MikroTik

hashbang

who will buy a $999 100g switch for basic switching only cannot do h/w offloading: mpls vxlan qinq bonding modes split horizon Port protection (bonding active backup) still dont know who to make acls with too many rules in single acl erps g8032 I purchased a switch 24 x 10g + 4 x 100g switch for the...

hashbang

hi,
ty,
Rt bridge id on instance 11 which is the instance active here shows 0000. It should show mac of route bridge and also the cost too.

int_print.png

bridge_int.png

regards

hashbang

Hi,
I see missing route bridge id on crs 317 running mstp mtik version 7.10.2. Is there any incompatibility with other brands of switches ?. its connected to huawei. Please find below screenshot.
ty and regards

hashbang

mpls - added FastPath support;

wht will be the advantage of this feature. We are planning to use crs 317 for mpls/vpls . Will it benefit in P case scenario
ty

hashbang

ty,
Couldnt find fixed netinstall 6.48.3.
Any tutorial how to do it with VM ?
The most bizarre thing is Mtik is silent and keeping its eyes closed. I'd already discussed this problem some months back on support email. I'd got 20 x86 server to be upgraded.
ty

hashbang

Hi, I'd always failed to install version 7.xxx on usb drive. I get error WhatsApp Image 2023-06-08 at 16.55.15.jpeg no issue with version 6.xxxx Now the problem is u cannot upgrade 6.xx to 7.xxx on x86. It looses licence. Your server goes down for alteast 4 5 days till I get replacement key You cann...

hashbang

I tried upgrading CRS 317 from 7.6 to 7.9. It failed lost the switch. After connecting with console. Switch was up but os version was old. Resetting the switch brought back the switch. The reconfigured it with older versihn 7.6
ty

hashbang

4 5 reboots a day

dont try 7.4++ for pppoe on x86

hashbang

Ok! Its a bug in 7.4 and 7.5 atleast. Whenever there is network flapping and lots of user gets disconnected. Queues dont get deleted and when user relogs it fails to connect bcoz of queue already there. Already sent details to support. I got the answer after long time "upgrade to 7.5" it w...

hashbang

You do CGNAT or all pppoe users have it's own Public IP?

ty,
natting with same target on mikrotik

hashbang

Hi, Recently on x86 version 7.4.xxx running pppoe server, I got this error login failed: failed to add queue: already have such name . I have other x86 running older versions 6.4xxx they never got this error. Only reboot resolved the error. Looks like not able to handle so many pppoe users. Earlier ...

hashbang

ty,
I dont find any reason why it has been deprecated
regards

hashbang

Hi,
In new 7xx version of user-manager I didnt find change password option in web interface. How user-manager user can change their password using browser ?
thanks

hashbang

hi,
Its x86 so no hypervisor. All huawei switches.
ty

hashbang

Hi, Recently I had a minor outage in the network. The cause was found Mikrotik running 6.47.1 on x86 with 2 10g cards became the root bridge. The mikrotik is running as pppoe server, no bridge, no stp. Just vlan interfaces and pppoe server. PortID=29, PreviousRootBridgeID=32768.bc62-0e0b-5f90, NewRo...

hashbang

Hi,
CRS doesnt support dynamic stp cost be it any speed. Its a feature request for adding dynamic cost of interface as per 802.1T.
If the interface is down (intial configuration) then u cannot change the mst port overrides which defaults to 10.
thanks

hashbang

Hi,
Running crs 317 ver 7.1.3 protocol mstp . I have 3 edge ports with p2p enabled and 4 ports involved in mstp. Disabling any of the edge port gives drops in ping (mgmt vlan). It shouldnt be like this. Any inputs on this.
thanks

hashbang

Hi, CRS 317 ver 7.1.1 started giving error snmp timeout while waiting for program xxxx. Winbox didnt showed any thing. No files to be seen. Rebooted the router still nothing seen in winbox, same with telnet. /int pr command stuck. Somehow managed to upload new version 7.1.3 upgraded it but issue did...

hashbang

Hi,
In CRS routeros
MSTP instance = 4
CRS = 4 switches
instance 1 root bridge crs1
instance 2 root bridge crs 2
instance 3 root bridge crs 3
instance 4 root bridge crs 4
how to do this
thanks and regards

hashbang

apologies tdw it was mstp with other brand mstp (not rstp) here is huawei manual for more info On an MSTP network where Huawei and non-Huawei devices are interconnected, if the Huawei and non-Huawei devices have the same region name, revision level, and VLAN mapping table but different BPDU keys, th...

hashbang

The screenshot showing sfp16 role root port indicates this is the path to the root bridge, it is not the root bridge itself - something else is. The path costs are shown in the screenshots, and you have actually set some as the configuration has entries such as path-cost=20000 . These are the cost ...

hashbang

Unlike other vendors which change path cost depending on interface speed the Mikrotik default is 10, just set it to what you wish it to be. The recommended defaults are widely available, e.g. https://en.wikipedia.org/wiki/Spanning_Tree_Protocol#Path_cost Note path costs have nothing to do with root...

