6.30 released

Wed Jul 08, 2015 5:02 pm

What's new in 6.30 (2015-Jul-08 09:07):

*) wireless - added WMM power save suport for mobile devices;
*) firewall - sip helper improved, large packets no longer dropped;
*) fixed encryption 'out of order' problem on SMP systems;
*) email - fix sending multiple consecutive emails;
*) fixed router lockup on leap seconds with installed ntp package;
*) ccr - made hardware watchdog work again (was broken since v6.26);
*) console - allow users with 'policy' policy to change script owner;
*) icmp - use receive interface address when responding with icmp errors;
*) ipsec - fail ph2 negitioation when initiator proposed key length
does not match proposal configuration;
*) timezone - updated timezone information to 2015e release;
*) ssh - added option '/ip ssh stong-crypto'
*) wireless - improve ac radio coexistence with other wireless clients, optimized
transmit times to not interfere with other devices;
*) console - values of $".id", $".nextid" and $".dead" are avaliable for
use in 'print where' expressions;
*) console - ':execute' command now accepts script source in "{}" braces,
like '/system scripts add source=' does;
*) console - ':execute' command now returns internal number of running job,
that can be used to check and stop execution. For example:
:local j [:execute {/interface print follow where [:log info "$name"]}]
:delay 10s
:do { /system script job remove $j } on-error={}
*) console - firewall 'print' commands now show all entries including
dynamic, 'all' argument now has no effect;
*) ipsec - increase replay window to 128;
*) fixed file transfer on devices with large RAM memory;
*) pptp - fixed "encryption got out of sync" problem;
*) ppp - disable vj tcp header compression;
*) api - reduce api tcp connection keepalive delay to 30 seconds,
will timeout idle connections in about 5 minutes;
*) pptp & l2tp & sstp client: support the case were server issues its tunnel
ip address the same as its public one;
*) removed wireless package from routeros bundle package,
new wireless-fp is left in place and wireless-cm2 added as option;
*) pptp & l2tp client: when adding default route, add special exception route for
a tunnel itself (no need to add it manually anymore);
*) improved connection list: added connection packet/byte counters,
added separate counters for fasttrack, added current rate display,
added flag wheather connection is fasttracked/srcnated/dstnated,
removed 2048 connection entry limit;
*) tunnels - eoip, eoipv6, gre,gre6, ipip, ipipv6, 6to4 tunnels
have new property - ipsec-secret - for easy setup of ipsec
encryption and authentication;
*) firewall - added ipsec-policy matcher to check wheather packet
was/will be ipsec processed or not;
*) possibility to disable route cache - improves DDOS attack
handling performance up to 2x (note that ipv4 fastpath depends on route cache);
*) fasttrack - added dummy firewall rule in filter and mangle tables
to show packets/bytes that get processed in fasttrack and bypass firewall;
*) fastpath - vlan interfaces support fastpath;
*) fastpath - partial support for bonding interfaces (rx only);
*) fastpath - vrrp interfaces support fastpath;
*) fixed memory leak on CCR devices (introduced in 6.28);
*) lte - improved modem identification to better support multiple identical modems;
*) snmp - fix system scripts table;

JanezFord · Wed Jul 08, 2015 5:31 pm

Hello,

is there any specific command needed to enable WMM Power Save of is it enabled by default. TNX ... upgraded RB951G without issues.

JF.

kristaps · Wed Jul 08, 2015 5:46 pm

No It just enabled by default on wireless-fp and cm2 packages

avantwireless · Wed Jul 08, 2015 6:12 pm

I could not get the regular wireless package be seen after uploading it to a router. What is the process to get it to be added to the packages? We see less disconnects with nstream with the regular wireless package than the fp or cm2 packages and need it for that reason.

CyberTod · Wed Jul 08, 2015 6:48 pm

*) wireless - added WMM power save suport for mobile devices;

Waiting a long time for this. Nice.

lekozs · Wed Jul 08, 2015 7:11 pm

I could not get the regular wireless package be seen after uploading it to a router. What is the process to get it to be added to the packages? We see less disconnects with nstream with the regular wireless package than the fp or cm2 packages and need it for that reason.

*) removed wireless package from routeros bundle package,
new wireless-fp is left in place and wireless-cm2 added as option;

... there is no "regular" wireless package anymore...

solaoxo · Wed Jul 08, 2015 7:53 pm

'execute' in not running scheduler.

moep · Wed Jul 08, 2015 8:15 pm

I just upgrades my CCR1009-PC
I now have problems on my bonding interface.
Its configured for active-backup. ether8 is master and ether7 is slave. If you disconnect the ethernet cable from ether8 you get an instant reboot with the following lines in the log/terminal:

System rebooted because of kernel failure
router was rebooted without proper shutdown, probably kernel failure

If you want to change the bonding interface, you will get the same instant reboot.

Is there a workaround/fix for this issue yet? Ist it even "on the radar"?

Otherwise the device is working well.

Thanks for any help

paoloaga · Wed Jul 08, 2015 8:36 pm

Am I wrong or all_packages-smips-6.30.zip doesn't contain the wireless package? And I can't find it separately for download (capsman2 package).

WirelessRudy · Wed Jul 08, 2015 8:42 pm

Am I wrong or all_packages-smips-6.30.zip doesn't contain the wireless package? And I can't find it separately for download (capsman2 package).

*) removed wireless package from routeros bundle package,
new wireless-fp is left in place and wireless-cm2 added as option;

NicK · Wed Jul 08, 2015 8:42 pm

It continues with the bugs that I reported. Apparently, you're not behind the comments of members. This is sad.

WirelessRudy · Wed Jul 08, 2015 8:44 pm

I hope so many changes is going to work out.. Lets wait and see for two week if we will get a storm on this tread or just a mild breeze...

WirelessRudy · Wed Jul 08, 2015 8:46 pm

It continues with the bugs that I reported. Apparently, you're not behind the comments of members. This is sad.

That's not completely fair.. they must have worked hard to correct and introduce so many features. Some bugs just take time to figure out and fix. Maybe the are fixed in next release. Or ask them polite to their status.

NicK · Wed Jul 08, 2015 8:55 pm

It continues with the bugs that I reported. Apparently, you're not behind the comments of members. This is sad.
That's not completely fair.. they must have worked hard to correct and introduce so many features. Some bugs just take time to figure out and fix. Maybe the are fixed in next release. Or ask them polite to their status.

The bug is from the 6.29 version and has been reported in fixes topic 6.30 RC

They had to give priority to bugs than the news.

paoloaga · Wed Jul 08, 2015 8:58 pm

Am I wrong or all_packages-smips-6.30.zip doesn't contain the wireless package? And I can't find it separately for download (capsman2 package).
*) removed wireless package from routeros bundle package,
new wireless-fp is left in place and wireless-cm2 added as option;

I was meaning I can't find ANY wireless package: I don't find both wireless-fp and wireless-cm2. I don't think that SMIPS ever had the plain wireless package at all. Are the features included into the system package?

bajodel · Wed Jul 08, 2015 8:59 pm

What version am I running ?

vys · Wed Jul 08, 2015 9:02 pm

*) wireless - added WMM power save suport for mobile devices;

but I still get the following in the log from my iphone 5s:

20:55:22 wireless,info wlan1: data from unknown device 54:E4:3A:27:06:F3, sent deauth 
20:55:47 wireless,info wlan1: data from unknown device 54:E4:3A:27:06:F3, sent deauth 
20:55:47 wireless,info wlan1: data from unknown device 54:E4:3A:27:06:F3, sent deauth 
20:55:47 wireless,info wlan1: data from unknown device 54:E4:3A:27:06:F3, sent deauth 
20:55:47 wireless,info 54:E4:3A:27:06:F3@wlan1: connected 
20:57:45 wireless,info 54:E4:3A:27:06:F3@wlan1: disconnected, extensive data loss

6.30, wireless-fp enabled

WirelessRudy · Wed Jul 08, 2015 9:04 pm

It continues with the bugs that I reported. Apparently, you're not behind the comments of members. This is sad.
That's not completely fair.. they must have worked hard to correct and introduce so many features. Some bugs just take time to figure out and fix. Maybe the are fixed in next release. Or ask them polite to their status.
The bug is from the 6.29 version and has been reported in fixes topic 6.30 RC

They had to give priority to bugs than the news.

And that can't be the fact. If bugs always would have priority above new features we would still work with pigeons to send a message. They still have bugs.....
Technology advances because new features are introduce to do better. Bugs are always around and some bugs only affect some specific users where new features probably benefit more and put a manufacturer still in the market.
Its a balance....
But lets not clutter this tread. I'm waiting for the first reports of success or failures from the wide users base....

NicK · Wed Jul 08, 2015 9:08 pm

And that can't be the fact. If bugs always would have priority above new features we would still work with pigeons to send a message. They still have bugs.....
Technology advances because new features are introduce to do better. Bugs are always around and some bugs only affect some specific users where new features probably benefit more and put a manufacturer still in the market.
Its a balance....
But lets not clutter this tread. I'm waiting for the first reports of success or failures from the wide users base....

It seems that managed to solve, after upgrading the firmware of my RB750. I will wait to see if the problem returns because the bug began after 15h uptime.

#Edit

The bug is still continuing. Only occurs when the port is connected but unused. Example: Using switch unmanaged.

By using the door with a computer for example, displays the correct data from the same use. Now, if the same stop using the door, and turns off when passing +/- 5 minutes, back to send packets from one port to the other, while the other is not in use, is connected only.

xavierbt · Wed Jul 08, 2015 9:17 pm

Be aware, There is a new routerboard firmware version 3.24 !!!

xavierbt · Wed Jul 08, 2015 9:33 pm

After upgrading, the Quick Set has changed from 'Home AP' to 'WISP AP'. The Wiressless Clients is not refreshing the Signal Strength.

It's a 951G-2HnD.

avantwireless · Wed Jul 08, 2015 9:46 pm

Yes there is just a regular wireless package in the all packages zip file... You have to get it though the zip file... Unfortunately when you put it into the packages on the router it doesn't add it in. I was looking for help adding it in, if possible, not the obvious, "It's been replaced in the bundle" .. Yes I do read the release notes, and yes I do understand what they say. So if anyone knows the magic to make it usable instead of fp and cm, I will give you karma!

Taylor · Wed Jul 08, 2015 9:49 pm

Why would you want to use it instead of fp? Its basically the same but improvements. I've been using it for around a year now.

mediana · Wed Jul 08, 2015 9:58 pm

Script to update DynDNS will fail after upgrade to v6.30.
Downgrade to v6.29.1, the same script works.
Have no idea what happened in the" /tool fetch" process, it just stop there......

