnat action=src-nat to-addresses=x.x.232.254
src-address=192.168.2.0/24 out-interface=ether6

1 chain=srcnat action=src-nat to-addresses=x.x.232.254
src-address=172.16.2.0/24 out-interface=ether6

2 chain=srcnat action=src-nat to-addresses=x.x.232.254
src-address=172.16.1.0/24 out-interface=ether6

3 chain=srcnat action=src-nat to-addresses=x.x.232.254
src-address=10.0.1.0/24 out-interface=ether6

If you want to move your uplink to your sfp interface, you should also change all the relevant config like NAT.
Change the ether6 to sfp interface.