hashbang

Unlike other vendors which change path cost depending on interface speed the Mikrotik default is 10, just set it to what you wish it to be. The recommended defaults are widely available, e.g. https://en.wikipedia.org/wiki/Spanning_Tree_Protocol#Path_cost Note path costs have nothing to do with root...

hashbang

even basic things are missing 1) How will mstp interact with other switches running RSTP 2) I couldnt find instance cost on ports settings. If anyone knows then plz tell me 3) No idea wht costs it sends 4) in above screenshot this switch has prio of 8000 then still its becoming a root bridge. how wi...

hashbang

Hi, how to change path cost in crs 317. No matter wht you do it sends 10. I'm having wrong root selection and its total mess. region 12 root should be x 0/0/4. network <.........><Huawei(A)>.....<CRS317>...............<Huawei(B)> HUAWEI A : 0 0.f898-ef73-c061 10000 2 20 15 XGigabitEthernet0/0/4 .......

hashbang

ty, done but still the same

Screenshot from 2021-10-04 13-50-38.png

hashbang

ty
please find screenshot

Screenshot from 2021-10-04 13-33-05.png

hashbang

this is what i did now simple scenario

Screenshot from 2021-10-04 13-18-40.png

no mac address on sfp 2
and no mac of 2607 on uplink sfp 16

Screenshot from 2021-10-04 13-18-40.png

any inputs
thanks

hashbang

I can imagine only two possible explanations why a single rule is sufficient on the Huawei: when learning the MAC address into the forwarding table for VLAN 2607 from an ingress packet with VID 167, the switch stores a remark that the VID has to be translated from 2607 to 167 on egress the VID tran...

hashbang

As others have said, the description doesn't exactly match the drawing, but if I understood what you're after, this is probably the config you need... note: I haven't tried this and it's been a minute since I've done VLAN translations so you might need to adjust this but it's a starting point ### M...

hashbang

Hi,
I want to do vlan translation from sfp 2 on crs 317 where I get tagged 2607 and 167 vlan. I want to map 167 to 2607 and out to trunk port sfp 16

crs317.png

hashbang

Without seeing the configuration from /export hide-sensitive it is difficult to say, likely missing port ingress filtering.

ty

crs317.rsc

hashbang

Hi, using crs 317 6.86.4. Eventhough sfp 16 is not allowed with vlan 2606. I see mac address on the this port with vlan 2606. /interface bridge> pr name="bridge" mtu=auto actual-mtu=1500 l2mtu=1584 arp=enabled arp-timeout=auto mac-address=2C:C8:1B:2F:C8:56 protocol-mode=mstp fast-forward=y...

hashbang

Hi,
I went the the mib but was unable to get oid for low power threshold of sfp module. Please add this oid
thanks

hashbang

Under /interface bridge you can either change pvid= to something other than 1, if you wish to have a different untagged interface on the bridge, or add frame-types=admit-only-vlan-tagged ingress-filtering=yes to remove untagged membership of any VLAN.

ty

hashbang

yes
Its our normal practice in large networks to remove vlan 1. Testing crs as another option

hashbang

Hi,
I want to remove default vlan from the bridge. How can i accomplish that ?

thanks

hashbang

Hi,
Had anyone done mpls on crs317 or 326 and wht is the performance as PE or P device

thanks

hashbang

Hi,
I'm having pppoe problem after adding vlan. On default vlans it runs fine. The problem is
PPPoE connection from xxx was already active - closing previous one. Its x86 server running 6.48.1
Tested with only one session = no
Can anyone help in where the problem is
thanks

hashbang

HI,
I request to add oid for SFP rx power low/high threshold
thanks

hashbang

hashbang

HI,
Please add support for mstp on swos
thanks

hashbang

Hi,
What is the oid for sfp low rx power threshold.
thanks

hashbang

HI, What kind of switch is this STP cost factor 10 (no dynamic changes on the port speed basis) STP loop guard missing ( I tried couldnt find in forum or doc) Flip thru docs to see wht can be done by switch controller and what will be done by cpu You have to write down a whole rule to just block bpd...

hashbang

Hi,
After the upgrade I'm not being able to advertise ip addresses in ospf. Running on x86. Is anyone having the same problem. I can see the peers on both side but no advertisment.

thanks

hashbang

hi,
same problem in 6.44.2
Interface bonding1(): Speed 2020-04-15 14:33:25 1 Gbps
any reply from mt when this will be resolved ?

hashbang

TY! there are 1000 pppoe connections that means 1k mac addr. I suspect missing modes of bonding in MT. So this problem only comes in encapped traffic like pppoe encap2+3 This policy uses the same formula as layer2+3 but it relies on skb_flow_dissect to obtain the header fields which might result in ...

hashbang

ty,
but when i create lag between huawei(switch) to huawei (router). It works fine but between huawei switch to mikrotik it doenst do load balancing. Any other method to make it working. Balance RR dont work.

