MikroTik

hknet

well, would like to go back to 3.20 with no HiDPI support - at least it worked :)
anyone got an archive link as I seem to dumb to find the previous releases in the download-section...

hknet

well, I'm now also in the having to zoom in three times for any window to get readable business...
please fix this.
Win10 2004, Surface Pro, HiDPI - if that makes any difference.

hknet

@hknet please, don't mess up count of NAT rules with conntrack table size limit.

not me

hknet

I'm curious - have you found the culprit bogging your cores to full load?

hknet

NAT Hardware Offloading ! Great news. When is this coming to the actual router products like CCR series or new router products ? Same Question from my side! word of advice: if you're on 1036 oder 1072 your box can usually handle any 10G nat traffic, of course depending on your setup you could also ...

hknet

based on your ROS release I'd assume this setup has been in production for some time. band-aid suggestions: switch on fastpath, yes you have to remove firewall-policies, also make sure your ip services are secured. as your network processing load seems incredibly high I'd suspect some kind of loopin...

hknet

could I ask for 5G supporting modems?
Location: Austria, Europe

hknet

+1 - seriously needed.

hknet

Maybe you could give us some feedback if you would like to use that feature in our CRS switches? What additional options you would like us to add for this feature? from a provider point-of-view - this is nice to have, but honestly I'd prefer other chip-features implemented available on crs317 hardw...

hknet

hknet

as this is trending and the initial target if I understand it correctly was VoIP priority, here is a small example on how we do it. the assumption is like this: download-link can be oversaturated, though this is not desireable we usually can't do much about it if incoming packets are thrown at the C...

hknet

after talking to support - this is kind of by design:
https://wiki.mikrotik.com/wiki/MikroTik ... t_Ethernet

S+RJ10 modules deliver MTU at max 1504 for linkspeeds up to 1Gbps....

this was unexpected.

hknet

stuff to read: https://www.intel.com/content/dam/www/public/us/en/documents/case-studies/att-cpu-impact-on-packet-processing-perfomance-paper.pdf you will get no where near the packets per second the CCR can handle. suggestion would be to rethink your setup, the added value of 10x full-routingtable ...

hknet

little update on this case: the problem seems to be related to S+RJ10 modules.
after moving to a "classic" sfp module - the payloadsize is back to at least 1500...

hknet

for the record L2MTU on all ports (and the bridge) is well above 1500 - if you never tested Q-in-Q on these switch-models please do not speculate

hknet

at first there was disbelieve, but when we try to transport QinQ frames through this switch (testing on CRS317) we obviously get into a frame-size problem. the usual setup works: one accessport (removing the outer-tag) and one provider port sending and receiving a trunk of outer tags. if we would li...

hknet

praise to @tomranson

indeed the criteria: two MTs having both their BGP-peers in VRFs then you _must_ use the routing rule to do a forced lookup for the other BGP peer IP and then force the policy-route to be resolved via the specific routing-table.

hknet

+1
nice idea

hknet

#funfact a little over three years ago we were teased about the working solution with ROSv7 and we were shown a printout on how it behaves, but that's about it I guess... actually I guess that's one main reason we use MTs for layer2-setups and as CPEs, the core is and probably will be junipers and c...

hknet

Hi regarding the issue: bridge - fixed packet forwarding when changing MSTI VLAN mappings could someone from MT please elaborate? we have been quite unsuccessfull integrating crs317 devices in our network using MSTP the RSTP from other devices arriving on vlans is simply not being replicated to othe...

hknet

Hi
upgraded CCR1072 - works fine.
but: snmp, warning arises with "timeout while waiting for program 79"
which is not ideal.

regards,
hk

hknet

*bridge - added per-port based "tag-stacking" feature trying to wrap my head around this vs the (previous) QinQ support by using ethertype=0x88a8 if I'm not mistaken this is now the more common support for QinQ using 0x8100 frames packing another 0x8100 frame into it (aka stacking). my question here...

hknet

Do I read this correctly as _no_ routerboard-firmware upgrade is usually required coming from 6.43rc or 6.42.7?

thx
hk

hknet

RSS is good, but will be nice to have some mailing list for security announcement and firmware update
+1 for security announcement mailinglist

+2

hknet

Well you guys did want a security blog and more awareness. Now you complain that the issues are not serious enough :D We will try to find a balance. I hope you understand that when you post a new version AND update the bugfix version at the same time, mentioning 4 security issues that have no descr...