The script code:

:local ddnsuser "myname";
:local ddnspass "mypass";
:local theinterface "pppoe-myisp";
:local ddnshost "ddnshost.dyndns.org";

:local ipddns [:resolve $ddnshost];
:local ipfresh [ /ip address get [/ip address find interface=$theinterface ] address ];
:if ([ :typeof $ipfresh ] = nil ) do={
:log info ("DynDNS: No ip address on $theinterface .");
} else={
:for i from=( [:len $ipfresh] - 1) to=0 do={
:if ( [:pick $ipfresh $i] = "/") do={
:set ipfresh [:pick $ipfresh 0 $i];
}
}

:if ($ipddns != $ipfresh) do={
:log info ("DynDNS: IP-DynDNS = $ipddns");
:log info ("DynDNS: IP-Fresh = $ipfresh");
:log info "DynDNS: Update IP needed, Sending UPDATE...!";
:local str "/nic/update?hostname=$ddnshost&myip=$ipfresh&wildcard=NOCHG&mx=NOCHG&backmx=NOCHG";
/tool fetch address=members.dyndns.org src-path=$str mode=http user=$ddnsuser \
password=$ddnspass dst-path=("/DynDNS.".$ddnshost);
:delay 1;
:local str [/file find name="DynDNS.$ddnshost"];
/file remove $str;
:global ipddns $ipfresh;
:log info "DynDNS: IP updated to $ipfresh!";
} else={
:log info "DynDNS: Don't need changes";
}
}

manitonetworks · Wed Jul 08, 2015 10:00 pm

Update to RB751 seemed to go well, including the new firmware. I appreciate the new features and the bug fixes as well, but it would be nice if we could remove the dummy firewall rule and associated mangle rules introduced in this release.

I've been unable to remove or disable the filter rule from the command line or in Winbox ("cannot remove builtin" error), and I don't want to have to add this new rule to our firewall documentation, or explain to our auditors what this rule is and why it's suddenly appeared since our last audit. On the networks where we're required to be compliant, because there's been a firewall addition that we can't undo we'll also have to re-scan our networks for compliance.

bajodel · Wed Jul 08, 2015 10:01 pm

Yes there is just a regular wireless package in the all packages zip file... You have to get it though the zip file... Unfortunately when you put it into the packages on the router it doesn't add it in. I was looking for help adding it in, if possible, not the obvious, "It's been replaced in the bundle" .. Yes I do read the release notes, and yes I do understand what they say. So if anyone knows the magic to make it usable instead of fp and cm, I will give you karma!

Disable/uninstall -fp and cm2 pkgs ..reboot ..upload std wireless pkg.. reboot

bajodel · Wed Jul 08, 2015 10:16 pm

Script to update DynDNS will fail after upgrade to v6.30.
Downgrade to v6.29.1, the same script works.
.. cut..

Take a look here ... http://forum.mikrotik.com/viewtopic.php ... 04#p489204

lekozs · Wed Jul 08, 2015 10:33 pm

Torrent link http://www.mikrotik.com/download/router ... 30.torrent is ready ...but no seeders, yet...

visalink · Wed Jul 08, 2015 10:44 pm

And that can't be the fact. If bugs always would have priority above new features we would still work with pigeons to send a message. They still have bugs.....
Technology advances because new features are introduce to do better. Bugs are always around and some bugs only affect some specific users where new features probably benefit more and put a manufacturer still in the market.
Its a balance....
But lets not clutter this tread. I'm waiting for the first reports of success or failures from the wide users base....
It seems that managed to solve, after upgrading the firmware of my RB750. I will wait to see if the problem returns because the bug began after 15h uptime.

#Edit

The bug is still continuing. Only occurs when the port is connected but unused. Example: Using switch unmanaged.

By using the door with a computer for example, displays the correct data from the same use. Now, if the same stop using the door, and turns off when passing +/- 5 minutes, back to send packets from one port to the other, while the other is not in use, is connected only.

This also occurs to me.

shahbazian · Wed Jul 08, 2015 11:23 pm

I don't see "*) fixed kernel failure after queues configuration has been changed" in RouterOS v6.30 change log, but this is exist in v6.30rc
where is this changelog?

barkas · Wed Jul 08, 2015 11:26 pm

After installation on rb450g after boot it takes all interfaces offline and there is no way to get at it anymore.
Did netinstall, worked at first, but after configuration restore it is unreachable again. Then I reset the configuration but it's still unreachable.
Nothing on serial console either.
Now I'm downgrading to 6.29.1 again.

Lupin · Wed Jul 08, 2015 11:35 pm

How to use execute? (mAP)

[admin@MikroTik] > :execute script={interface print} file=test.txt
[admin@MikroTik] >

No error in sintax

[admin@MikroTik] > file print
 # NAME                        TYPE                              SIZE CREATION-TIME       
 0 flash                       disk                                   jan/01/1970 00:00:00
 1 pub                         directory                              jul/08/2015 20:13:55
 2 flash/skins                 directory                              jan/01/1970 00:00:03
[admin@MikroTik] >

but no file

shahbazian · Wed Jul 08, 2015 11:47 pm

Where is RouterBOOT v3.24 changelog in "http://wiki.mikrotik.com/wiki/RouterBOOT_changelog"?

noyo · Wed Jul 08, 2015 11:51 pm

What version am I running ?

in my rb711 is ok

Where is RouterBOOT v3.24 changelog in "http://wiki.mikrotik.com/wiki/RouterBOOT_changelog"?

+1

xavierbt · Thu Jul 09, 2015 12:17 am

After upgrading, the Quick Set has changed from 'Home AP' to 'WISP AP'. The Wiressless Clients is not refreshing the Signal Strength.

It's a 951G-2HnD.

Safari and Firefox reports errors executing the Quick Set page

[Error] Failed to load resource: http://192.168.75.2/jsproxy/?%00%00%00% ... 0%C2%8B¿ÖÑ
the server responded with a status of 404 (Not Found) (jsproxy, line 0)

[Error] TypeError: undefined is not a function (evaluating 'this.udpateRow({ufe0001:i,ufe0013:1})')
updateTable (engine-4c6d1994ff94.js, line 442)
load (engine-4c6d1994ff94.js, line 441)
update (engine-4c6d1994ff94.js, line 233)
lstn (engine-4c6d1994ff94.js, line 222)
notify (engine-4c6d1994ff94.js, line 299)
load (engine-4c6d1994ff94.js, line 460)
lstn (engine-4c6d1994ff94.js, line 458)
notify (engine-4c6d1994ff94.js, line 299)
onreply (engine-4c6d1994ff94.js, line 828)
decrypt (engine-4c6d1994ff94.js, line 53)
(función anónima) (engine-4c6d1994ff94.js, line 792)
onreadystatechange (engine-4c6d1994ff94.js, line 782)

paoloaga · Thu Jul 09, 2015 12:37 am

I was meaning I can't find ANY wireless package: I don't find both wireless-fp and wireless-cm2. I don't think that SMIPS ever had the plain wireless package at all. Are the features included into the system package?

I try to be more clear. I still can't find any wireless*.npk in the "extra packages" .zip file for the SMIPS architecture (hAP). Am I missing something?

G2Dolphin · Thu Jul 09, 2015 12:58 am

Please, add a checkbox for disabling dummy fasttrack rules.

regi · Thu Jul 09, 2015 1:26 am

upgraded from 6.27 (because of problems with l2tp/ipsec in 6.28 and 6.29)

on RB2011UAS-2HnD:
L2TP/IPSec still not working

20:24:42 ipsec,error key length mismatched, mine:128 peer:256. 
20:24:42 ipsec,error authtype mismatched: my:hmac-sha1 peer:hmac-md5

default configuration with IPSec enabled on L2TP Server

on RB1100AHx2:
Packet marking not working as in 6.27.
In 6.27 i can easy mark (in mangle table) incoming packet:
- protocol 50-ipsec -> mark 1
- 17-udp (port 4500) -> mark 2
Next using mark matcher i can create separate filter rulles for:
- L2TP/IPSec (mark 0x1)
- L2TP/IPSec with NAT-T (mark 0x2)
- non encrypted L2TP (no mark)

Now in 6.30 decrypted traffic has no mark assigned to IPSec packet in mangle table.

In 6.30 I can use IPSec Policy "in ipsec", but there is no possibility to separate native IPSec from NAT-T

mediana · Thu Jul 09, 2015 1:30 am

Script to update DynDNS will fail after upgrade to v6.30.
Downgrade to v6.29.1, the same script works.
.. cut..
Take a look here ... http://forum.mikrotik.com/viewtopic.php ... 04#p489204

Hi bajodel:

Thanks. Seems working.

mediana · Thu Jul 09, 2015 1:38 am

Another problem:

Site to Site IPsec is broken.
RB850Gx2 (Dynamic IP) to Vigor2920 (Fixed IP) don't work after upgrade to v6.30.
Haven't tried between Mikrotik routers yet.

kometchtech · Thu Jul 09, 2015 1:59 am

Where is RouterBOOT v3.24 changelog in "http://wiki.mikrotik.com/wiki/RouterBOOT_changelog"?

According to heard, is Protect RouterBOOT related fixes.

For your device only change is that protected routerboot was not working
correctly. Now it is. You can disable pin hole reset now.

StubArea51 · Thu Jul 09, 2015 2:35 am

Thanks MikroTik.

Just tested the new IPSEC key feature of EoIP and it works very well.....Great feature!

mrz · Thu Jul 09, 2015 9:01 am

upgraded from 6.27 (because of problems with l2tp/ipsec in 6.28 and 6.29)

on RB2011UAS-2HnD:
L2TP/IPSec still not working
Code: Select all
20:24:42 ipsec,error key length mismatched, mine:128 peer:256. 
20:24:42 ipsec,error authtype mismatched: my:hmac-sha1 peer:hmac-md5 
default configuration with IPSec enabled on L2TP Server

But error says it all you have mismatched auth types.

Now in 6.30 decrypted traffic has no mark assigned to IPSec packet in mangle table.

See these examples
http://wiki.mikrotik.com/wiki/Manual:IP ... ed_Traffic

mrz · Thu Jul 09, 2015 9:06 am

Site to Site IPsec is broken.
RB850Gx2 (Dynamic IP) to Vigor2920 (Fixed IP) don't work after upgrade to v6.30.
Haven't tried between Mikrotik routers yet.

*) ipsec - fail ph2 negitioation when initiator proposed key length
does not match proposal configuration;

Make sure key lengths are identical.

Thu Jul 09, 2015 9:09 am

#Edit

The bug is still continuing. Only occurs when the port is connected but unused. Example: Using switch unmanaged.

By using the door with a computer for example, displays the correct data from the same use. Now, if the same stop using the door, and turns off when passing +/- 5 minutes, back to send packets from one port to the other, while the other is not in use, is connected only.

@Nick
If there is a broadcast traffic, then it is normal behaviour that router sends it out to all connected bridge ports. You should check and verify it with packet sniffer on ether4 port.

In case you are not expecting router to do it, write email to MikroTik support with included supout.rif file to investigate this issue.

mrz · Thu Jul 09, 2015 9:10 am

How to use execute? (mAP)

[admin@MikroTik] > :execute script={interface print} file=test.txt
[admin@MikroTik] >

No error in sintax

[admin@MikroTik] > file print
 # NAME                        TYPE                              SIZE CREATION-TIME       
 0 flash                       disk                                   jan/01/1970 00:00:00
 1 pub                         directory                              jul/08/2015 20:13:55
 2 flash/skins                 directory                              jan/01/1970 00:00:03
[admin@MikroTik] >

but no file

Currently "file" is not working, we will add this feature correctly in next ROS release.

Thu Jul 09, 2015 9:16 am

I was meaning I can't find ANY wireless package: I don't find both wireless-fp and wireless-cm2. I don't think that SMIPS ever had the plain wireless package at all. Are the features included into the system package?
I try to be more clear. I still can't find any wireless*.npk in the "extra packages" .zip file for the SMIPS architecture (hAP). Am I missing something?

you are right. we will fix it. however, the package does exist for the "check for updates" procedure to work. It is just not in the ZIP.