hashbang

Hi,
What hashing method to be used on bonding 802.3ad for pppoe encapsulated traffic. I'd tried l2 l2+3 l3+4 but the traffic doesnt balance equally or so. Its something like 90%+10%.
thanks

hashbang

address list having random /32 ip address and mapping them to netmap subnet.

hashbang

Hi
how will source/dst address list will work with netmap target ??

hashbang

its all default 10s.
Reboot resolves the problem

hashbang

Hi, Recently it has been found on platform x86 pppoe-users dont logout when there is a disconnection in physical connectivity (router reboot etc). You have to manually disconnect from winbox. Earlier I thought its the becoz of memory consumption. As it happened on one of the MT running 6.44.2 but to...

hashbang

Hi, There has been memory leakage in all the version beyond 6.4*. Earlier I'd uptime of year or more. Now I have to reboot every 1 month. Memory consumption goes down to dangerous levels. My h/w All Dell, IBM etc server on intel x 82599 cards. Running as pppoe server Contacted support but same old s...

hashbang

@ghusson As you are a consultant for proxmox. Can you please tell me the performance of proxmox on Mikrotik running as pppoe-server. I got very poor performance on all the servers. Mikrotik couldnt give me any answer other than "try hyper-V". In proxmox performance degrades after 800 pppoe...

hashbang

Hi
Need help in installing chr on xen server. I tried importing vmdk image in xen center but it ends up in error failed to import.

ty

hashbang

I'd been struggling to get any kind of acceptable performance from chr on proxmox. Sever with fairly good configuration, dual Xeon 26xx at 2.2 ghz 8 cores 16gb ram intel x520 10g running proxmox. Mikrotik chr is running as pppoe concentrator. Some vlans and pppoe services running on them. User logge...

hashbang

just installed this version on x86 pppoe-server showing red in color where as server is running ok. people can authenticate surf. Seems like cosmetic bug

hashbang

On Dell R-220 xeon 1220 I'm running 1500 pppoe sessions and 1200mbps that is max on intel x520 10G lan cards

hashbang

Hi,
I installed mikrotik 6.38.5 on Hp dl-360 server but it shows only one core whereas it has 28 cores (14 x 2). Using any live linux cd for booting it show all the cores available.

hashbang

Looks like I have the same problem

http://forum.mikrotik.com/viewtopic.php?f=15&t=117306

hashbang

thnx, That means virtualbox have performance problem. In my test I found latency jitters, high cpu usage when using paravirtualized drivers. Performance seems better when using intel card but then it strips the vlan tags. How good is kvm ?

hashbang

Hi, I tested chr 6.38.1 and x86 on the same server with virtual box. Virtual box setup : Dual Xeon 56xx 32gb ram. (total of 16 cpu shown in MT) Centos 7 running as host with latest intel drivers x520. Paravirtualized drivers used Virtualization acceleration used KVM 2 bridges one for lan and other f...

hashbang

repeat the exercise with 6.36.4 everything is working fine. Later on added the key and upgraded to 6.37.1

hashbang

same problem here
fresh install on pen drive
version 6.37.1
dell 710

hashbang

same problem here dell 710, usb pen drive version 6.37.1
1) installation
2) 1st boot ok
3) 2nd boot ok config done
4) 3rd boot demo expired and reboot

hashbang

did you mean accepting in filter table ? If yes then I didnt tried but no problem after shifting to x86

hashbang

/tools profile

hashbang

scenario is more or less same as my post sluggish behaviour. Now a very strange problem. 1200 users running on pppoe. Ping to 8.8.8.8 (running on peering connected with 10g fibre) there were drops from subscriber side. I connected laptop on ether6 (ccr-1036) to check if its the problem with network....

hashbang

I just got up and read this post. Same kind of thoughts were going in my mind when I got up and I found this discussion. My already recent post regarding sluggish performance as BRAS. In MUM Mikrotik people gave all the diplomatic answers. Questions were : 1) What x86 h/w do u suggest as CCR is insu...

hashbang

no tested till 3.6 then moved over to x86 but I have more cases. Just upgraded ccr 1036 having some strange problem. Underplan to shift to x86

hashbang

Hi, Scenario, Mikrotik running as PPPOE AC, users : 1500*, authentication method : radius , Queues : users + 4, Vlans : 8, PPPOE-Server on each vlans, Natting rules : 10*, Mangle rule : 2 + MSS, HW : CCR-1036-8g, Bandwidth : 1200-1400mbps I got complaints about slow speed. I checked the cpu usage it...

hashbang

A small feature request in winbox : 1) A tab inside winbox to show all the windows panes opened inside winbox. As all the windows overlap eacher other it becomes tedious to change from one pane to other. For example : bgp, filters, routes. When they are opened there should be tab somewhere on the to...

hashbang

hi,
Mikrotik iso/chr not taking any data from vlans on Vbox 5.1.
Server Dell R220
Network settings
Mode : Bridge
Driver : Intel pro 82450
Promiscous : Deny & Allow all & Allow VMs