hknet

Hi this might be a tricky one, any food for thought would be appreciated (and consulting offers are welcome too if you know what you're doing). The scenario goes like this: there is one QinQ port that is shared between several users of this shared infrastructure. the s-tags are generated per infrast...

hknet

Hi
feeling a bit dumb, but has anyone ever managed to use a microSD in the ccr1072?

we seem to be unable to solve this riddle.

and while we are at it any tested smartcards for the smart card slot?

thx
hk

hknet

Hi on a cisco router we do something like this: route-map RK permit 50 match ip address RK set ip next-hop 192.168.1.1 this route-map is then bound to an interface that is the incoming gateway for packets like interface vl10 ip policy route-map RK in other words if the source IP matches the RK addre...

hknet

after seeing issues with CRS317 an 1G SFPs (not 10G) not linking with latest 6.43rc44 (getting auto negotiation as incomplete) we tested a downgrade to the current 6.42.6 and see auto negotiation failed. copper 1G SFPs do not see any advertised 10/100/1000M link-speeds while the other end gets auton...

hknet

model: CRS317-1G-16S+
release: v6.43rc44

we use some ports with non-10G SFPs (ie 1G-SX and 1G-TX modules with fiber and copper).

those modules get auto negotiation 'incomplete', while claiming the link is ok but no working datatransmission can be established.

regards,
hk

hknet

I like: "crs3xx - added initial Q-in-Q hardware offloading support (CLI only);"
could someone please point me to the correct manual/wiki entry on this one?

hknet

I don't know the inner workings of CRS3xx but...

q.e.d.

hknet

@RoadkillX you do not test your setup on the mentioned HW therefore it's invalid and not what anybody was asking for. @TestCRS thx for the ethertype reminder from RoadkillX I found the relatively simple case to do: set the vlan added to the bridge with "use service tag" and you should be fine. -> se...

hknet

QinQ requires intervlan. it's a vlan interface inside another vlan interface so revisit the docs ;-), to be more clear QinQ requires 2 vlan interfaces which can't be hw-offloaded because they are L3. you might brush up your knowledge, you are misinformed and did not read the first post to understan...

hknet

it should do 10Gb within the same vlan but not intervlan

the original poster was never talking about inter-vlan-anything maybe you should revisit the original problem-report.

hknet

This is not a bug, its clearly misconfiguration respectfully disagree - the CRS and the new bridge-hardware-accelleration config to use it as a switch and get hardware-forwarding is what the original poster tried to achieve imho (and I see no flaw in the config); you are referencing the implementat...

hknet

I cant se your C tag in your setup. A ctag is a vlan to a vlan interface... You need to paste your complete setup with vlan tagged to another vlan interface. well, I guess there is no specific c-tag configured (as the CRS317 wouldn't support qinq in hardware anyway), therefore the claimed s-tag (vl...

hknet

As other people wonder - Are we/Have we gotten hardware support for QinQ?

Doesn't look like it, though the Marvell 98DX8216 is a metro chipset (marvell's words) which would suggest 802.1ad should be available at the HW level.
But my guess is we won't see this with this HW.

regards,
hk

hknet

Did not found how to pm here. Wonna honest answer? Go mx80 for your business. second that. if you could talk your upstreams into limiting your bgp-view (ie only local routes and no full feed) you might still have fun with CCRs. Would you mind elaborating a bit more in detail, why your suggestion to...

hknet

12 x SFP+ (10G)
4 x QSFP (4x10G)
1 x Management (no PoE)
1 x serial management

second that.

hknet

EoIP Ethernet Frame issue is still there (introduced in 6.42 breaking fragmenting big frames somehow broken) verified on RB2011 and sent supout to MT-support.

hknet

it seems we lost packet fragmentation for EoIPs connecting bridges between two MTs we no longer can transport big pppoe-frames there; was working with 6.41.4

hknet

Web, winbox, ssh etc access to the router works only in main routing table. It isn't totally correct. Because if you've got access to the Internet from main route table, mikrotik gets inbound request on VRF interface and sends reply via Internet interface in main route table. Is it possible to chan...

hknet

well after fixing the threshold issue on the ciscos-side we see a MLPPP issue on the mikrotk end... it can be reproduced quite simply: if we max out all our mlppp-pppoe-links (using bw-test) then we see dropped links which are established immediately, but the linkd-down causes loss in bandwidth and ...

hknet

it's a cisco issue, the box believes it has to drop link though fully loaded...