Thu Jul 09, 2015 9:18 am

How to use execute? (mAP)

[admin@MikroTik] > :execute script={interface print} file=test.txt
[/quote]

known issue,  v6.30.1  will have fix.

Thu Jul 09, 2015 9:20 am

'execute' in not running scheduler.

known issue, v6.30.1 will have fix

Thu Jul 09, 2015 9:20 am

After upgrading, the Quick Set has changed from 'Home AP' to 'WISP AP'

could it be that you used regular wireless.npk before? now this package is discontinued, wireless-fp is now default. it is possible that changing the package changed the mode.

Thu Jul 09, 2015 9:20 am

What version am I running ?

please send us the supout.rif file to support@mikrotik.com so we can see what the system sees

Thu Jul 09, 2015 9:22 am

After upgrading, the Quick Set has changed from 'Home AP' to 'WISP AP'. The Wiressless Clients is not refreshing the Signal Strength.

It's a 951G-2HnD.

Safari and Firefox reports errors executing the Quick Set page

[Error] Failed to load resource: http://192.168.75.2/jsproxy/?%00%00%00% ... 0%C2%8B¿ÖÑ
the server responded with a status of 404 (Not Found) (jsproxy, line 0)

[Error] TypeError: undefined is not a function (evaluating 'this.udpateRow({ufe0001:i,ufe0013:1})')
updateTable (engine-4c6d1994ff94.js, line 442)
load (engine-4c6d1994ff94.js, line 441)
update (engine-4c6d1994ff94.js, line 233)
lstn (engine-4c6d1994ff94.js, line 222)
notify (engine-4c6d1994ff94.js, line 299)
load (engine-4c6d1994ff94.js, line 460)
lstn (engine-4c6d1994ff94.js, line 458)
notify (engine-4c6d1994ff94.js, line 299)
onreply (engine-4c6d1994ff94.js, line 828)
decrypt (engine-4c6d1994ff94.js, line 53)
(función anónima) (engine-4c6d1994ff94.js, line 792)
onreadystatechange (engine-4c6d1994ff94.js, line 782)

clean browser and proxy cache. the file is probably cached, and is looking for things that are no longer there.

solaoxo · Thu Jul 09, 2015 9:27 am

QQ截图20150709142253.jpg

QQ截图20150709142300.jpg

'fasttrack connection' closes, dynamic rules are not automatically deleted.

Thu Jul 09, 2015 9:50 am

QQ截图20150709142253.jpg

QQ截图20150709142300.jpg
'fasttrack connection' closes, dynamic rules are not automatically deleted.

in mi case rebooting clean the dummy rule rb951Ui

xavierbt · Thu Jul 09, 2015 10:28 am

After upgrading, the Quick Set has changed from 'Home AP' to 'WISP AP'
could it be that you used regular wireless.npk before? now this package is discontinued, wireless-fp is now default. it is possible that changing the package changed the mode.

This is what I have.

[admin@ROUTER] /system package> print
Flags: X - disabled
# NAME VERSION SCHEDULED
0 routeros-mipsbe 6.30
1 system 6.30
2 X wireless-cm2 6.30
3 X ipv6 6.30
4 wireless-fp 6.30
5 hotspot 6.30
6 dhcp 6.30
7 mpls 6.30
8 routing 6.30
9 ppp 6.30
10 security 6.30
11 advanced-tools 6.30

But I think that is related to the javascripts error that breaks the construction of the Quick Set window.
I have cleared the cache, tried from other browsers, etc. I have connected from another computer and browser never connected to the router with no avail.

[Error] Failed to load resource: http://192.168.75.2/jsproxy/?%00%00%00% ... IYÚ®%15%06 the server responded with a status of 404 (Not Found) (jsproxy, line 0)

[Error] TypeError: undefined is not a function (evaluating 'this.udpateRow({ufe0001:i,ufe0013:1})')
updateTable (engine-4c6d1994ff94.js, line 442)
load (engine-4c6d1994ff94.js, line 441)
update (engine-4c6d1994ff94.js, line 233)
lstn (engine-4c6d1994ff94.js, line 222)
notify (engine-4c6d1994ff94.js, line 299)
load (engine-4c6d1994ff94.js, line 460)
lstn (engine-4c6d1994ff94.js, line 458)
notify (engine-4c6d1994ff94.js, line 299)
onreply (engine-4c6d1994ff94.js, line 828)
decrypt (engine-4c6d1994ff94.js, line 53)
(función anónima) (engine-4c6d1994ff94.js, line 792)
onreadystatechange (engine-4c6d1994ff94.js, line 782)

Thu Jul 09, 2015 10:30 am

[Error] Failed to load resource: http://192.168.75.2/jsproxy/?%00%00%00% ... IYÚ®%15%06 the server responded with a status of 404 (Not Found) (jsproxy, line 0)

I can repeat the issue, thanks, looking into it.

marrold · Thu Jul 09, 2015 10:56 am

*) wireless - added WMM power save suport for mobile devices;

This interesting, is there any way to view if it's being utilised, buffers etc?

lordkappa · Thu Jul 09, 2015 11:43 am

*) wireless - added WMM power save suport for mobile devices;

Wow. I've been forced to run Ubiquiti APs for the last 3 years due to exactly this.

MrOrkyZ · Thu Jul 09, 2015 12:11 pm

Hi . script dont working in 6.30 . At 6.29.1 all work fine

:foreach i in=[/ip firewall connection find dst-address~":55777" protocol~"udp"] do={ /ip firewall connection remove $i };

Thu Jul 09, 2015 12:15 pm

[Error] Failed to load resource: http://192.168.75.2/jsproxy/?%00%00%00% ... IYÚ®%15%06 the server responded with a status of 404 (Not Found) (jsproxy, line 0)

This error is unrelated to HomeAP changing to WispAP. This change can only happen if you change wireless mode, channel width or some other wireless setting. Did you change something outside QuickSet?

jza · Thu Jul 09, 2015 12:31 pm

I upgraded my 951G from 6.28 to 6.30 and all my NAT rules were missing. The default one as well. Friend of mine did the same with Basebox and same result, however in this case rules were there but interfaces went to unknown state.

This seems to be a quite big problem, since it's a suicide to upgrade any router from a distance.
751G took upgrade without any problems. Could it be something with hotspot? These ones that broke did had hotspot running. 751G did not.

Thu Jul 09, 2015 12:45 pm

I upgraded my 951G from 6.28 to 6.30 and all my NAT rules were missing. The default one as well. Friend of mine did the same with Basebox and same result, however in this case rules were there but interfaces went to unknown state.

This seems to be a quite big problem, since it's a suicide to upgrade any router from a distance.
751G took upgrade without any problems. Could it be something with hotspot? These ones that broke did had hotspot running. 751G did not.

Could you make a supout.rif file and email it to support@mikrotik.com with brief problem description?

khatab · Thu Jul 09, 2015 1:03 pm

Dear All,

Still there is a problem with SNTP- client, I tried so may NTP server ip but the clock still not Syncing, any one else facing this problem?

Thu Jul 09, 2015 1:07 pm

Dear All,

Still there is a problem with SNTP- client, I tried so may NTP server ip but the clock still not Syncing, any one else facing this problem?

what is the message in SNTP status? did you try other servers?

paoloaga · Thu Jul 09, 2015 1:24 pm

I try to be more clear. I still can't find any wireless*.npk in the "extra packages" .zip file for the SMIPS architecture (hAP). Am I missing something?
you are right. we will fix it. however, the package does exist for the "check for updates" procedure to work. It is just not in the ZIP.

We already deployed over than a hundred hAP units, not all have direct internet access and the automated procedure doesn't involve "check for updates". I'll wait to find the wireless package into the zip file, thanks.

mikruser · Thu Jul 09, 2015 1:48 pm

CCR1009-8G-1S-1S+
after upgrade to v6.30
traffic through IPsec tunnel stops for ~30 min every ~2 hour

upgrade630.png

regi · Thu Jul 09, 2015 2:19 pm

upgraded from 6.27 (because of problems with l2tp/ipsec in 6.28 and 6.29)

on RB2011UAS-2HnD:
L2TP/IPSec still not working
Code: Select all
20:24:42 ipsec,error key length mismatched, mine:128 peer:256. 
20:24:42 ipsec,error authtype mismatched: my:hmac-sha1 peer:hmac-md5 
default configuration with IPSec enabled on L2TP Server
But error says it all you have mismatched auth types.

Yes, that's what it says

but it's not true

only sha1 is selected in proposal on mikrotik, and i use windows client with default settings (sha1+3des, sha1+aes128). same problem is when i use android 5.1.1 client.
after downgrade to 6.27 everything works fine

Now in 6.30 decrypted traffic has no mark assigned to IPSec packet in mangle table.
See these examples
http://wiki.mikrotik.com/wiki/Manual:IP ... ed_Traffic

thx. i saw that, but my configuration is very complicated (IPSec traffic filtering) and
I need packet marking work as previously to meet company requirements.
Have this functionality been removed? or it's just a bug?

mikko · Thu Jul 09, 2015 2:23 pm

LTE interface number changes in update. I have many Baseboxes with Sierra MC7710 LTE card in remote locations. During update LTE1 interface changes to LTE2 interface. This breaks up DHCP client and firewall rules that are still trying to use LTE1 interface after update. This prevents updating remote sites as connection won't come up after update automatically. So beware if you are upgrading remote site with LTE connection.

I have sent supouts to support and waiting their reply.

mrz · Thu Jul 09, 2015 2:36 pm

Now in 6.30 decrypted traffic has no mark assigned to IPSec packet in mangle table.
See these examples
http://wiki.mikrotik.com/wiki/Manual:IP ... ed_Traffic
thx. i saw that, but my configuration is very complicated (IPSec traffic filtering) and
I need packet marking work as previously to meet company requirements.
Have this functionality been removed? or it's just a bug?

It was removed to fix other bugs. You can still use DSCP and priority instead.

w32pamela · Thu Jul 09, 2015 3:06 pm

Groove 52Hpn new from box. Upgraded from v5.26 to v6.30. Webfig Quick Set page freezes when I try to connect to an AP. Chrome, Firefox, IE etc.

v6.30 is working great in Winbox

gius64 · Thu Jul 09, 2015 4:08 pm

You forgot to mention that when you add a user in user-manager the username field was named "name" before.
In 6.30 its name changed in "username".
You should mention it in the changelog.

Using /tool user-manager user print the username is not printed anymore

Also, IPsec+L2TP VPN stopped working on CCR1036-12G-4S after upgrading from 6.27 to 6.30. It worked using iPhone and Linux as client, now it doesn't work anymore and I'm trying to configure an OpenVPN instead. PPTP works but it's old and insecure...

mrz · Thu Jul 09, 2015 4:19 pm

Also, IPsec+L2TP VPN stopped working on CCR1036-12G-4S after upgrading from 6.27 to 6.30. It worked using iPhone and Linux as client, now it doesn't work anymore and I'm trying to configure an OpenVPN instead. PPTP works but it's old and insecure...

As I mentioned before in this topic, make sure that key length match on server and client.

gius64 · Thu Jul 09, 2015 4:24 pm

@mrz:
Please release a fix urgently for user-manager. It broken all my API scripts, but I want 6.30 fixes on hardware watchdog for CCR, so I don't want to downgrade! Please!

About L2TP+IPsec, I don't have a key lenght settings on the iPhone or on the Mac. It worked before...

