RB951G-2HnD
Reboots every minute after upgrade to 6.38.1. High CPU usage for a short time leads to constant reboots.

951G-2HnD
firmware 3.33
upgrade to 6.38.1 good
CPU load 1-4%, for "speedtest" - 14% max

When trying to recover faced another problem.

Problem description:
Ports order flip when restoring configuration from backup

Equipment used in test:
A. Mikrotik RB951G-2HnD firmware version 6.38.1
B. Mikrotik RB951Ui-2HnD firmware version 6.37.3

Steps to reproduce:
1. Backup configuration via Files (copy/paste in Winbox) on A
2. Copy backed up configuration via Files (copy/paste in Winbox) from A to B
3. Restore configuration via Files on B
4. Reboot B
5. Ports order flipped upside down (and wlan disabled)

For example, on A I have port 1 named ether1-master and port 5 named ether5-gateway. After restoring configuration on B port 1 named ether5-gateway and port 5 named ether1-master. Firewall rules and filters following invalid ports order.

1. Backup configuration via Files (copy/paste in Winbox) on A
2. Copy backed up configuration via Files (copy/paste in Winbox) from A to B
3. Restore configuration via Files on B

Backups are not meant to move/copy configuration between different routers. Do not use it to move the configuration even between two routers of exactly the same model. It not only may lead to an undesirable effect like the one you've just described, but it also copies values which are meant to be different on different boards (for instance interface MAC addresses).

When you need to transfer the configuration from one router to another use /export instead.

When you need to transfer the configuration from one router to another use /export instead.

Appreciated for this information. I've moved disputable configuration which leads to rebooting to a different router (RB951Ui-2HnD) with 6.38.1 firmware. Will see.

hAP ac, ROS 6.38.1, RouterBOOT 3.34:
Firewall/Address-List/Creation-Time shows GMT time, not time according to time-zone settings.
Can someone confirm this?

Regards,

ditonet - Yes, that is correct. We will try to fix this in next RouterOS release.

ditonet - Yes, that is correct. We will try to fix this in next RouterOS release.

Thanks.

Regards,

when will be fixed WIFI problem with old devices on Intel 2200BGN ? Still can`t connect to hAp Lite with old Toshiba laptop. Version prior 6.37 works perfect!

Well.... it's next beta.
It bricked my hAP ac on the bench - it had default config (I did an reset like 5 minutes before) with WiFi disabled. It just went into dead boot loop - only netinstall was working. I used update feature from Quick Set and the previous ROS loaded was 6.37.1.

I don't see (or I missed that) nothing about EoIP+IPSec bug fix forhttp://forum.mikrotik.com/viewtopic.php?f=2&t=116589&p=580377#p576524 but quickly testing it has been half-fixed - when local IP is changed it's also changed in IPSec policies.

p.s. Is there any way to safely upgrade large CAPsMAN network without great disturbance in services?

is there an update on the IPsec issue (see some posts above)?

every time responder size changes its IP adress and reconnect before the old dynamic policy is flushed, there is an invalid policy which prevent any communication.

every time responder size changes its IP adress and reconnect before the old dynamic policy is flushed, there is an invalid policy

Do you have DPD enabled?

Oh dear, it turned into one of those mornings. Upgraded device to 6.38.1 only to discover the problem with VLAN on a bridge interface not working, pushed onto 6.39rc20 where this all came right and must be fixed but now I'm finding a 4 port Ethernet switch trunk link between 2x CRS226-24G-2S+ devices will only work if only one of the 4 ports is enabled, the minute I enable port 2,3 or 4 the link goes down.

Is anyone else seeing similar problems?

Cheers, Dominic.

I have an hAP ac lite, after updating my ip system telephone stopped working.
I have Elastix distribution.

every time responder size changes its IP adress and reconnect before the old dynamic policy is flushed, there is an invalid policy

Do you have DPD enabled?

yes, Interval 5 max. failures 3. but it does not do anything. the old policy ist still there after 15 seconds and even after several minutes and does not get removed at all.

devices used: ccr1009 as responder (with aes-ctr and sha256 to avoid reordering problem) and RB3011, RB751G, hAP ac lite, hEX, two RB951G as initiator

Hi all
I have problem with 6.38.1 wireless connection between 2 RB.
1 ap bridge, 2 station bridge. Both connected but no traffic.
6.37.3 works OK.
BTW now 1. 6.37.3 & 2. 6.38.1 also works OK ????