hknet

Hi we test MLPP on ROS 6.41.4 using two ethernet-interfaces and a PPPoE interface that uses those interfaces and connect successfully. We tested using 2 and 4 (i.e. using eth1 twice and eth2 twice) links. the interesting phenomenon is that we always loose the second pppoe-session (pppoe1.2) no matte...

hknet

Did not found how to pm here. Wonna honest answer? Go mx80 for your business.

second that. if you could talk your upstreams into limiting your bgp-view (ie only local routes and no full feed) you might still have fun with CCRs.

hknet

We see a development (and others too) port scanning for winbox 8291 is dropping while port 2000 (bandwidth-test-service) is going strong.
Also telnet ist still highly active with those hijacked mikrotiks.

hknet

please enable the dude to connect to routeros devices on other than the default port (eg. multiple devices on different winbox-ports or multiple devices behind one nat-gateway and portmapping active); connecting to those boxes via winbox is done using ip:port but the dude does not allow this, specif...

hknet

Just to make it clear: only devices running a not up-to-date RouterOS version are affected, whose HTTP port (TCP/80) are open and provides the login facility and management GUI, right? I never allow unencrypted connections and always disable the HTTP and HTTPS interfaces. Only SSH and Winbox is ena...

hknet

as we are approaching v6.42 is no-QinQ-in-hardware still the case with the upcoming release?

hknet

Well "Nexthop self" in an bgp environment is an abomination. It should not be needed - agreed, but real world teaches us: One looses an external interface and therefore the nexthop is removed from the IGP. In a small network this converges fast and causes virtually no service disruption. Think about...

hknet

Hi. RR's need not to be in data path (most often aren't) so please consider your own setup before fiddeling with above statement. ahem, the nexthop delivered by RRs was not implying the nexthop in fact is the RR, in fact the nexthop is usually the IP set by "next-hop self" (or similar) by BGP-route...

hknet

Hi saw this thread reactivated :) The only workaround we have seen so far for iBGP IPv6 routes to get active is to add a static ipv6 route for the loopback IP for the next-hop delivered through the route-reflectors. (again tested with RouterOS 6.40.2) If there is any workaround, I'd be glad to hear ...

hknet

as we won't get ROSv7 anytime soon, anyone with knowledge how to crosscompile for the CCR platform and build a system to run quagga or similar? But does it support multi thread? Last time I read quaggas docs it says that quagga is planned to support but so does not have the right libs for that... T...

hknet

Hi tunnels sound like a plan, you then could talk ospf on those tunnels and between routers A+B and C+D. In order to get your upstream in the datacenter to make connectivity redundant you would need to speak some routing protocol there too, usually BGP might be an option, alternate you can form VRRP...

hknet

if you just want to suppress this advertisement - do it using a route-filter.

hknet

Posting questions and reading answers in forums is already doing some reading. :) Where did you get the two ISP idea? I'd suggest doing some reading on BGP, two ISPs having the same AS is not two ISPs :) probably from: add name=toISP2 add name=toISP1 if you would like to have two links to one ISP a...

hknet

sidenote: you might also notice that you won't see the advertised prefixes in bgp advertisements as vrf advertised prefixes aren't shown there.

hknet

I'd suggest doing some reading on BGP, two ISPs having the same AS is not two ISPs

hknet

The mentioned 154 Mbps by your carrier make the access technology smell like ATM. LACP over ATM might create some issues :) Therefore as you have a CCR you might ask your provider to host a small mikrotik for you at one of their PoPs near you. using eg. EoIP-tunnels over each 100M link you get you c...

hknet

as we won't get ROSv7 anytime soon, anyone with knowledge how to crosscompile for the CCR platform and build a system to run quagga or similar?

hknet

"ppp optimized for multicore" - on all platforms?

hknet

Is anyone else having issues with the tools > Winbox tool not showing up on the client for the Dude? I tried version 6.38 and 6.39rc4 and am getting the same result. I upgraded from the previous version 6.37.3 and had no issues. Thanks, Lynn the end of the starting post should help :) -> winbox can...

hknet

any details on the obviously also new RouterBoard Firmware v.3.36?

hknet

Hi, first of all: loving it ;) specific to CHR I'd ask for a way to automate deployment, be it to do a dhcp-client-request on the first interface found and get a config file specific for this box by eg requesting a tftp-file based on the mac# of this interface. and for an easy system-reset a specifi...

hknet

Hi
to add to our l2tp-server experience: we have fastpath active - fine.
We also "allow fastpath" for the l2tp-server.