Thu Jul 09, 2015 4:26 pm

thx. i saw that, but my configuration is very complicated (IPSec traffic filtering) and
I need packet marking work as previously to meet company requirements.
Have this functionality been removed? or it's just a bug?

Can you provide an example of what can not be achieved with IPsec policies, IPsec policy templates and this new ipsec-policy traffic matcher in firewall?

mrz · Thu Jul 09, 2015 4:36 pm

@glus64: default on mikrotik is aes-128, most likely your devices use different length. Maybe 256, just add aes-256 in ipsec proposal. It worked previously because of security issue

*) ipsec - fail ph2 negitioation when initiator proposed key length
   does not match proposal configuration;

cnt · Thu Jul 09, 2015 4:38 pm

After upgrade to 6.30 version on (homemade) x86 router the system freeses. After reseting the router works some time but when speed growe on interface (bigger of 10Mb/s), the router freese. We try to chnaged Network cards, don't help. After downgrade to 6.29.1 everything works ok.

mandrade · Thu Jul 09, 2015 5:01 pm

I was meaning I can't find ANY wireless package: I don't find both wireless-fp and wireless-cm2. I don't think that SMIPS ever had the plain wireless package at all. Are the features included into the system package?
I try to be more clear. I still can't find any wireless*.npk in the "extra packages" .zip file for the SMIPS architecture (hAP). Am I missing something?
you are right. we will fix it. however, the package does exist for the "check for updates" procedure to work. It is just not in the ZIP.

Still waiting for fix on download packages....

Thu Jul 09, 2015 5:09 pm

Maybe still cached? Because files were added some 8 hours ago:
http://mikrotikdownload.s3.amazonaws.co ... -smips.npk

hngjared · Thu Jul 09, 2015 5:10 pm

*) wireless - added WMM power save suport for mobile devices;

Is this enabled by default?

Does "WMM support" need to be set to "enabled" in the wlan settings?

mandrade · Thu Jul 09, 2015 5:19 pm

Maybe still cached? Because files were added some 8 hours ago:
http://mikrotikdownload.s3.amazonaws.co ... -smips.npk

Here i did the download via torrent, and missing cm2 and fp package of skips *AND* cm2 for tile too...

BR

avenn · Thu Jul 09, 2015 5:42 pm

Hi,

Two questions. One for users and One for Mikrotik

For Users:
I know it’s an early question considering just released but has anybody tried the wireless ac improvements? I am about to roll out a new ac site and interested in feedback/problems/improvements?

For Mikrotik/Normis:

Is this the start of new release approach?
Major 6.30/6.31/6.32…..
Fixes 6.30.1/2/3….

Kind Regards

Aidan Venn

hngjared · Thu Jul 09, 2015 5:47 pm

*) wireless - added WMM power save suport for mobile devices;

Is this enabled by default?

Does "WMM support" need to be set to "enabled" in the wlan settings?

After testing I don't see any noticeable improvement. When I have my device in power saving mode latency jumps into the 1000's with significant packet loss. As soon as I change my device back to max performance all is well again. I tested with "WMM support" enabled, disabled and required. Is there something I'm missing? Appreciate any feedback.

kristaps · Thu Jul 09, 2015 5:53 pm

best results we achieved with AR9200 chips and newer and n /ac clients,
also try to look for frequencies with less interference.

khatab · Thu Jul 09, 2015 5:57 pm

Dear All,

Still there is a problem with SNTP- client, I tried so may NTP server ip but the clock still not Syncing, any one else facing this problem?
what is the message in SNTP status? did you try other servers?

Yes, I tried other IP servers, still not working, the time is not syncing..

hngjared · Thu Jul 09, 2015 6:00 pm

Dear All,

Still there is a problem with SNTP- client, I tried so may NTP server ip but the clock still not Syncing, any one else facing this problem?
what is the message in SNTP status? did you try other servers?
Yes, I tried other IP servers, still not working, the time is not syncing..

I haven't had any problems with SNTP since the upgrade.

maxpower · Thu Jul 09, 2015 6:02 pm

After upgrading CCR1009-8G-1S-1S+ from 6.28 to 6.30 I have problem with multiple IP Addresses on one VLAN interface, only first IP is replying to ping. Any workaround for that?

If I use downgrade to 6.28 all config will be erased? I've also updated firmware from 3.22 to 3.24 -- is this sensitive to downgrade?

avenn · Thu Jul 09, 2015 6:19 pm

Hi,

Netinstall wireless-cm2 on hex lite for testing. Default select of cm2 auto ticks system not routeros-mipsbe? Cannot change. This is my first netinstall on hex lite so maybe normal or change has occurred in 6.30?

Anybody have any info if this is normal (for hex) or new?

cheers

Aidan Venn

visalink · Thu Jul 09, 2015 8:27 pm

Please, add a checkbox for disabling dummy fasttrack rules.

+1, please.

Thu Jul 09, 2015 8:41 pm

We don't even list the torrent on the download page, so fixes were not applied to it. Please download from our homepage

Sent from my iPhone using Tapatalk

Thu Jul 09, 2015 8:42 pm

Hi,

Two questions. One for users and One for Mikrotik

For Users:
I know it’s an early question considering just released but has anybody tried the wireless ac improvements? I am about to roll out a new ac site and interested in feedback/problems/improvements?

For Mikrotik/Normis:

Is this the start of new release approach?
Major 6.30/6.31/6.32…..
Fixes 6.30.1/2/3….

Kind Regards

Aidan Venn

Yes, the next will be with fixes only, 6.30.1

Sent from my iPhone using Tapatalk

maxpower · Thu Jul 09, 2015 10:07 pm

Was there any fix in 6.30 preventing from using multiple IP addresses on a single interface with same subnet mask?
I've had two addresses 10.10.0.2/29 & 10.10.0.5/29 -- for 6 months it was working fine (on 6.23-6.28) after upgrade to 6.30 it stopped.
In torch I see incoming ICMP to 10.10.0.5 but no reply is send to it. After I've changed mask to /32 on the second IP, it seems like it is working...

bajodel · Thu Jul 09, 2015 10:34 pm

What version am I running ?
please send us the supout.rif file to support@mikrotik.com so we can see what the system sees

I've managed to fix it ..downgraded to 6.30rc28 and than upgraded to 6.30 final (downgrading to 6.30rc29 didn't fix the problem).
Now I can correctly see version 6.30 on "System Resources" and also in "Partition Section" (and everywhere).

[note refresh

] Maybe I'm blind ..but I still miss the Extra Packages for x86 on mikrotik download page ..

gius64 · Fri Jul 10, 2015 12:12 am

@glus64: default on mikrotik is aes-128, most likely your devices use different length. Maybe 256, just add aes-256 in ipsec proposal. It worked previously because of security issue
Code: Select all
*) ipsec - fail ph2 negitioation when initiator proposed key length
   does not match proposal configuration;

Thank you! Problem solved

Please fix user-manager issue, it's driving me crazy

lotnybartek · Fri Jul 10, 2015 12:17 am

DynDNS update script not working after update to 6.30

# Set needed variables
:local username "XXXXXX"
:local password "XXXXXX"
:local hostname "XXXXXX.dyndns.org"

:global dyndnsForce ""
:global previousIP "$previousIP"

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

# get the current IP address from the internet (in case of double-nat)
/tool fetch mode=http address="checkip.dyndns.org" src-path="/" dst-path="/dyndns.checkip.html"
:delay 1
:local result [/file get dyndns.checkip.html contents]

# parse the current IP result
:local resultLen [:len $result]
:local startLoc [:find $result ": " -1]
:set startLoc ($startLoc + 2)
:local endLoc [:find $result "</body>" -1]
:local currentIP [:pick $result $startLoc $endLoc]
:log info "UpdateDynDNS: 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 DynDNS!

#:set dyndnsForce true

# Determine if dyndns update is needed
# more dyndns updater request details http://www.dyndns.com/developers/specs/syntax.html

:if (($currentIP != $previousIP) || ($dyndnsForce = true)) do={
:set dyndnsForce false
:set previousIP $currentIP
:log info "$currentIP or $previousIP"
/tool fetch user=$username password=$password mode=http address="members.dyndns.org" \
src-path="nic/update?system=dyndns&hostname=$hostname&myip=$currentIP&wildcard=no" \
dst-path="/dyndns.txt"
:delay 1
:local result [/file get dyndns.txt contents]
:log info ("UpdateDynDNS: Dyndns update needed")
:log info ("UpdateDynDNS: Dyndns Update Result: ".$result)
:put ("Dyndns Update Result: ".$result)
} else={
:log info ("UpdateDynDNS: No dyndns update needed")
}

I have two 2011UiAS-2HnD.

One has 6.29 and script works fine, second has 6.30 and script is not updating DynDNS with new IP.

docmarius · Fri Jul 10, 2015 1:19 am

It seems IPv6 ND/RA is broken in this version.
Windows hosts don't get a IPv6 address via RA (the interface got an correct address from the pool, but that's all).
Also I get an invalid dynamic prefix in ND which shouldn't be there (from another box running also 6.30).
Reverting to 6.29.1 fixes the issue.

michaelcarey · Fri Jul 10, 2015 4:33 am

I added the IPsec key option to my EoIP tunnel interfaces... and it seem to break the L2TP server.

EoIP works fine with IPsec.

"Windows error 792: The L2TP connection attempt failed because security negotiations timed out."

Is an IPsec enhanced EoIP tunnel mutually exclusive with an L2TP VPN server?

I remove the IPSec secret on each end and the L2TP VPN server works fine again and client can connect.

MTeeker · Fri Jul 10, 2015 8:01 am

DynDNS update script not working after update to 6.30
One has 6.29 and script works fine, second has 6.30 and script is not updating DynDNS with new IP.

Not sure if you are aware but there is a built-in dynamic DNS feature using the router's serial number.

/IP Cloud.

Neat and simple.
Cheers

Kindis · Fri Jul 10, 2015 10:19 am

Hi,

Two questions. One for users and One for Mikrotik

For Users:
I know it’s an early question considering just released but has anybody tried the wireless ac improvements? I am about to roll out a new ac site and interested in feedback/problems/improvements?

For Mikrotik/Normis:

Is this the start of new release approach?
Major 6.30/6.31/6.32…..
Fixes 6.30.1/2/3….

Kind Regards

Aidan Venn
Yes, the next will be with fixes only, 6.30.1

Sent from my iPhone using Tapatalk

This is great but I have a follow up question which I did not manage to ask before the other tread was locked

If you change this will there be more than one channel/branch regarding updates?
I would love to have a long time service branch which will give support for a year or so. Only fixes and update to this.
Then also a beta branch which is voluntary to use and is not pushed via auto update.

I think this will give more options for us users. Those who want to be on the bleeding edge can run beta builds and those that can wait for functions and add-ons run the stable version.

Would also be awesome if you in /system/packages the check for update there will be an option for which channel/branch you would like to use. Default is stable but we can change to beta (or what you want to call it).
Would be easy to see then which build you use.

Fri Jul 10, 2015 10:30 am

Yes, we plan to make something like that eventually. Download page will also list both, latest fix-only releas and the latest feature release

zyzelis · Fri Jul 10, 2015 10:43 am

Hello,
Yesterday upgraded whole network to 6.30 from different versions (mainly from 6.27).
devices upgraded(~200 devices): CCR, sxt, sxt-ac, 433, 2011,951,711,750 and omnitiks.
Upgrade process was clean, without surprises and all devices went back without any issues.
Services used: bpg, routing ,firewalling, vlans, gre tunnels, wireless 2G/5G/ac bridging.
So far so good
PS. I like fastpath for bridges it lowered AP cpu consumption.

mrz · Fri Jul 10, 2015 10:48 am

I added the IPsec key option to my EoIP tunnel interfaces... and it seem to break the L2TP server. EoIP works fine with IPsec.