I have problem with 6.38.1 wireless connection between 2 RB.
1 ap bridge, 2 station bridge. Both connected but no traffic.

I saw similar problem, There is a traffic, but can't connect via winbox to RB with station-bridge mode set and no response to ping.
Unfortunately I haven't free time to deeply investigate this .

Regards,

Please would Mikrotik consider a global bridge setting to restore previous per VLAN (R)STP configuration. We got Netgear to update firmware on M4300 switches to pass through STP as-is and were finally able to implement redundant routers where two sets of bridges could provide VLAN bridging. eg vlan14-vlan10 bridged to vlan20-vlan99

The STP changes may conform to a 2014 standard but break service provider redundancy services which is a real problem...

every time responder size changes its IP adress and reconnect before the old dynamic policy is flushed, there is an invalid policy

Do you have DPD enabled?

yes, Interval 5 max. failures 3. but it does not do anything. the old policy ist still there after 15 seconds and even after several minutes and does not get removed at all.

devices used: ccr1009 as responder (with aes-ctr and sha256 to avoid reordering problem) and RB3011, RB751G, hAP ac lite, hEX, two RB951G as initiator

I discovered that this:
ike1 - fixed ph1 rekey in setups with mode-cfg;
seems to be fixed for phase1.
but now the mode-cfg responder ip can get lost at phase2 rekey (not phase1 rekeay anymore).
it happend with quite a load on the interface, but not 100% load, cpu was almost idle at that time.
UPDATE:
when the problem occurs the dynamic allocated IP is lost on responder and therefore no traffic can flow. it seems to be possible to add a duplicated ip address on a dummy/loopback bridge to circumvent the problem for now, but this is clearly not intended behaviour
this does not solve the still present "invalid policy" problem
UPDATE2:
I opened a ticket for the "invalid policy" bug

please also fix this
thank you

Still having problems when on local bridge enabled rstp. After disabling rstp, everything works as it should. This was broken in release before this one.
Two Mikrotik routers connected to each other. When rstp is enabled, I can't ping the other Mikrotik.

Normally I am using just bugfix releases, but RSTP sounds very well, so I tried to upgrade one CRS112-8G-4S and enable RSTP. It is working on all ports except the ether6 - there is no forwarding nor learning, see the screenshot. Why? There is an old RB250GS 1.17 connected. It is not compatible? Newer RB260GS on ether8 is working fine.

Still having problems when on local bridge enabled rstp. After disabling rstp, everything works as it should. This was broken in release before this one.
Two Mikrotik routers connected to each other. When rstp is enabled, I can't ping the other Mikrotik.

I have the same problem, but I had short time to look using Wireshark what happens: bridge does not send Topology Change Notification BPDU, so both bridges think I'm root bridge and then disconnect the other bridge ("split brain")

I am not sure if it is related to this version, but I have problem with the LEDs. After reset they are all off. I use exported config to restore configuration (because I am teaching and students have various RB models) and there are these commands in export: The problem is that these commands are not working (error: no such item) and the only way to restore LEDs is manually set them using WinBox. Is the syntax in export wrong? Or is there a problem to set LED using commands? Tested on hAP and RB951Ui-2HnD.

I have no idea if this is related to the VLAN/Bridge/(r)stp troubles being reported, but I just discovered a bunch of instances of something interesting on my network.

A bunch of my VLAN interfaces had set on "Loop Protect" Send Interval of 00:00:00 and Disable Time of 00:00:00. Defaults are 00:00:05 and 00:05:00. I had to fix these settings before Winbox would allow me to make other changes.

6.38.3 version has been released:
http://forum.mikrotik.com/viewtopic.php?f=21&t=118642