Though as soon as a client connects fastpath is deactivated.

edit: fixed it, got to set "change tcp mss" to no it seems, not quite sure if this is a working way to go...

best,
hk

hknet

Problems with server l2tp!!!

fyi - tested l2tp-server on CHR v6.36.3 - works fine - maybe you should elaborate your problem.

regards
hk

hknet

We'd like to use some model-configs built for MikroTik CPEs. At the moment a simple way to roll this out is backing up the whole config and restoring it into the new devices, downside is the MAC# are also copied. Is there a sane way to roll out custom "default configs" or more ideally is there a bes...

hknet

ahem, is there by design no dude-client for 6.36.3?
(just let me know if going back to 6.36.2 is the way to go pls)

hknet

Hi
it seems the dude client is missing for now?

Regards
hk

hknet

Hi
I'm also quite impressed with CHR, the only thing that got me thinking so far is the KVM inside the CHR

Should this work somehow or should it just be removed?

Regards
hk

hknet

Hi currently testing your script on four small RBs. 2 x RB750UP - installed and works fine. 1 x hEX PoE lite - installed and works fine. another hEX PoE lite yet fails: /sys scr run updateBlacklist status: failed failure: closing connection: <400 Bad Request> 172.102.241.58:443 (4) Therefore I'd ask...

hknet

This initiative by IntrusDave makes for an interesting read! I'd ask IntrusDave to consider delivering this blacklist in another format if possible, this would allow different use-cases, especially multiple 10G+ uplinks make it hard to handle stuff using firewall policies and blackhole-routes would ...

hknet

http://wiki.mikrotik.com/wiki/Manual:Interface/SSTP

-> only one portforward on your side needed at the firewall you can control and you're done.
(lots of things to config though)

hknet

and there's one more thing... in case you have routing-marks for anything anywhere in your system, this will prevent fastpath, even if it's an inactive RIP-setting ;) and - as far as I found out - there is no way to clear the routing-mark table in the kernel via RouterOS and therefore you're stuck o...

hknet

YMMV, but we usually prefer per packet load-sharing (ppls), especially for customer-connections. typically because: a) lower cpu overhead, b) independent layer3 connections (monitoring is easy), c) policies can decide for different paths in case then again, while activating ppls is just an interface...

hknet

well after several support-mails to and from Mikrotik (thanks guys!)
we got fastpath active - the last issue was an ipv6 firewall policy - this also breaks fastpath.

hknet

I'm a little talking to myself here :) To make the analysis a bit harder in terms of performance, if I read all this stuff correctly one has to have onnectiontracking on auto in order to get fastpath/fasttrack. On the other hand onnectiontracking seems to take cpu cycles, which is again not good for...

hknet

It is best to avoid BFD until v7 is released.

well, removed it for the time being, as I like stable

might try to run it in the future, if MT care to comment on the issue.

hknet

Ah has that been implemented? That was a feature I once requested... Will certainly have a look into that! However, currently I am not able to check for new RC version, maybe it has been taken down due to discovering a serious error? releasenote for 6.37rc10 says: snmp - added script table which ex...

hknet

we try to be good admins and do comment interfaces, instances and so on... quite unexpected is the behaviour if I add or change a comment of an ospf interface definition via winbox's fancy yellow comment icon. pressing "enter" to finalize the comment kicks the ospf neighbor association and forces a ...

hknet

well, snmp by the standard would of course be fine, but the current RC promises to have a snmp to script interface which should allow for some interesting private OIDs

hknet

just as a "fun fact" note: RouterOS v6.36 on a CCR1036 speaking bfd on a single 1G copper link (OSPF) with a Cisco (IOS 15.2(4)S7) does seem to carry a (at least for now) stable BFD interface (29 minutes and counting)

hknet

Sounds in the end you'd monitor your routereflectors

*scnr*

hknet

second that!

hknet

Hi pe1chl,
I bet you, you don't want this hovering-feature on bgp peers transmitting a full v4 (or even v6) table as this would simply take quite long

Regards,
hk

hknet

Well I'm finally totally puzzled about fastpath/fasttrack - in /ip settings we see this: https://oc.kapper.net/public.php?service=files&t=c28166da8ec363eca0adadfb288edf02&download while on the other hand the interface stats show: https://oc.kapper.net/public.php?service=files&t=80d3e250b65378fb33d16...