"Windows error 792: The L2TP connection attempt failed because security negotiations timed out."

Is an IPsec enhanced EoIP tunnel mutually exclusive with an L2TP VPN server?

I remove the IPSec secret on each end and the L2TP VPN server works fine again and client can connect.

EoIP and L2Tp connects from the same IP?

tboernert · Fri Jul 10, 2015 11:00 am

What's new in 6.30 (2015-Jul-08 09:07):

*) wireless - improve ac radio coexistence with other wireless clients, optimized
transmit times to not interfere with other devices;

Is there any detail on that feature ?

is it now possible to sync the tdma timing on all devices on a tower ? (as for example mimosa it does ? http://mimosa.co/tdma)

Thomas

WirelessRudy · Fri Jul 10, 2015 11:26 am

Hello,
Yesterday upgraded whole network to 6.30 from different versions (mainly from 6.27).
devices upgraded(~200 devices): CCR, sxt, sxt-ac, 433, 2011,951,711,750 and omnitiks.
Upgrade process was clean, without surprises and all devices went back without any issues.
Services used: bpg, routing ,firewalling, vlans, gre tunnels, wireless 2G/5G/ac bridging.
So far so good
PS. I like fastpath for bridges it lowered AP cpu consumption.

Ok, that can be good news. Because 6.29.1 works out to be a disaster on my network....

How is the uptime now for your devices?
Can you give us an update on the status later today?

I am forcefully updating units on 6.29.1 now to 6.30. It looks the network becomes more stable (port flapping, traffic stop to pass through ethernet interfaces, routers just crash after hours or days without problems..)
But since in 6.29.1 it could take up to some days before a crash occurs I'd really hope 6.30 is keeping units 'up' for much longer than that. Keep us updated please!

Fri Jul 10, 2015 12:07 pm

Bridge fastpath is great thing. Thanks for that.

But, can it be related individually to each bridge, if there is more than one bridge? For example, having one bridge with firewall rules and other bridge without. Is there possibility to have fastpath active on the bridge without firewall rules even there is other bridge using firewall rules?

Maybe I should post it in the feature request section... hope you dont mind.

pchott · Fri Jul 10, 2015 12:13 pm

Is it safe to upgrade also CRS226?

http://forum.mikrotik.com/viewtopic.php ... 5&p=489527

WirelessRudy · Fri Jul 10, 2015 12:20 pm

Just had one rb911G-5HPnd crash 2 hours after upgrade from 6.29.1 to 6.30; kernel failure...

On 2.29.1 we ocurred ethernet ports flapping, now that still goes on and on top a kernel failure. After the reboot the link towards this unit becomes very unstable with lots of package losses.....

WirelessRudy · Fri Jul 10, 2015 12:42 pm

One rb433AH updated this morning fm v6.29.1 to 6.30 to fight interface port stops receiving traffic from rb750UP behind it.
After the upgrade unit runs for some other and port still stops passing traffic. Send support to MT.

One more (port) crash and I have to revert to to 6.27 again... pfff

zyzelis · Fri Jul 10, 2015 12:58 pm

How is the uptime now for your devices?

1-2 days (since release). the most important was CCR. Now, im glad that it works without any issues, because before
i had problem with bug in queue trees, which stopped accidently traffic shaping with versions >6.19.

Can you give us an update on the status later today?

Sure. Maybe you interesting with more specific details?

I am forcefully updating units on 6.29.1 now to 6.30. It looks the network becomes more stable (port flapping, traffic stop to pass through ethernet interfaces, routers just crash after hours or days without problems..)
But since in 6.29.1 it could take up to some days before a crash occurs I'd really hope 6.30 is keeping units 'up' for much longer than that. Keep us updated please!

I had few devices with 6.29.1 (2011, not very important omnitik and few SXT's. No problems was found.
OK i will let you know.

bajodel · Fri Jul 10, 2015 1:41 pm

Tested 6.30 update on some (2x RB1100x2, 4x RB2011, 1x RB912G, half-dozen of RB912Ui) ..without any particular problems

Only on one RB2011 the update had some oddness (showed 6.30 on system resource and previous version on other config sections), but it's a heavy testing box coming from every 6.30rcXX step (and probably also all the previous 6.2[789]rcXX

mikko · Fri Jul 10, 2015 2:08 pm

LTE interface number changes in update. I have many Baseboxes with Sierra MC7710 LTE card in remote locations. During update LTE1 interface changes to LTE2 interface. This breaks up DHCP client and firewall rules that are still trying to use LTE1 interface after update. This prevents updating remote sites as connection won't come up after update automatically. So beware if you are upgrading remote site with LTE connection.

Support confirmed bug with Sierra MC7710 card changing it's interface name in 6.30 upgrade. Should be fixed to v6.30.1 and v6.31rc4.

michaelcarey · Fri Jul 10, 2015 2:23 pm

EoIP and L2Tp connects from the same IP?

Yes... the client I used for testing is "behind" one of the routers that hosts the EoIP endpoint, NAT would make it appear to have the public IP address of the router. The public IP addresses are used at each end of EoIP.

mrz · Fri Jul 10, 2015 2:27 pm

Then you have to set up ipsec peer manually and do not use new feature.

Fri Jul 10, 2015 3:11 pm

Then you have to set up ipsec peer manually and do not use new feature.

Can you please provide more technical details regarding this limitation?

mrz · Fri Jul 10, 2015 3:29 pm

If I understood correctly then just look at dynamically created ipsec peers.
one should be with address 0.0.0.0 other with eoips remote address. But since L2TP client uses the same IP address then of course peer setting with more specific IP will be used not 0.0.0.0.

So obviously in such setups you have to make static peer config to suite your needs.

eworm · Fri Jul 10, 2015 4:26 pm

My devices report:

certificate,error CRL size exceeds free memory, aborting update

I have CAcert certificates for https, in addition Class 1 PKI Key and Class 3 PKI Key. Probably the devices try to download and use these certificate revokation lists:
* https://crl.cacert.org/revoke.crl
* https://crl.cacert.org/class3-revoke.crl

michaelcarey · Fri Jul 10, 2015 4:42 pm

If I understood correctly then just look at dynamically created ipsec peers.
one should be with address 0.0.0.0 other with eoips remote address. But since L2TP client uses the same IP address then of course peer setting with more specific IP will be used not 0.0.0.0.

So obviously in such setups you have to make static peer config to suite your needs.

I have access to several public IP addresses at each location. Can I use one IP for the IPsec EoIP tunnel and a different one for the L2TP server?

Chupaka · Fri Jul 10, 2015 4:45 pm

I have access to several public IP addresses at each location. Can I use one IP for the IPsec EoIP tunnel and a different one for the L2TP server?

yep, that should work

michaelcarey · Fri Jul 10, 2015 5:05 pm

yep, that should work

Thanks for that... I'll give it a go.

mrz · Fri Jul 10, 2015 5:07 pm

My devices report:
certificate,error CRL size exceeds free memory, aborting update
I have CAcert certificates for https, in addition Class 1 PKI Key and Class 3 PKI Key. Probably the devices try to download and use these certificate revokation lists:
* https://crl.cacert.org/revoke.crl
* https://crl.cacert.org/class3-revoke.crl

How much memory does your router have (what RB model)? Both files in total requires at least 8MB RAM just to download.

ebreyit · Fri Jul 10, 2015 5:26 pm

Torrent link http://www.mikrotik.com/download/router ... 30.torrent is ready ...but no seeders, yet...

Seeding away merrily now from my Kimsufi box

We don't even list the torrent on the download page, so fixes were not applied to it. Please download from our homepage

Why the move away from Torrents. I keeps a lot of updates and open source software available via bittorrent on one of my servers to help ensure other have quick and easy access to things I also find useful. Also helps take the pressure of their servers.

aleprolit · Fri Jul 10, 2015 6:06 pm

User manager problem (PPC)
/tool user-manager user> print
...
... 0 ;;; may/30/2015 10G by ...t
customer=xxx actual-profile="10G"
password="zzz" first-name="zzz" last-name="zzz"
phone="3706xxxx" location="xxx"
email="aa@com.aa" shared-users=1
wireless-psk="" wireless-enc-key="" wireless-enc-algo=none
uptime-used=24w1d15h38s download-used=350335579097
upload-used=34653664424 last-seen=jul/10/2015 15:00:34
active-sessions=1

not included parameter
name="name"
So API not work.
After downgrading all OK.

upower3 · Fri Jul 10, 2015 6:46 pm

I really don't know the reason but as I compare 6.30rc... and 6.30 release I can say release uses much more CPU on the same load.

Here is the graphs of cpu for CRS125-24G-1S. First graph for the last day (24h) and in the night there was ROS upgrade to 6.30 release. The second graph is a day before, the same firmware but ROS is 6.30rc30. I can see difference easily. Really interested what's the reason for:

Last day:
http://prntscr.com/7r3yd0

Day before that day:
http://prntscr.com/7r3yu4

Any way to lower CPU again without reverting to rc?

Chupaka · Fri Jul 10, 2015 7:10 pm

Average CPU load 29% vs 31%?..

Trackboy · Fri Jul 10, 2015 7:55 pm

The speed is very bad with 6.30. The Ethernet and wireless too. Where i had 170 Mbit/sec now i have 80-90.

TheRealJLH · Fri Jul 10, 2015 10:26 pm

I just upgraded one of my CRS125-24G-1S Switches to 6.30 hoping it would resolve some issues i noticed since upgrading from 6.27 sadly this is not he case.

issue 1. I am not sure why but my CPU utilization spikes 100% and spends its time around 10% on this switch there is only about 50-60Mbit of traffic.
yes there a way I could view the process list and find out what is causing the high cpu spikes. ?

Edited( Looked like the process that is eating up the CPU time is called Management is there anyway I can find out why it is doing so. )

Issue 2. when i mirror a port to dump to wireshark or use torch to vew packet activity, I am unable to see interface TX Traffic.

eworm · Sat Jul 11, 2015 1:20 am

My devices report:
certificate,error CRL size exceeds free memory, aborting update
I have CAcert certificates for https, in addition Class 1 PKI Key and Class 3 PKI Key. Probably the devices try to download and use these certificate revokation lists:
* https://crl.cacert.org/revoke.crl
* https://crl.cacert.org/class3-revoke.crl

I've see my RB750GL update the CRL... So this is not a general problem with the release.

How much memory does your router have (what RB model)? Both files in total requires at least 8MB RAM just to download.

/system resource print 
                   uptime: 1d23h42m24s
                  version: 6.30
               build-time: Jul/08/2015 09:07:45
              free-memory: 8.9MiB
             total-memory: 128.0MiB
                      cpu: MIPS 74Kc V4.12
                cpu-count: 1
            cpu-frequency: 600MHz
                 cpu-load: 8%
           free-hdd-space: 105.9MiB
          total-hdd-space: 128.0MiB
  write-sect-since-reboot: 151718
         write-sect-total: 25148376
               bad-blocks: 0.2%
        architecture-name: mipsbe
               board-name: CRS125-24G-1S
                 platform: MikroTik

No idea what eats my memory...

