v7.16.2 [stable] is released!

[boingolover]

I posted about occasional FIN and RST packets from local subnets leaking out of the WAN interface here:

http://forum.mikrotik.com/t/occasional-fin-or-rst-packet-showing-up-on-wan-from-my-private-subnets/178176/1

As of 7.16, I can no longer reproduce this issue. So one of two things are happening as far as I can tell:

1. the AT&T gateway is no longer logging this leaked traffic from rfc1918 addresses

2. Mikrotik fixed this issue and didn’t call it out in the changelogs

I’m going to try to rule out #1 later today, but just curious if anyone else has seen this?

[pe1chl]

The firewall rules that I have added to catch such traffic still have hits, so I doubt it has been fixed…

[Kernal87]

I am now connected such a router that no longer reacts on wan interface via an internal interface. This is an LTE router LHGGM. For example, when I do /ip/route/print, the input simply hangs and no output is generated.
The same happens when I want to display Wireguard configs or BGP configs. The memory utilization is at half 128 MB.
I wonder how these things can die like this.

[Spirch]

from 7.12.1 on hap ac2 to 7.16, i had to unplug the power and replug, waited 15 minutes before doing it

i had a single device on it that got an ip and was actually working
every other one were not able to get an ip, same for the router itself, mac address connection was not working too

[mustang1986]

*) wifi-qcom-ac - improved memory allocating process;

I’m afraid this is not fixed yet.
cAP AC, reboot after 2d22h due to kernel failure.
PRTG monitoring shows increasing memory usage until reboot happens.

Back to daily scheduled reboot for now 8)

I already have an open ticket, will send autosupout there.

Not good news…
I use 44 piece cap AC with qcom-ac (more VLAN and ~70 piece wifi client with 802.11r fast BSS transitions ( roaming), routeros 7.15.3 and 7.16rc1-5).
After 7-10 days runs out the cap’s memory.

[boingolover]

The firewall rules that I have added to catch such traffic still have hits, so I doubt it has been fixed…

I guess I was just impatient, I did eventually see them. So it’s 3) none of the above

[infabo]

*) console - added "about" filters for "find" and "print where" commands;

How does this about parameter work? Can’t figure it out. Docs unavailable.

bump. This is discussion strictly related to release 7.16

[abrehmc]

My CRS310-8G+2S+ simply refuses to upgrade from 7.12.2. The arm package is there, but it doesn’t get picked up on reboot.

[normis]

1. Check the Log file and it will tell you why

[abrehmc]

1. Check the Log file and it will tell you why
2. Why this old version? You should be using 7.16

It came from the supplier with that version. Going to 7.15.3 and then to 7.16 worked. I have no idea why it did that, I should have saved the log.
If it happens again on the other switches, I will check it.

[holvoetn]

My CRS310-8G+2S+ simply refuses to upgrade from 7.12.2. The arm package is there, but it doesn’t get picked up on reboot.

Most likely space issues because moving from 7.12.x going higher, will probably also install wireless. On a switch …
(on the CRS devices I upgraded, I saw it each time)

Log file should have indicated space problems so if you have others to process, be aware.

[normis]

Sorry I misread the issue Glad you solved it .

[prawira]

–deleted–

[prawira]

–deleted–

[prawira]

*) arp - fixed possible issue with invalid entries;
has #[SUP-136090]: flag DIH on /ip arp been solved ?

*) ip - added max-sessions property for services;
is it to limit max session on the same time of each services ?
please add “idle timeout” as well.

TIA

[deluxor]

My RB5009 had a kernel failure today:

It is running this version.

[nmt1900]

Word ‘probably’ should not have a place in log messages…

[nexusds]

Thanks @ToTheFull, on my hAP EX2 I had the same problem from v.15.x onwards.
I will stay with v.14.3.

Regards.

agreed. something is different since 14.3 that using multiple AX APs or some other similarity just doesnt work. going back to 14.3 and all is stable and good.

[hknet]

it seems vrf-routing is trouble, coming from 7.15.3 static vrf routes were marked inactive and we found no way to get those active, neither deleting, adding new ones, basically all vrf-routes are inactive and show things like:

 4  IsH  dst-address=0.0.0.0/0 routing-table=main gateway=10.100.6.5
         immediate-gw="" distance=1 scope=30 target-scope=10
         vrf-interface=MGMT

in addition to static routes within vrfs not working also dynamic routes (added by pppoe client) were inactive within vrfs. this is seriously messed up.

finally had to downgrade to 7.15.3 to get this working again.

Config looks to be incorrect, either you use routing-table to determine to which vrf this route should belong or you use vrf-interfce. But not the both especially if routing table config does not match the vrf to which vrf-inteface belongs to.

tried all ways imaginable - couldn’t get it to work besides downgrading…

[hknet]

Config looks to be incorrect, either you use routing-table to determine to which vrf this route should belong or you use vrf-interfce. But not the both especially if routing table config does not match the vrf to which vrf-inteface belongs to.

tried all ways imaginable - couldn’t get it to work besides downgrading…

revisited the issue - well - before upgrading one has to remove the parameter “vrf-interface” from any static vrf routes that do typically carry routing-table=vrf and vrf-interface=NAME which obviously breaks the routes after v7.16 upgrade.
also dialer-interfaces that enable “add default route” do not add a working default route into its assigned vrf, one has to add a route like:

add disabled=no distance=1 dst-address=0.0.0.0/0 gateway=PPPoE-VRF@vrfO routing-table=vrfO scope=30 suppress-hw-offload=no target-scope=10

it’s a pitty an upgrade breaks the config this way.