On paravirtualized drivers cpu usage very high. Through less than half compared to intel pro

hashbang

Hi,
I"m having high cpu usage on CCR-1072 when using torch. It consumes all the cpu available. Normally its 20-30% but using torch it goes upto 100%

hashbang

queues are resource hungry. As soon as traffic passes beyond a cetain limit the consumption of cpus goes up to alarming levels CCR-1036 : max 200kpps passing though a single queue is enuff to take all 36 cores to 100% iptables mangle rules like change tos, packet marking are resource hungry it reall...

hashbang

Hi, This problem came up last year. Mikrotik only suggestion was to upgrade the router. I was saw this problem 6.23 then upgraded to 6.27. Now on a different router different place version 6.30.4. Mikrotik had failed to give any explanation why this is happening. I cannot upgrade the router to late...

hashbang

Hi,
I'd been working on linux from last few year and winbox was running just about ok. Now Im running windows 7 64bit. Winbox 3.4 crashes when downloading plugins. Kindly plz help

hashbang

Hi, I have 1036 running as nas ver. 6.35.4 . 1800 PPPOE users with radius authentication and 1800 simple queues for them. One single queue on top of dynamic queues fot cache hit bypass. 100kpps to 150Kpps passing through that queue. CPU usage goes to 100%. Disable this queue cpu goes down to 30-35%...

hashbang

i was simulating udp flood. Can you really stop udp flood ?? As the packet had reached to your router so your upstream pipe will very likely be choked.

hashbang

Hi, I'd already posted few times 1.5 yrs back regarding extremely poor handling of packets when CCRs are used with firewall or queues. Currently I experienced some mysterious reboots of CCR 1072 running 6.35.4. Total traffic was 5gbps download. Two rules in firewall mangle changing tos, 4 simple qu...

hashbang

make a template for auto discovery (low level discovery) of pppoe interfaces and then graph them

hashbang

Cha0s - Make sure that you have administrator privileges and in file path there are no specific symbols; hashbang - This is problem of Wine not Winbox. You have to adjust it yourself. Winbox is tool for Windows; kometchtech - This is not issue of Winbox. We will tyr to fix this in upcmoning RouterO...

hashbang

hi,
a basic feature request about column of service name of pppoe-server in pppoe interfaces and active. It will make it easy to view online connections by filtering on service name where there are many pppoe services running.

hashbang

Will 3.1 resolve cursor problem on wine? The cursor displays slightly left of actual position which makes it very difficult to edit the fields.

hashbang

Hi,
Mikrotik-Host-IP attribute to be used in access-accept/reject. Currently its is only for statistical purpose. It makes network more secure. when mac address authentication is being used.

hashbang

1g sfp are connecting setting are auto-neg = off sfp rate select=high
tested on sfp2,3,4,

hashbang

it crashed again

hashbang

i'm unable to connect ccr 1072.........huawei........ccr 1036. unable to ping huawei from ccr 1072 on 1g modules. Fiddled with all settings auto of on low high but unable to ping. Tomorrow I'll confirm. On 10g modules things are running fine. The same modules are working on all the series of ccrs. L...

hashbang

yes seems so. If u change the default behaviour of queue order in hotspot/profile queues stopped working. Ticket was raised few months back but no idea about resolution. It took time to make MT support understand the problem. There is no mentioning this bug in any of the release. I dont know why u w...

hashbang

2morrow i'll connect ccr 1072 to huawei switch with 1g module and further to ccr 1036. Will update in few days

hashbang

normis < already sent and interacting with them. they are doubting ping utility crash

hashbang

no not like that. 1) It is connected to 10g switches 2) Data was going fine but pinging stopped. services were affected because my routes had check-gateway=ping enabled. which made router think the gateway is unreachable. Explanation is one of the link got down and it had redundant routes. Routing f...

hashbang

hi, I just replaced ccr 1036 with ccr 1072, after 2 days of uptime It stopped pinging all the connected routers. I could see them in neighbours but no ping. As i'd static routes with different distances added which made all my static routes invalid because check ping was enabled. Reboot resolved the...

hashbang

and how much traffic can you pass by disabling the queue ?

hashbang

A very basic and needed feature :
PPPOE service name in the column list which can be sorted or filtered under ppp active interfaces list.
addition of rx and tx rate in ppp active list

hashbang

make a new queue type with pfifo and queue size to 200-300 and attach that queue type to ur simple queue

hashbang

do the tools/profile. How much is cpu usage

hashbang

Hi, I did trunking with crs and ccr. trunking on crs side and bonding on ccr side with balance-xor loadbalancing was working properly but people started complaining about slow and sites not opening. I myself tested speedtest.net. Sometimes it used to work sometimes given error object not found. I'd ...

hashbang

thanks, true I'll have to check y total bandwidth on bridge is limited to 950mbs where as traffic is coming from 2 x 1g port on bridge and going out to 2 isp on different 1g ports.