Do I have to be concerned about the sector writes?

ufm · Sat Jul 11, 2015 4:10 am

In 6.30 broken traffic flow.
Field "Input interface" always contain 0.

WirelessRudy · Sat Jul 11, 2015 10:29 am

Screenshot 2015-07-11 09.24.03.png

Some 17-18 hours after upgrading rb911G-5HPnD from 2.29.1 to 2.30 had a system reboot due kernel failure.

Also noticed (seen it before on some other board now I think of it) that the log is not in chronologic order.... weird....

zyzelis · Sat Jul 11, 2015 11:38 am

@WirelessRudy
I checked important network nodes and:
AP's - 433GL+52Hn, SXT's SA/ac, OmniTiks,
Routers - CCR1036, 2011, 750G
Switches - 2011,750UP
All devices are up since upgrade. No abnormal CPU consumption, no unexpected reboot,no unexpected link downs,
No wireless degradation
Nice

rextended · Sat Jul 11, 2015 11:39 am

if on EoIP "keepalive=10,10" the system export as "!keepalive"

please, what is this on 1100AHx2???
/system resource irq rps
set ether1 disabled=yes
set ether2 disabled=yes
set ether3 disabled=yes
set ether4 disabled=yes
set ether5 disabled=yes
set ether6 disabled=yes
set ether7 disabled=yes
set ether8 disabled=yes
set ether9 disabled=yes
set ether10 disabled=yes
set ether11 disabled=yes

WirelessRudy · Sat Jul 11, 2015 1:13 pm

This morning rb750UP OR rb433AH stops passing traffic again. Its the rb433AH that has ether2 port everytime again stopping receiving any traffic (tx=0) and each time need to reboot that unit to bring the network back.
After it there either are still a lot of package losses and the port goes down in a short while again, or the package losses are minor and the port stays up for hours...

But since I suspect the rb750 connected to that port giving the problem I'm downgrading that one first. See if that helps....back to 6.27...

Sat Jul 11, 2015 1:48 pm

Please, add a checkbox for disabling dummy fasttrack rules.
+1, please.

Vote against dummy rules in firewall: http://forum.mikrotik.com/viewtopic.php?f=1&t=98471

WirelessRudy · Sat Jul 11, 2015 2:59 pm

This morning rb750UP OR rb433AH stops passing traffic again. Its the rb433AH that has ether2 port everytime again stopping receiving any traffic (tx=0) and each time need to reboot that unit to bring the network back.
After it there either are still a lot of package losses and the port goes down in a short while again, or the package losses are minor and the port stays up for hours...

But since I suspect the rb750 connected to that port giving the problem I'm downgrading that one first. See if that helps....back to 6.27...

Well, that didn't help. Unstable ethernet, almost impossible to loginto the rb750
Downgraded the rb433AH in front of it. (I enter via wireless backhaul into this rb433AH and ether2 is connected to this rb750. ether1 is connected to another rb750 with similar config and v.6.30 running as the one I now downgraded. It is always ether2 of the rb433AH that stops rx data. Tried to swap to ether3 but that gave the same result before. Hence I'd suspected the rb750....)
After downgrade to 6.27 again links are back stable again. <1% package loss over 3x wireless backhaul and 9 nodes....
I can breath again....

But for how long? Update to v.6.30 yesterday gave me peace the first 18 hours. Than link started to become unstable again resulting in multiple port crashed (and one kernel crash at another node in the route towards gateway).

Since I use in almost al boards some sort of bridge to bridge ports (wireless to wire or wire to wire or wireless to wireless) to thus create sort of 'bridged' network route towards end AP's I was hoping the new fastpath features for bridging was going to improve my throughputs. And in fact I got the feeling that was indeed happening. But the collateral damage is too big to deal with. So back to 6.27 if my links stay stable now.

And if not I presume I have to start changing routers. When is MT finally making some rb980UP with gigabit ports?
Now I have to start finding some workarounds with other hardware now....
This is a remote tower with 8 radio's and running on 24v/Solar that I want to be in control with watchdog etc.

Camis · Sat Jul 11, 2015 3:21 pm

After upgrading, the Quick Set has changed from 'Home AP' to 'WISP AP'. The Wiressless Clients is not refreshing the Signal Strength.

It's a 951G-2HnD.

Confirm, i have 951G-2HnD too.

I can't change from WISP AP to Home AP. After change winbox automatic revert to WISP AP.

Please help with this.

Another problem is with Android smartfones connected to this Mikrotik. On any "Google Play" not showing images or loading them very slow. When i change router to my other 951G-2HnD with RouterOS 6.20 (This same configuration) all smartfones working ok !.

Any solutions ?.

Basdno · Sat Jul 11, 2015 4:38 pm

After reading this thread, it seems like I will be waiting for ROS 6.30.1 before I consider upgrading any production router atleast! ;D

But I really LOVE the new system of "MAIN/New Feature" Releases(6.30), and then the sub point "BUG FIX" releases(6.30.1)!

This is a really great new way of releasing RouterOS, this way we can see how new mainreleases work, and then on production/critical routers wait for even more stable and "BugFree" subpoint releases without worrying about something new that might mess up ur router!

Keep this system for all future plz!

Keep up the wonderful work Mikrotik does!

avenn · Sat Jul 11, 2015 6:42 pm

After reading this thread, it seems like I will be waiting for ROS 6.30.1 before I consider upgrading any production router atleast! ;D

But I really LOVE the new system of "MAIN/New Feature" Releases(6.30), and then the sub point "BUG FIX" releases(6.30.1)!

This is a really great new way of releasing RouterOS, this way we can see how new mainreleases work, and then on production/critical routers wait for even more stable and "BugFree" subpoint releases without worrying about something new that might mess up ur router!

Keep this system for all future plz!

Keep up the wonderful work Mikrotik does!

+1000

akthar123 · Sun Jul 12, 2015 6:41 am

after upgrading my RB450G to 6.30 UM shows following error frequently

A synchronization error has occurred. Click "Ok" to re-synchronize

Please help

WirelessRudy · Sun Jul 12, 2015 2:29 pm

Ok, rb433AH + rb750UP combination after downgrade 6.30 => 6.27 again up and stable for 25 hours! Finally seem to get my network back under control! No more port flaps and no more ethernet traffic stopping. No more excessive package loss.

I have some other routers still running 6.30 (3.29.1 was really bad, going to 3.30 improved things) and as long as they aren't giving problems I leave them. But no more new routers to 6.30. I think I stick with 6.27 for a while more until some new version really showing more confidence by users.....

dadoremix · Mon Jul 13, 2015 12:42 am

where to download older relese ? mipsbe

WirelessRudy · Mon Jul 13, 2015 3:06 am

after about 36 hours running on the 6.27 again had a crash on rb433AH. I took a deeper look where this now could come from and found the AP-side of the bridge serving this board also still was serving a v.6.29.1 unit! hmm, must have missed that. Downgraded that one too back to 6.27 and counter at '0' again......

lordkappa · Mon Jul 13, 2015 3:42 am

where to download older relese ? mipsbe

Older releases are considered "Not officially supported" and are not available on the mikrotik download page.

If you have a special need, make your case to support@mikrotik.com; they might send you an older version.

jebz · Mon Jul 13, 2015 4:18 am

where to download older relese ? mipsbe

For the current 6.30 it's -
http://download2.mikrotik.com/routeros/ ... e-6.30.npk
If you copy and edit this URL you can change it to 6.27 for example -
http://download2.mikrotik.com/routeros/ ... e-6.27.npk

soyelpulpo · Mon Jul 13, 2015 7:13 am

Update to RB751 seemed to go well, including the new firmware. I appreciate the new features and the bug fixes as well, but it would be nice if we could remove the dummy firewall rule and associated mangle rules introduced in this release.

I've been unable to remove or disable the filter rule from the command line or in Winbox ("cannot remove builtin" error), and I don't want to have to add this new rule to our firewall documentation, or explain to our auditors what this rule is and why it's suddenly appeared since our last audit. On the networks where we're required to be compliant, because there's been a firewall addition that we can't undo we'll also have to re-scan our networks for compliance.

Do a reset device and check the box that says "no default configuration".

Sent from my Nexus 4 using Tapatalk

dibatech · Mon Jul 13, 2015 7:35 am

Upgraded a CCR1009+ from 6.25 to 6.30
Since a CCR1009+ will not connect reliably via SFP+ with high end Cisco kit with auto-negotiate enabled on the module, we needed to disable auto-negotiate for the SFP+ slot.
When we upgraded, auto-negotiation was re-enabled by Mikrotik.
Of course we were locked out of our device.
Not nice. Another oversight.

TomosRider · Mon Jul 13, 2015 9:22 am

after upgrading my RB450G to 6.30 UM shows following error frequently

A synchronization error has occurred. Click "Ok" to re-synchronize

Please help

Which ROS you had before upgrade?

rufee · Mon Jul 13, 2015 12:12 pm

Still got problems with *) pptp - fixed "encryption got out of sync" problem;
On CCR1009-8G-1S-1S+ v6.30

User can stay connected for a very long time, but when you start pushing traffic the tunnel dies within a few minutes.

Mon Jul 13, 2015 3:37 pm

Autonegotiation on a CCR1036-12G-4S / S-31DLC20D does not work. Plugging the same SFP-Module in a GSP-Switch autonegotiation works. Setting the other side to 1000/FDX work but Flow Control does only work with autonegotiation. The interface monitor command does not show the actual state of negotiated Flow Control it shows the setting.

ismets · Mon Jul 13, 2015 5:21 pm

Hi

When upgrade 6.30 tile,mips l2-tp default route problem.
If select "Add default Route" then "Connect to address" added routing table.
Why?

Best Regards

Chupaka · Mon Jul 13, 2015 6:55 pm

When upgrade 6.30 tile,mips l2-tp default route problem.
If select "Add default Route" then "Connect to address" added routing table.
Why?

have you read changelog?

*) pptp & l2tp client: when adding default route, add special exception route for
a tunnel itself (no need to add it manually anymore);

ismets · Mon Jul 13, 2015 7:06 pm

Ooops!
I cant see changelog.
Its why answer.Thanks

But 1. Blue line is my "Connect to Address" route.
2. blue line is other isp route.How can i sort routing table dynamicly.

Other proble is dynamic route.Gateway adress and dst addres is same.
On l2-tp connec to address=x.x.x.x
on Routing table dst-addres=x.x.x.x and gateway=x.x.x.x distance=0

Thanks alot.

fikkar2 · Mon Jul 13, 2015 10:28 pm

I just signup to this forum just to post this.

I use RouterOS since version 5.14 on my Main Router.
My hardware platform is x86.
I am using Intel D410PT. With 512MB of DDR2 RAM and D-Link 528T (Gigabit NIC).
I also use VLAN quite heavily especially VLAN Trunking to compensate the lack of HW interfaces on my router. And I am doing it by pairing my router with Mikrotik CRS 125 series.

My Backbones are:
1. Fiber Optic WAN1 (75Mbps 1:1)
2. Fiber Optic WAN2 (30Mbps 1:1)
3. Cable DOCSIS 3.0 WAN3 (10Mbps 1:8)
4. ADSL+ WAN4 (8Mbps 1:8)

Some of key Technology that I use on my main router:
IPv4, Dynamic & Static Routing, DHCP Client & Server, Hotspot, VLAN, EOIP, IPIP, DNS, Firewall, NTP Client & Server, email, scripting, PPPoE Client, PPTP Server, L2TP Server, Simple-Queueing, etc.

