Spanning Tree Protocol

Hi,
will be added the STP and or RSTP to RB260GS switch ?
If so, what standard will be used for STP ? the IEEE 802.1D or a proprietary STP ?

I trust in you mikrotik! :slight_smile:

They have simply said ā€œsometime in the futureā€ā€¦ but that was a couple of years ago! It may never be added :frowning:

:sunglasses: And so, NO official response from mikrotik ?

Spanning Tree Protocol support in SwOS is not planned in the near future.

Unfortunately. STP, resp. RSTP is definitely important feature.
Can you generally estimate if RB250/260 hardware is capable to handle it or isn’t?

The hardware can probably handle it fine… but this feature is not longer planned for the near future (this means at least a few years away, maybe longer), see the post above.

(R)STP is not a protocol I would like to depend on for redundancy. Is there no way to implement OSPF?

SWOS is what it is. If it isn’t there, it won’t be added anytime soon. You should look at using the regular Miktorik router if you want ospf or stp.

True, that’s what I meant :slight_smile: Maybe TS can solve his ā€˜redundancy problems’ using OSPF on a router instead of STP on a switch.

Currently they are planning to adapot it.

If you’re wanting a 1G switch and STP, the only option is a RB750GL … as far as I know.

ā€œwhoā€ is planning to adopt ā€œwhatā€ ???

I’m sure you are not saying that Mikrotik is planning to put R/STP into SWOS, at least not anytime soon.

I’m wondering if there was any progress in adding some spanning-tree protocol (RSTP maybe) to the Mikrotik switching platform (RB260GS and CRS125 or CRS226).

Does anyone can share some new info?

Thanks,
Gustavo.

+1 stp is a must

at the market its very rare a switch under 8-10 gigabit ports to support stp or rstp, looks like there is no economically viable to implement strong features on 5 port switch, many features are only seen on 10+ ports switches.

not only on mikrotik, other brands are the same, looks like chipset limitation

It’s not a chip limitation, maybe the chip isn’t good at RSTP and thats why it’s not implemented.
Here is the feature list for the chip: (it’s a lot of goodies :slight_smile: ie QinQ)

AR8327/AR8327N Features

  • Supports 802.3az Power Management
  • The AR8327N chip includes the Hardware NAT (Network Address Translation) function
  • The AR8327 chip (without the ā€˜N’ designation) does not contain the Hardware NAT function
  • ACL Mask Rule from Layer1~4. Port No, DA, SA, Ethernet Type, VLAN, IP Protocol, IPv4/v6 Source/Destination Address, TCP/UDP Source/Destination port
  • 96 ACL Mask Rule for Pass/Drop, VLAN/Q.O.S./DSCP Mapping/Translation
  • User define ACL up to 48 bytes depth in Layer 4/3/2
  • Q.O.S mechanisms include Weight Round Robin, Strict, Hybrid Up Queue
  • Port Base VLAN & 4K 802.1Q VLAN Group
  • IVL & SVL
  • IGMP Snooping V1, V2 & V3. IPv6 MLD V1/V2 forwarded to CPU
  • Supports Light Hardware IGMP snooping v1/v2/v3, MLDv1/v2 and Smart Leave
  • Hardware Looping Detection
  • QinQ function for SVLAN & CVLAN Translation
  • IP Packet/PPPoE bypass to reduce CPU loading on Video packet
  • 16 PPPoE session support/PPP Session Header Removal/Addition
  • Scalable Ingress/Egress Bandwidth Control
  • 40 MIBs Counter/Port & Port Status.
  • 1M Bit Packet Buffer
  • Supports 9K Jumbo Frame
  • Port Mirror, 802.1X Security, Rapid Spanning Tree
  • Rule-based Bandwidth Control
  • Programmable Wake on LAN
  • Half Power Mode for Cable length less than 30m (for home installations)
  • Supports Internal/External Loopback
  • Supports Reduced AFE circuit
  • 2K MAC Table. Edit, Search, Add & Delete.
  • MAC Limit by Port/Chip/VLAN
  • Trunking Function
  • Supports Trunking and auto-failover
  • Power Saving on Cable no Link, short Cable & 10BASE-Te Idle
  • Supports 1K NAPT entries and 128 hardware based host routing (ARP) entries
  • Supports hardware-based IP source guard, ARP inspection, routing/L3 switching
  • Supports VLAN translation and mapping with 64 Translation entries

ok is not a chipset limitation,

maybe a economical reason? too expensive to implement on small switches to be viable?? who knows…

it may even be a cultural issue, many people do not see the purpose of a small switch with strong manageable functions, many only think on big switches, and many only think on no manageable small switches, is a small niche market i think have to evolve, to increase the demand of this devices

I think is a very good idea to have small cheap very manageable switches, in certain scenarios you save money and space on cabling, i think is a modular approach to networking

Are there plans for STP on RouterBoard models (like RB2011/3011/CCR1009/…) or CRS? I would never use a switch which does not support STP.

STP is already supported in ROS when using bridges. STP is currently not implemented on CRS, and it doesn’t look like MT has any real plans of implementing it despite a lot of desire for it from the forums.

Is there any reason why mikrotik refuses to implement STP in any switch products?
In my opinion, any switch missing STP is not really an option for use in production environments. It may be good for home use, not more, not less. And no, loop protect is not a good STP replacement. STP is a lot more than loop protect.