hashbang

hi, a simple question what is the total bandwidth bridge can pass i:e total of physical interfaces for example I'd got a bridge ether3 and ether4 on ccr. So total bandwidth should be 2gbps or 1gbps. As I'm seeing my total bandwidth doesnt go above 1gbps even if the traffic is on both the ports and g...

hashbang

sorry it was 1016.
thns

hashbang

make a support file and send it to support. Please remember CCR wont be able to take heavy traffic on queues. There is already one post by me sometime last december showing it cannot handle heavy traffic(pps) on queues. Make a queue of around 500mb and send heavy udp 64b packet through that queue. M...

hashbang

can you plz check if any of the queue is being hit by dos attack or high pps. What is the rate of drop when traffic hits ur firewall rule. Queue wont be able to handle more than 1200kpps on 1072 to maxout the cpu. Correct me if I'm wrong

hashbang

please be more clear. you are talking about 1036 and showing pics of 1009. No traffic going through interface/firewall rules is shown.

hashbang

PPPOE works well only on perfect networks. I have 1000s of users running on pppoe and mysteriously some nas / users logs out at the same time. Comparing with hotspot it has more resiliency to sustain network drops. Even latency going high will make the users logout. I'd shifted most of the users to ...

hashbang

this rule is present in all the routers but strangely one ccr 1009 having just one natting rule showed the same problem. I dropped a mac address. After disabling the rule, the mac address which was a mobile device still was blocked. Rebooted the mobile still no browsing, rebooted the ccr and browsin...

hashbang

Hi,
Version 6.24 onwards (afaik) whether you delete the firewall rule or disable they keep on working. I'd seen it on many occassions. I had to reboot the router (seen on ccrs).

hashbang

I have 4 bgp peers 3 isp and 1 cdn. 3 isp sends just default routs. CDN advertises 400 prefixes. I have to do policy routing with subnets src ips going to each of isp and all received routes desitnation going to cdn but as soon as i do policy routing rules they failed to go to cdn peer. Moreover I w...

hashbang

Hi, I have 2 isp and cdn bgp peers. I'd got 400 prefixes received from my cdn peer. Since I started doing policy routing for isp2. The subnet on policy route (with source addr subnet to lookup table for isp2)could not be reached to cdn. I filtered 400 prefixes received to a separate table "CDN&...

hashbang

I have few ccr new with 1 or 2 ethernet non functional.
Sometimes I see one ether port is connecting 1gbps and other with same cable router/switch at other end doesnt connects on 1gig only connects at 100mbps. This happened on 1009 brand new

hashbang

I'd posted same problem earlier with mikrotik 6.23-6.27. After studying the problem I found : When there is hotspot and pppoe-server running on Mikrotik on vlans and hotspot user profile setting is insert queue before=bottom then hotspot queues dont work. If I change the profile setting to insert qu...

hashbang

Hi,
H/w 1100 ahx2 version 6.30.4 running hotspot and pppoe on various vlans. dynamic queues of hotspot are not working but pppoe queues are working well. Anyone having the same problem

hashbang

hi,
very basic feature request. Prevention of duplicate ip address on different interface or warning before submitting.

hashbang

then why only few of them are giving problem out of 100 ?

hashbang

i'm talking about wired ethernet. Sxt lites have level 3 licence so cannot be used as ap.

hashbang

already explained. tweaked ether of both sxt and windows
not all are giving problems. I purchased 100 pcs few of them are giving problem

hashbang

does firewall/antivirus has something to do with physical connectivity of lan card ?

hashbang

Hi, I just got a whole lot of sxt lite 5 returned because they all were unable to connect with ethernet to pc. Strangely all the sxt's were able to connect with ethernet on linux but on windows none of them connected. They all have 5.26 version installed. Upgraded to 6.30.4 but without any results. ...

hashbang

right, after so many years the problem is still there but what is the cause. I'm experiencing the same thing. One my my networks runniing pppoe likethis MT.....switch.....nanobeam.........nanobeam....switch....subscribers. The number of subscribers is low around 20. They all experience disconnection...

hashbang

This is not a bug. Simple queue had never been working with unlimited limits afaik in all versions.

hashbang

HI,
Will ver. 7 resolve undermentioned issue ?
http://forum.mikrotik.com/viewtopic.php?f=3&t=92834

hashbang

Hi, I have 2 isp isp1 isp2 and 1 cdn provider peering with me. I get 100 prefixes from cdn provider others are just default route. Traffic is divided between isp1 and and isp2 with the help of policy routing. ISP1 traffic goes through main table and ISP2 traffic goes through isp2 routing table. Now ...

hashbang

Sorry, some confusion in my writeup earlier. framed-ip-pool was wrongly mentioned. I'd given ip pool in radius server, radius server correctly selects the ip address from that pool and sends the ip in framed-ip-address to the mikrotik pppoe-server. Still pppoe-server gives the ip address from local ...

hashbang

True bridging is not possible on wireless 802.11. Ubiquity uses arpnat to bridge the network on station side. Use wds for pure bridging to get your customer side mac address.