My Average trafic was around 30Mbps.

I've subscribed to Mikrotik email newsletter so I've been notified every time there is a new released version of RouterOS.
And I usually upgrade all my RouterOS everytime I've received an update, So I've use almost every single release version since version 5.14. There were no problem at all.

But when I've upgraded my Main router to this version (6.30). My main router keep crashing after 1-2 days of uptime.
No reboot. Just Hangs. (No Image from VGA, No Keyboard response, nothing, just FAN and my LED light on my case and on my NIC was light up and blinking properly)
If I do power off & on cycle, I will boot and working normally, but no more than 2 days.

I cannot afford the cost of finding and debugging the problem.
So I've just downgraded to Version 6.29.1
And now it works just fine (again).
And this is the first time in my life that I use RouterOS downgrade command/function.

I've learned a lot from this version (6.30).
My learning was "Never 100% trust the 'release' version from Mikrotik". I've received numerous complains from my clients and some of them are just unsubscribe from my services because of this events.

Sorry if anyone were offended by my post.
But I just want to share my disappointment with someone.
Thank you.

sirkike · Mon Jul 13, 2015 11:34 pm

There are a bug, when try to move many items at the same time in Winbox with mouse. Only get moved one. (sorry bad english) 6.30 versión x86.

vortex · Tue Jul 14, 2015 12:20 am

Default autonegotiation of RB2011 with Samsung SUHD TV does not work.

The TV advertises 100. I had to disable 1000 advertising on the RB2011 for the link to be established.

WirelessRudy · Tue Jul 14, 2015 1:05 am

Well, new discovery. My rb433AH that was previously with other routers upgraded to 6.29.1 which worked out to be a disaster, After that we went to 6.30 which improved some things but still traffic stopped after so many hours so we went back to 6.27. Traffic runs much 'cleaner' (less package losses) and so my conclusion is that after 6.27/6.28 (never tried 6.28 though) things went wrong. What was new? Well, the new fastpath feature was extended to bridged interfaces etc. Something I use quit a lot.....

So we are back to 6.30. (This must be "6.27" I see now. Typo, corrected 14/7) So far, after 23,5 hours the link still runs, but what I see (I am monitoring with a special graphic ping plotter tool) is that package losses are gradually increasing.
After the reboot they are around 1%, but during the day the went up to about 8 to 9% with more and more cuts. Ok, during the day more and more traffic runs over these 'overcrowded' link. How come?
Well, obvious the new firmware is still in place....

So decided to do a reboot. Unit comes up and yes, package loss is at least 6% lower. (I am talking average!). Traffic even jumped up a bit more, obvious, clients get better connections I would say.....

So, is it the firmware making the issue?
How can I downgrade the firmware?
Ok, just found out. Downgraded and package loss back to 2.5% with 10-20Mb up and down traffic from several branches in the tree over the wlan1 wireless with CCQ of 60-80% not bad.
See if I can keep it that way for the next 24 hours....

elfking · Tue Jul 14, 2015 2:13 am

For anyone getting on Mikrotik's case about bugs and having a hard time with their production equipment failing / crashing or otherwise... Lets get real, look through the forums, do some research. You will constantly have someone complaining of issues in their production environment.

If you have a production environment (for example someone before me is upset he had downtime on his production setup due to bugs in 6.30 and lost customers or money and time) having great luck blinding upgrading previous versions is not going to guarantee a success in the future. Any upgrade is a risk which may involve rolling back! Even more so in a production environment where you want things as stable as possible with as little downtime as possible.

You want to test something new? Get a second identical model router /hardware and upgrade that first, export /import configs to the test, TEST TEST TEST then roll it out to your production environment WITH keeping the older version tested and true as fall back. This just makes sense to me, sure it may cost you a bit more money to buy the second hardware but you should have a backup handy in case of a hardware failure anyways!

Everyone chooses to upgrade to the latest, it is not forced on you. I have rolled back plenty of times. I even have some internal hardware running v3.25 still because they get the job done, without any crashes and no need for the new features. All export their configs on a scheduled routine, as well as have backup hardware ready in case of failure.

I generally test the latest at my home, even when I have a test environment setup to give extra effort to try and catch a bug before rolling an update out to my production environment. Even then I roll out to ONE and let it run for a month and decide if I want to roll that out to the rest ( which yes may put me behind with some updates but at least I know its stable, or what features may be broken)

In conclusion, nothing is perfect. Programming included. TEST TEST TEST before production rollout!

I enjoy all of my Mikrotik products when I don't screw up my own config and spend hours finding my mistake

WirelessRudy · Tue Jul 14, 2015 3:38 am

@elfking

Don't know how your 'real' world works but the way you prescribe is not real, let me show you why:

1. Most of the times units can only be put to the full test IN an working environment!
I seldom seen units that did not work on my desk with the latest software. They always do. Bugs usually only come up in working environments. Because only then they need to perform like in THE real environment. So unless you run 2 identical networks, one for use and one for test, you statement is worth nothing.

2. Off course is it good conduct to 'taste and feel' new releases first. Hence you put them in new client units first and test it towards a nearby tower from a location you can easily reach. Then put it in the working environment at easy to reach and of lesser important status. Only after several units this way seem to work you might start thinking of upgrading more important units like AP's, backhauls or remote units that need lengthy roll-outs.

3. Off course you read the forum first to see if there are not too many failure reports. And if the ones that are mentioned might be having an impact on your configs. (If you don't use PPoE but some that do report issues you can expect its probably not going to hurt you. If you work with MT for many years, you know that bugs will always be there, or not real but results of poor configs by users. So you try to filter and 'read' the messages and shown issues with possible workarounds. Only if 2. and 3. are satisfactory start with no. 1.

4. If you have existing problems due growth in the network or traffic, and you are desperately waiting for solutions to improve your network that can only come from software and/or firmware upgrades, or maybe because you do have a known bug bothering you and the new release is promised to have it solved, then it makes sense to try the new release to see if it indeed does help. Yet again, take steps 2, 3 and 1 in the back of your head.

5. If you have a problem you really don't seem to be able to solve and you really need help that isn't be found on the forum, than you can ask a support request of MT. Their first, almost standard, reply will be; "upgrade to the latest version". Basically your statement is now shoved into the bin by MT! And they are the ones that should know! (?)

6. Some bugs only show up after long term running in working environments. And sometimes issues only evolve on very specific boards / setups where similar boards / setups show no problems in same software version. I was happy with 2.29.1 for more than a week. We indeed saw improved throughputs on the bridges. Until the shit started to hit the fans...
How could I ever have expected that, let alone tested?

Every serious technician / operator will be cautious to use the latest, newest and 'best'. But you can't always wait untill everybody else proved it worked all fine. If you would follow same argumentation on transport means we would all be still walking to work or ride a horse. Cars and planes still crash at times due technical issues and errors not known before.
Most improvements on technology (so also software improvements) are only done because someone somewhere sometime found the error in a working environment and a crash occurred. You can't test all.......

Tue Jul 14, 2015 8:59 am

No reboot. Just Hangs. (No Image from VGA, No Keyboard response, nothing, just FAN and my LED light on my case and on my NIC was light up and blinking properly)

If you are still here. It would be great if you at least made a supout.rif file after the reboot (or check, if there is one made automatically) and send it to mikrotik. Just from this description, it is impossible to tell what is the cause of the hang.

Ansy · Tue Jul 14, 2015 10:05 am

WirelessRudy, +100500!
You save me a full screen of great words

MT, you've completely broken Bridge + Firewall function in 6.29-6.30.
RCs sometimes worked, somewhere not... but let "get the graphs" now:

mt_traf_2015-07-12.png

mt_traf_2015-07-13.png

mt_traf_2015-07-14.png

AFAICS, last RC with Bridge - Settings - Use IP Firewall is ON (and so on, Simple Queues & some Firewall NAT rules) RB750UP works with reboots by Watchdog (to default gateway) or kernel crashes every some hours.
Weekend, not so much traffic. BTW, not every crash spikes shown because graphing script (not on MT device) get data minutely, but RB750UP reboots quickly in 15-20 seconds, being configured as Bridge.

Then (last graph), main business users awaken, eating traffic... device began to crash more and more frequently, got customers complaining.
I had to work around it, and at 13:30 I tried to get device back to 6.27, but failed. May be just because of crashes every 2-3 minutes.

Well... then I've turned Bridge - Settings - Use IP Firewall check box OFF!
Voila! Device works rock stable for now -- as DUMB Bridge with PoE, but -- no Firewall, no NAT, no Queues and any QoS. Very sad.

P.S. Last night reboot/spike is my own, downloaded 6.31rc5, nothing changed (I've hoped on snmp scripts fix in changelog)

Tue Jul 14, 2015 11:27 am

Ansy - have you conacted MikroTik support about it ?

HussamRefai · Tue Jul 14, 2015 1:51 pm

Script to update DynDNS will fail after upgrade to v6.30.
Downgrade to v6.29.1, the same script works.
Have no idea what happened in the" /tool fetch" process, it just stop there......

The script code:

:local ddnsuser "myname";
:local ddnspass "mypass";
:local theinterface "pppoe-myisp";
:local ddnshost "ddnshost.dyndns.org";

:local ipddns [:resolve $ddnshost];
:local ipfresh [ /ip address get [/ip address find interface=$theinterface ] address ];
:if ([ :typeof $ipfresh ] = nil ) do={
:log info ("DynDNS: No ip address on $theinterface .");
} else={
:for i from=( [:len $ipfresh] - 1) to=0 do={
:if ( [:pick $ipfresh $i] = "/") do={
:set ipfresh [:pick $ipfresh 0 $i];
}
}

:if ($ipddns != $ipfresh) do={
:log info ("DynDNS: IP-DynDNS = $ipddns");
:log info ("DynDNS: IP-Fresh = $ipfresh");
:log info "DynDNS: Update IP needed, Sending UPDATE...!";
:local str "/nic/update?hostname=$ddnshost&myip=$ipfresh&wildcard=NOCHG&mx=NOCHG&backmx=NOCHG";
/tool fetch address=members.dyndns.org src-path=$str mode=http user=$ddnsuser \
password=$ddnspass dst-path=("/DynDNS.".$ddnshost);
:delay 1;
:local str [/file find name="DynDNS.$ddnshost"];
/file remove $str;
:global ipddns $ipfresh;
:log info "DynDNS: IP updated to $ipfresh!";
} else={
:log info "DynDNS: Don't need changes";
}
}

i have the same problem , its very important here in UAE , all VPN connections here using DYNDns

apathy · Tue Jul 14, 2015 2:02 pm

What about traffic flow? In 6.29 broken it. Please fix it.

HussamRefai · Tue Jul 14, 2015 2:05 pm

LTE interface number changes in update. I have many Baseboxes with Sierra MC7710 LTE card in remote locations. During update LTE1 interface changes to LTE2 interface. This breaks up DHCP client and firewall rules that are still trying to use LTE1 interface after update. This prevents updating remote sites as connection won't come up after update automatically. So beware if you are upgrading remote site with LTE connection.

I have sent supouts to support and waiting their reply.

same here , i have 3 remote site using LTE , i have to go to the sites by my self to get back the connections

Ansy · Tue Jul 14, 2015 2:20 pm

Ansy - have you conacted MikroTik support about it ?