hknet

Hi,
no language problem here, he still has access

I can only second your opinion on MikroTik's BGP, while it likes restarts for changes to its configuration it on the other hand needs an extra push to use updated filterrules

regards
hk

hknet

not ruling the junior out, but I'm a quite trusting guy on the other hand, because people at our company are allowed to make errors and not get expelled

on the positive side: we now do extensive bgp monitoring for the mikrotik boxes in our network...

hknet

Dear Pincio, thank you for updating your code, maybe you'd like to put this together in one command request like this: my @cmd = ("/routing/bgp/peer/print","=status=","?remote-address=". $ng->get('peer'),"=.proplist=prefix-count,state"); this way you can simply check the results like this: result: =...

hknet

to fast-path or not to fast-path... Dear MikroTik could you please update http://wiki.mikrotik.com/wiki/Manual:Fast_Path ? in RouterOS 6.36 on CCR1036 we currently see: VLAN outgoing traffic is not fastpath, while the interface the VLAN is tagged to shows its packets sent via fastpath? on the other ...

hknet

Thanks Cha0s,
this definitely did the trick!

finally the prefix-count can be checked for a remote peer IP like this:

my @cmd = ("/routing/bgp/peer/print","=status=","?remote-address=". $ng->get('peer'),"=.proplist=prefix-count");

regards
hk

hknet

Hi
would you care to share how you put those filters in place?

thank you in advance,
hk

hknet

https://github.com/pincioc/check_bgp_mikrotik_nagios Hey pincio, thanks for the API idea, though if possible I'd ask you to go with https://github.com/efaden/MikroTikPerl as the api library. (ssl support and so on) based on your script I was able to check the route-count in a specific vrf: my @cmd ...

hknet

after reading this thread I'm still trying to make sense, so I conclude: OSPFv3 and Loopback-bridge-interfaces with /128 IPv6 addresses assigned in RouterOS will only be shown reachable if one sets an admin-mac to the bridge (named eg Loopback0). well, that's not 100% intuitive, but I guess that's s...

hknet

maybe some kind soul at MikroTik could update the wiki page http://wiki.mikrotik.com/wiki/Manual:API#Query_word to mention the count-only attribute to queries, it took me hours to figure this one out...

hknet

thanks, but this is the linux-style cpu-load value imho, therefore no percent usage for the whole box.
also getting 72 cpu load values and making some sense for alert values makes me quite clueless here.

hknet

well, I'm a bit lost, anyone knowing which perl API might be maintained? it seems the later "MikroTik.pm" is no more working for scripts originally based on the "Mtik.pm" by Hugh. our nagios checks use the older lib. now I ran into a problem - I'm trying to implement something like: /ip/route/print ...

hknet

Hi we're trying to monitor critical values of CCR systems in our network. the idea was: there's linux running, so we should be able to monitor ample parameters. well, wenn it comes to CPU monitoring only one value is readable: ssCpuSystem.0 therefore I'd like to know: how to monitor cpu usage? and o...

hknet

Hi strange thing happened to us today: CCR1036 running RouterOS v6.36 today suddenly disabled (all by itself!) a single BGP instance running. We investigated the issue, but the system is only manageable via private IPs and not exposed to remote access, therefore we are quite sure it did this thing b...

hknet

It is indeed a slippery slope. What if MikroTik took us to the test and started some outside funding project on kickstarter (or whatever crowdfinance portal) in order to get major development done? This way we could (hopfeully) keep our beloved MikroTik HW+SW, but we could help advance their efforts...

hknet

What is the business reason to force the DFS on? Mikrotik should explain it really in details. There are rules that when radar is detected, the Ap should lower its tx power, sometimes, so there is not really necessary to leave the frequency. When radar is stopped, it can revert the tx power back to...

hknet

There are several not so subtle things broken too. E.g. - BFD - L3VPN + PE-CE BGP NLRI updates (NLRI updates do NOT occur when the best path changes) - IPV6 recursive next-hop lookup - VRF + BGP Passive Peers - Advertised Routes for BGP peers in a VRF - BGP often stops retrying connection attemp...

hknet

We'd a similar issue, as we were running two RRs we could easily show to MT it doesn't work as expected.
They told us we should wait for ROS v7 as this is the only way they would fix this issue.
ergo do not try to use MT as RR for IPv6 running ROS v6.xx

Search found 105 matches