hashbang

hi, I'm using radius manager with mikrotik 6.28 on pppoe server. If i use ip pool on some user, mikrotik gives ip address from local ip pool rather than framed-ip-pool. I'd tested using radtest radius server is sending framed ip address correctly but ignored by mikrotik giving ip address from local ...

hashbang

hi,
we had installed few 850gx2 as nas. After creating vlans on ether2 they had started to crash. Unable to create support file as they are too far and unreachable. Mikrotik version is 6.29.1. mail has been sent to support. Had anyone had this kind of issue ?

hashbang

in rc 10 last settings are not save. I have to manually add ip address turn on the advance mode. It was fine in rc9. Running on wine. Connect to field is always empty

hashbang

Policy routing is very cumbersome adding all directly connected subnets to main table manually. Mikrotik should have local table lookup to make it easy.

hashbang

mt-6.27-1.png

hashbang

on 1036-8g after 3 weeks of uptime cpu load barely 2% doesnt show anything on print command or in winbox everything blank. Trying to make support file but its taking ages. Mikrotik version 6.27

hashbang

any resolution on this post

http://forum.mikrotik.com/viewtopic.php?f=3&t=95539

no reply from support

hashbang

what hardware are you using ?
Motherboard/ cpu

hashbang

hi, hotspot queues dont work when you have "insert queue before bottom". Topmost queue is icmp marked packets (target 0.0.0.0/0) . All traffic of hotspot users pass without queues and users get full speed. Seen on 6.24 upgraded to 6.27, hardware is ccr 1016-12s. Anybody having similiar pro...

hashbang

ty all.
All i wanted to do is advertise all the subnets to all upstream providers and prioritize some subnets on 1 upstream providers and others on 2nd upstream.

hashbang

FYI there was a simple queue on top with target 0.0.0.0/0 with packet mark for bypassing icmp traffic from hotspot queues. Strangely this queue was not allowing the lower hotspot queues to pass traffic through them. So there was no traffic in hotspot queues. Changing the target from 0.0.0.0/0 to ip ...

hashbang

hi,
doing connection mark with icmp proto it marks other proto too udp tcp etc.platform ccr 1016-12s version 6.24. Hotspot queues stops working. Support file sent to support

ty

hashbang

outdated list

hashbang

Hi,
I have 2 upstream provider with upload going on ecmp. IPs subnets advertised with prepend-path for incoming traffic. Even though everything is right but still incoming traffic comes from both the ips (if i test it from route views). When do prepend-path fails to deliver what we want to accomplish.

hashbang

Normis < how long is not too long ????

hashbang

Hi,
Does Mikrotik support intel x99 chipset and i7 5960x cpu ? Which board someone can recommend and which version of MT

thanks in advance

hashbang

tried with or without ntp package. Servers were domain names and were resolved properly.

hashbang

hi,
like others on this forum. I'm having problem with ntp updates. Router is CCR-1036, tested on 6.24, 6.27. Tried so many servers but update not happening.
anyone can shed some light on this problem

thanks

hashbang

winbox rc2 shows all the filters in routing as 0.0.0.0 (prefix) where as in cli everything is ok. Tested on ccr-1016. 2 people at different location seeing 2 different things with diferent version of winbox of the same filters. Too lazy to edit the screenshot and post it here.

hashbang

Hi,
I tried to install MT 6.25 on HP Dl-380e gen 8 server. It failed to detect hdd. Maybe MT doesnt support SAS drives.
Any help will be appreciated

Vyos was successfully installed on the same system

hashbang

1.2 Mpps (syn) is not a bad amount of load on ccr 1036 but too bad if it hits any of the simple queue and see ccr 1036 bouncing on the table

hashbang

ccr 1016 screwed up @ 300k pps udp 64b packet by a simple queue of 900Mbps. Support asked me to upgrade to 6.24. Even after upgrading problem still is same

hashbang

As soon as there is ddos attack or high packet rate on simple queue CCR cpu usage goes to 100%. My other post testing ccr have details

hashbang

Looks like i'd hit another soft spot of CCR. On the above test I found that no traffic was passing thru the queues. I disable all queues and enabled just one queue with target 0.0.0.0/0 dst ether6 max limit 900k and there it goes. The cpu usage rose to 100% winbox disconnected no telnet no ssh. I wa...

hashbang

here is my test running with 4 xeon server with iperf, packet size 64b, proto udp and 10 parallel connection. It reached 2.7Mpps, cpu usage 25%. Though it showed erroneous cpu usage under winbox title bar after doing the tools profile the cpu usage got corrected too. Config : allow fast path off con...

hashbang

eagerly awaiting for the results
Mikrotik should make cross platform traffic generator or atleast post the source or rpm/deb package for the same.

hashbang

More Updates added 2 more xeon servers running centos and iperf. Traffic reached 2.6gbps aggregated on sfp plus 1 with cpu usage 1 -2%. Three simle queues were added with target 0.0.0.0/0 on the interfaces where tests were conducted. Conntrack was on. Details of test may come up late as i forgot the...