Yes, contacting from early 6.28. Last [Ticket#2015071466000444]

Support suggests NAND/partition format errors. But I've updated firmware & ROS RC's without problem many times and RC's, System - Resourses says 0 Bad Blocks too (is it NAND?)

I can't understand how does well working device (without Use IP Firewall option) depend of NAND errors for so many RC's long? And totally fails when I turn this checkbox ON? After version 6.27 only?

P.S. Uptime is 15+ hours for now without issue.
As dumb bridge, of course.
While winbox monitoring 4 ether interfaces, etc.. managing 20Mbps traffic ~10% CPU load.
But I want some NAT rules in IP Firewall and a dozen of Simple Queues, as it was for some recent years.

Clbh · Tue Jul 14, 2015 4:58 pm

What about traffic flow? In 6.29 broken it. Please fix it.

The change log for the most recent 6.31 RC appears to have disappeared, but last time I looked at it earlier today, it included a fix to restore the previous "birectional" traffic capture behaviour for traffic-flow.

wombat · Wed Jul 15, 2015 10:15 am

Hello,

I have (had) I was working configuration L2TP / IPSec (MK / clients) for any reason after UPG v6.30 stopped working on connecting mobile clients via Android but Windows works normally. Google did not help anything. I tried a lot of variations according to the log, but not much, including configuration changes helped. Any idea what it could be?

Connecting the first takes a long time - he writes on mobile Connecting ...
2. MK had already client is connected, but does not pass any data
3. The client is disconnected after some time on the phone to give a "Connection failed" ..

Thanks.

mrz · Wed Jul 15, 2015 10:18 am

enable aes-256 in ipsec proposal

wombat · Wed Jul 15, 2015 10:20 am

Enabled it .. But doesn't work.

wombat · Wed Jul 15, 2015 10:24 am

for example log:

05:06:19 ipsec,debug no policy found: CLIENTIP/32[0] ROUTERIP/32[0] proto=udp dir=in 
05:06:19 ipsec,debug no policy found: ROUTERIP/32[0] CLIENTIP/32[0] proto=udp dir=out 
05:17:14 ipsec,debug no policy found: CLIENTIP/32[0] ROUTERIP/32[0] proto=udp dir=in 
05:17:14 ipsec,debug no policy found: ROUTERIP/32[0] CLIENTIP/32[0] proto=udp dir=out 
09:19:21 ipsec,debug no policy found: CLIENTIP/32[0] ROUTERIP/32[1701] proto=udp dir=fwd 
09:19:21 ipsec,debug such policy already exists. anyway replace it: ROUTERIP/32[1701] CLIENTIP/32[0] proto=udp dir=out 
09:19:21 ipsec,debug such policy already exists. anyway replace it: CLIENTIP/32[0] ROUTERIP/32[1701] proto=udp dir=in

mrz · Wed Jul 15, 2015 10:26 am

There is nothing wrong with these log messages.
Contact support with attached supout file.

rextended · Wed Jul 15, 2015 12:22 pm

WARNING: "/tool user-manager user export" DO NOT EXPORT USERNAMEs

minor bug on export:
if you have keepalive ON with 10,10 as value, the export still report it as "!keepalive" instead "keepalive=10,10"

Actually ALL my core devices have 6.30 without any problem (from the day of 6.30 release) except the two reported here.

On future i test the 6.30 on wireless devices. and on the S-RJ01 module, because on previous version (6.29.x) i have problem with ethernet drivers for S-RJ01 and on ethernet freeze on "ac" models.

rextended · Wed Jul 15, 2015 12:54 pm

[...]
I've learned a lot from this version (6.30).
My learning was "Never 100% trust the 'release' version from Mikrotik".
[...]

First of all you have to learn:
Never use just released hardware and/or software (of any company) for production environment.
If you do that, you accept any risk and you can not complain the productor.

[...]
I've received numerous complains from my clients and some of them are just unsubscribe from my services because of this events.
[...]

If one client unsubscribe for only one problem, the client is worthless

Wed Jul 15, 2015 1:21 pm

WARNING: "/tool user-manager user export" DO NOT EXPORT USERNAMEs

minor bug on export:
if you have keepalive ON with 10,10 as value, the export still report it as "!keepalive" instead "keepalive=10,10"

Actually ALL my core devices have 6.30 without any problem (from the day of 6.30 release) except the two reported here.

On future i test the 6.30 on wireless devices. and on the S-RJ01 module, because on previous version (6.29.x) i have problem with ethernet drivers for S-RJ01 and on ethernet freeze on "ac" models.

Do you use MPLS/LDP ?

rextended · Wed Jul 15, 2015 1:37 pm

WARNING: "/tool user-manager user export" DO NOT EXPORT USERNAMEs

minor bug on export:
if you have keepalive ON with 10,10 as value, the export still report it as "!keepalive" instead "keepalive=10,10"

Actually ALL my core devices have 6.30 without any problem (from the day of 6.30 release) except the two reported here.

On future i test the 6.30 on wireless devices. and on the S-RJ01 module, because on previous version (6.29.x) i have problem with ethernet drivers for S-RJ01 and on ethernet freeze on "ac" models.
Do you use MPLS/LDP ?

Not, only "pure" routing, BGP and VRRP on edge routers.

mjoksimovic · Thu Jul 16, 2015 8:00 pm

Where can I download all packages for x86 platform, version 6.30? After upgrade from 6.28 to 6.30, I've lost NTP server package from main CORE router. Thanks.

WirelessRudy · Thu Jul 16, 2015 10:12 pm

For those that wondered what happened with my units that give me lots of headaches with 6.29.1 and later 6.30... well, back to 6.27 brought peace.
One troubled board that had a failing ethernet port is now running for 26 hours without fails. Still got some 0-1,5% package losses at times but with 6.30 inmediately after a reboot it started with 3-5% and went slowly up with the traffic to about 20-30% and than it stopped passing traffic after all...

Two other boards having similar issues but on top of it developed kernel crashes (after more than a year uptime!) that made the board reboot every so many hours (on top of the port failure) still has the kernel crash, but only after 20 hours or so.
So so far it improved a lot to go back to 6.27 (also downgraded the firmware. With the 6.30 combined firmware I still had lots of issues...)

I must say I still have several CPE's running on 6.30. They seem to be fine. Even some with 6.29.1 are still fine.
I also have still some rb750UP en some rb433AH with 6.30 running that never gave, and still don't give any problems...

The boards that still give the issue at time are both boards where one of the ports doesn't work anymore. Actually they are up for renewal.
One other board improved a lot by just changing a powersplit cable.

So it looks to me that boards that already might have some issue on the ethernet part of connections (mismatch in settings on both ends or board with Gigabit connected to fastehternet port connected and both on 'auto' are a disaster.
Under 6.27 my rb800 (gigabit ports) connected to a rb433AG (fastE) had some issues before if both ends were on auto/full/10/100/1000. Once in so many weeks we had a port stopping to pass traffic. So we set both ends to manual/full/100 and they worked fine for months.
Under 6.30 we also tried the auto/full/10/100/1000 again but that made the ports stopped passing traffic almost within minutes. Both end on 'auto/full/10/100/1000 was much better but then we still had the kernel crash...

Imho with the introduction of fastpath for bridges and in a mixed environment and/or some less perfect ethernet connections, the ethernet ports are making problems.
I don't know about the process or drivers but its clear to me something is gone wrong in the new feature.

I'm wating for newer releases or bug finding before any important router gets the new feature installed again.....

padrecc · Fri Jul 17, 2015 10:57 pm

... after UPG v6.30 stopped working on connecting mobile clients via Android but Windows works normally...

Same here

Fri Jul 17, 2015 11:20 pm

Many of my laboratory devices running 6.30 have lost dns server dynamically obtained from dhcp server so they were not able to resolve domain names after few days of runtime. Dhcp client renewal did not helped. Only reboot was effective. Testing 6.30.1 now and hoping it could be better.

macsrwe · Mon Jul 20, 2015 2:46 am

Frankly, the feature that excites me most in 6.30 is the ability to set the owner of a script.

Because we provide a somewhat rich environment in each of our CPEs, we spend more time than we should have to when we set up new customer CPEs. The setup procedure for our standard configuration is unwieldy:

We netinstall them to our standard release of ROS.
We log in to load three macros and run the first, which resets the configuration and runs the second one on reboot.
We log in again to run the third macro, because it contains two (prior to 6.30) operations that cannot be performed in the reboot macro:
1. creation of the scripts with the real admin user as the owner (or else schedule will not run them — Ticket#2013090666000085); and
2. the setup of /system upgrade upgrade-package-source, which demands the login password interactively and has no option to provide the password on the command line (unlike every other command in ROS — Ticket#2013012566000027).

6.30 has just removed the first impediment to our being able to provide the configuration macro as part of the netinstall, where it really should be. Now, all we need is for somebody to allow the OPTION to provide the password to /system upgrade upgrade-package-source command, and our setup time will be cut to a fraction of what it is now.

When I asked for this feature over two years ago, I was told it was "done due to security, to avoid username password appearing in plain text." It makes no sense for MikroTik to be squeamish about this one command, given that every other ROS command that deals with passwords/secrets/encryption keys allows input on the command line. (You may prefer to suppress its output in an export, but that's an entirely different issue than allowing it to be input; plus the subsequent introduction of /export hide-sensitive solves this issue anyway.)

This would seem to be a truly simple change to ROS. Can you folks please add it? Pretty please?

Ansy · Mon Jul 20, 2015 9:57 am

So it looks to me that boards that already might have some issue on the ethernet part of connections (mismatch in settings on both ends or board with Gigabit connected to fastehternet port connected and both on 'auto' are a disaster.
Under 6.27 my rb800 (gigabit ports) connected to a rb433AG (fastE) had some issues before if both ends were on auto/full/10/100/1000. Once in so many weeks we had a port stopping to pass traffic. So we set both ends to manual/full/100 and they worked fine for months.
Under 6.30 we also tried the auto/full/10/100/1000 again but that made the ports stopped passing traffic almost within minutes. Both end on 'auto/full/10/100/1000 was much better but then we still had the kernel crash...

Imho with the introduction of fastpath for bridges and in a mixed environment and/or some less perfect ethernet connections, the ethernet ports are making problems.
I don't know about the process or drivers but its clear to me something is gone wrong in the new feature.

WirelessRudy, the same suggestions in my case:
http://forum.mikrotik.com/viewtopic.php ... 91#p490689

I have very noisy environment because of high power FM-radio transmitter near on the same roof, and Ethernet lines didn't work properly untill they've got ferrite filters on the both ends. Now only one line (ether3) of my RB750UP doesn't have filters (hanging too high on the tall mast), and I've only 10Mbit/s rate there.

So, my Bridge became dumb after 6.28+, crashing and rebooting by Watchdog timeout, loading CPU 100% by management process, logging weird "snmp, warning timeout while waiting for program 20" while I've SNMP totally OFF now... but still investigating under MT support control.

Mon Jul 20, 2015 10:06 am

Many of my laboratory devices running 6.30 have lost dns server dynamically obtained from dhcp server so they were not able to resolve domain names after few days of runtime. Dhcp client renewal did not helped. Only reboot was effective. Testing 6.30.1 now and hoping it could be better.

I can confirm this to be problem of 6.30.1 too.

[Ticket#2015072066000218] created.

The easiest solution I have found so far is to make "release" of dhcp client, "renew" only does not help. Hope it will be corrected soon.