hashbang

Updates on the test

one pc running windows was replace by hp dl 320 server running centos 7 . Centos was running iperf as client. I got
600mbps* and 700kpps throughput. CPU usage hovering between 1 to 2% (conntrack on) More tests will come up. Preparing 4 hp-320 servers for testing

hashbang

test was conducted for 900 sec. and many times. will try again 2morrow. MT tester is behaving strangely cant connect with packets size 200 or below udp

hashbang

Hi, Suggestion needed for testing ccr-1036-8g for sfp+ ports. I have 2 ccr 1 40ge+2 10ge port switch I need help in testing sfp+ ports as some people are dissatisfied and I'm about to put into production. What software to use for testing It should be cross platform i;e windows and linux. I tried ipe...

hashbang

uhm... Thanks, but no, this is a router in actual production networks. Well over 5K established TCP sessions. CPU sitting at less than 20% also, so it's not a load issue either. A simple single threaded wget in linux gets over 3Gbps. So yes, it is most certainly a Mikrotik issue. If Mikrotik's prod...

hashbang

Suffering with same issue. I have ccr1036-12g-4s with sfp modules, setting up to 1g link manually set the LINK to OK. but no data pass. Tried every suggestion mentioned here and at forums, but no luck. Can you please explain how you had connected and which modules you are using. This will benefit o...

hashbang

hashbang

hashbang

so queues does have some problems with handling multicores (for what i think). I recently did a test A(pc)...........ccr1036(vlan).............B(pc) flooded the packets with hping3 from A to B and to my surprise the cpu got 100% @ 200kpps with tcp. I was shocked to see the cpu usage. As cpu was 99% ...

hashbang

Hi,
While going thru health checks of ccr 1036-8g. I found sudden rising of cpu usage for unknown reasons. I was unable to do profiling as it was detected in graphs. Normally the cpu usage hovers around 18%. CPU nearing at 100% at extemely lean hours 2.30am

hashbang

Hi.
Is this kind of slow escalation in memory is ok ? As on x86 its not seen after months of uptime. Screenshot of ccr-1036-8g ver 6.19

hashbang

thanks all for this discussion Port throughput max of a single gigabit port = 1Gbps down + 1Gbps up = Full-Duplex 1Gbps non-blocking internal connection Traffic that the CPU processes is 1Gbps in total doest that means tilera can only process 1gbps of total bandwidth on 1gbps port and 10gbps of tota...

hashbang

Brocade specs says 6610 has forwarding capacity of 432Mpps.

hashbang

Hi, This is my confusion. Anyone help me clear it. Why should anyone buy ccr1036 ? ccr1036.........1000usd approx practical thru 2Mpps practically Layer3 Switch for example brocade 6610 price approx 4000 usd thruput 100s of mpps then what is the reason for layer 3 switch being so cheap ? Why should ...

hashbang

This is device throughput - amount of traffic that device can receive, process and then send out to next device.
The same traffic is entering and leaving router - so why count it twice??

routers are meant to forward data. What use a router is if it cannot forward. So why count twice ??

hashbang

:O
if i'm using 1036 with just 2 ports lets suppose ether1 as lan and ether2 as wan. Will 1036 deliver 1gbps up + 1gbps down on ether1 (lan) ??

hashbang

MT seems to be not interested in x86 arch. The supported H/W list needs updates. Found problems with newer chipsets. Still in darkness about support for newer chips like e350, x99 etc

hashbang

running bgp with 3 peers, 1gbps bandwidth running, version 6.19. Running from 17days uptime (rebooted because of upgrade). The only problem i faced was bgp filters not behaving the way it should. Memory consumption increasing. It started up with 34xxMb free and after 17 days its 3260Mb free. I'm che...

hashbang

better late than never someone responded from MT, Thanks. Main question is why 1gb ports get saturated @ 512Mbps+512mbps (approx) why not 1gbps+1gbps ?

hashbang

one more question :
As 1036 is capable of doing 28mbps. Testing 1036-8g-2s+ with just 2 ports of sfp+, will it give 20mbps aggregated throughput i:e 10mbps up 10mbps down.

hashbang

In benchmark of 1036 the max throughput is equal to total of ports CCR1036-8G-2S+ 8 ports gig + 2 ports 10gigs = 28gbps. In benchmark max throughput is 27,942.9 CCR1036-12G-4S 12 ports gig + 4 port sfp = 16gbps. In Benchmark max throughput is 15,963.3 Does that means aggregated throughput of each po...

hashbang

No noticeable change
peak cpu usage 24%

hashbang

ty, Rule orders changed /ip firewall filter add action=add-src-to-address-list address-list=level3 address-list-timeout=\ 6h chain=input comment=l3 connection-state=new dst-port=22,23 protocol=\ tcp src-address-list=level2 add action=add-src-to-address-list address-list=level2 address-list-timeout=\...

hashbang

/ip firewall filter add action=drop chain=forward comment="zeus drop" connection-state=new \ dst-address-list=zeus add action=add-src-to-address-list address-list=level1 address-list-timeout=\ 1m chain=input connection-state=new dst-port=22,23 protocol=tcp add action=add-src-to-address-lis...

hashbang

5 rules to be precise. 4 rules for bruteforcing of telnet and ssh as per wiki
1 rule for blocking dst address list which has 142 ip addresses (zeus network block)
1 rule for making packets based on dscp 14
no nat rule

hashbang

Mine is having 200k pps and 950mbps, cpu usage reaching peak 24%. Screenshots below

hashbang

which part ? Does MT needs more optimization ? Practical thru is noway near what tilera advertise.
MT is silent about release date of ccr 1072

hashbang

I installed ccr 1036-8g-2s+, a brief explanation of config : 1) 2 bgp peers + 6 filters 2) 2 isp (ether 7 & 8) 3) 2 lan outs with one lan having 10 vlans (ether 4 & 5) 4) 41 queues 5) 3 rules for ssh bruteforcing protection 1 rule for packet marking. all other ports are free Bandwidth runnin...

hashbang

I forgot to mention I'm using 1036. OID you had mentioned will only graph cpu 1. I need oid for total cpu usage i;e cpu-load.

hashbang

why there is difference in cpu usage seen on winbox (below titlebar) and /tools/profile.
On winbox it shows 12% in profile it shows 82% idle.
which should be considered accurate

Is there any OID for snmp monitoring of total cpu usage (zabbix etc)

hashbang

adyb76 please post the screenshot
vox +1

hashbang

+1 winbox 3 is buggy. It doesnt run properly on wine. The cursor will blink on the extreme right of ip address but if u use back space it will delete the 1st octet of ip address. So painful for anyone who uses keyboard and doesnt watch the screen :P. Specially when u have a long list of devices in y...

hashbang

i found a minor bug in bgp filters. I dont know whether it was display or it was really the config. Router ccr 1036 8g version 6.15 BGP with 3 peers 2 peers to isp - isp1 and isp2 1 peer to client - c1 yesternight I added c1 to peers and added his subnet on filters accept. Strangely the IPs which we...

hashbang

then atleast please update the chipsets supported

hashbang

does basebox 5 support custom channels. whatever I do in custom channels I get invalid channels in scan list.

hashbang

added all the subnets to look for main table
4 subnets of lan and 2 subnets of isp1 and isp2
it resolved the problem

hashbang

I did it but of no use.

hashbang

i did the same but nothing happened. I'll do it again and report back(maybe I used src )
thanks

hashbang

Hi, I'm doing policy routing of two /24 subnets going to isp2 and two /24 going to isp1 throught default gateway. The only problem I experience is isp2 subnets cannot ping to isp1 subnets. They go directly to routing table of isp2 and leaves the route using default gateway. As far as my knowledge sa...

hashbang

is MT readinng this discussion. As I sent them support file and got the answer of upgrading to 6.19. You are corroborating that there are drops on 6.19 also.

hashbang

i changed the rule order as per said but on isp2 it was still advertising all subnets. I deleted all rules of isp2 and chain. Added all 3 rules again. Everything working as per needed. ty all. Next problem is policy routing

hashbang

no its not something to do with rule order. As discard all rule is 2nd so subnet 2 should be advertised.
I tried to change the rule order with discard all in the last but still same problem

hashbang

which version r u using ? I'll reboot later on. Queue size 500 is already too big for small bandwidths. Wrong screenshot posted earlier

hashbang

thanks, I'm not using any queue with hw queue. All the queues are default-small. Now changed to ethernet-default with pfifo queue size to 500 packets but packets are still dropping

hashbang

thanks,
This question i always wanted to ask. What is the match criteria of bgp filters.
all rules are read
rules are read till match is met.

hashbang

ty, here is my config 0 chain=isp1-out prefix=x.x.7.0/24 prefix-length=24-32 invert-match=no action=accept set-bgp-prepend-path="" 1 chain=isp2-out invert-match=no action=discard set-bgp-prepend-path="" 2 chain=isp2-out prefix=x.x.53.0/24 invert-match=no action=accept set-bgp-pre...

hashbang

hi,
I'm getting vlans packet drops on ccr 1036-8g version 6.15. I'd sent the support file. Had anyone experienced this kind of problem ? Scared to upgrade to 6.19

hashbang

hi, i have 2 isps and 4 /24 subnets lets take it as isp1 and isp2, networks as 1, 2, 3, 4. I have to advertise 1 and 2 on isp1 and 3 and 4 on isp2 My filter list is subnet 1 accept chain isp1out subnet 2 accept chain isp1out discard all chain isp1out subnet 3 accept chain isp2out subnet 4 accept cha...

hashbang

screenshot at around 800mbps few firewall rules, queues and vlans. PBR using route marking, 2 bgp peers. What maximum throughput can be expected on the basis of these inputs ?

Search found 202 matches