PCC and Hotspot isn't possible on the same machine.

Arman

PCC and Hotspot isn't possible on the same machine.

Arman

PCC and Hotspot isn't possible on the same machine.

Arman

PCC and Hotspot are very possible on the same router. I have it running successfully at around 50 locations for well over a year.

Here is one example of a working setup.
http://forum.mikrotik.com/viewtopic.php ... 25#p175525

The problems you may run into is if you are doing proxy or transparent proxy on the same box along with the Hotspot and PCC, but even this is still possible. I successfully got that working in a lab environment.

The main difference between that configuration and the one in the Wiki is that is specifies "hotspot=auth" on the PCC rules. That's what is doing most of the work in that case. You can also accomplish the same thing with a NAT rule.

Code: Select all

/ip firewall nat
add action=accept chain=pre-hotspot hotspot=auth dst-address-type=!local

The reason why PCC does not work with the hotspot "out of the box" is because of the way the hotspot functions. Behind the scenes it redirects all traffic to itself, and this breaks the PCC configuration due to traffic now being on a different firewall chain than expected. The NAT rule above stops the hotspot from processing packets from authorized users, so things like PCC work normally. This rule however does break things like the advertising functions available from the MikroTik, and no load balancing will happen for people before they are authorized on the network.

You can create that firewall rule before or after, just the custom pre-hotspot chain will not exist until the hotspot has been created. By using the pre-hotspot firewall chain, you are basically applying a NAT rule before the hotspot does any processing on the packet.

The packet flow diagram shows you how packets are processed through the MikroTik. It is a very good document to have as it shows you how a packet flows through the router so you know where any why certain things work the way they do.
http://wiki.mikrotik.com/wiki/Packet_Flow#Diagram

/ip address
add address=192.168.2.2/24 disabled=no interface=WAN1
add address=192.168.1.200/24 disabled=no interface=WAN2
add address=10.0.0.2/24 disabled=no interface=hotspot

/ip firewall address-list
add address=10.0.0.0/24 disabled=no list=Local_NAT_Networks

/ip firewall mangle
add action=mark-connection chain=input connection-state=new disabled=no in-interface=WAN1 new-connection-mark=WAN1_connection passthrough=yes
add action=mark-connection chain=input connection-state=new disabled=no in-interface=WAN2 new-connection-mark=WAN2_connection passthrough=yes
add action=mark-routing chain=output connection-mark=WAN1_connection disabled=no new-routing-mark=to_WAN1 passthrough=yes
add action=mark-routing chain=output connection-mark=WAN2_connection disabled=no new-routing-mark=to_WAN2 passthrough=yes
add action=accept chain=prerouting disabled=no dst-address=192.168.2.0/24 src-address-list=Local_NAT_Networks
add action=accept chain=prerouting disabled=no dst-address=192.168.1.0/24 src-address-list=Local_NAT_Networks
add action=mark-connection chain=prerouting connection-state=new disabled=no dst-address-type=!local hotspot=auth new-connection-mark=WAN1_connection passthrough=yes per-connection-classifier=src-address:2/0 src-address-list=Local_NAT_Networks
add action=mark-connection chain=prerouting connection-state=new disabled=no dst-address-type=!local hotspot=auth new-connection-mark=WAN2_connection passthrough=yes per-connection-classifier=src-address:2/1 src-address-list=Local_NAT_Networks
add action=mark-routing chain=prerouting connection-mark=WAN1_connection disabled=no new-routing-mark=to_WAN1 passthrough=yes src-address-list=Local_NAT_Networks
add action=mark-routing chain=prerouting connection-mark=WAN2_connection disabled=no new-routing-mark=to_WAN2 passthrough=yes src-address-list=Local_NAT_Networks

/ip route
add check-gateway=ping disabled=no distance=1 dst-address=0.0.0.0/0 gateway=192.168.2.1 routing-mark=to_WAN1 scope=30 target-scope=10
add check-gateway=ping disabled=no distance=1 dst-address=0.0.0.0/0 gateway=192.168.1.254 routing-mark=to_WAN2 scope=30 target-scope=10
add check-gateway=ping disabled=no distance=5 dst-address=0.0.0.0/0 gateway=192.168.2.1 scope=30 target-scope=10
add check-gateway=ping disabled=no distance=10 dst-address=0.0.0.0/0 gateway=192.168.1.254 scope=30 target-scope=10

What line specifically are you talking about not working? The mangle rules look right and they paste fine into one of my routers.

If you are talking about only one line being used, it's likely because of your PCC classifier. It it only taking it based off of the src-address, so this means that every connection coming from the same host will always get the same connection mark, and therefore always go out the same link. We use both-addresses for the classifier. It's random enough where both links will be used fairly evenly, but stable enough that it won't break forum web sites or banking web sites, as all connections from these same host to the same destination will always use the same link.

The classifier takes the paramiters you tell it to (src, dst address or both, src, dst port or both, or all of them together) and does a hash function on it and divides it by a given amount. Based on that value returned it gives it a different connection mark that you specify. Obviously the more variables you feed it, the more random it will be and the more evenly the connections will go up, but if it's too random you run the risk of breaking certain web sites like banking or forms.

The PCC setup was designed as and is a load balancing method, that's why you use it and that is its advantage. No one connection can exceed it's maximum throughput, but because it load balances you have a good chance of having your 2nd or 3rd or 4th connection go out of the line that is free. Because multiple connections will be balanced across multiple internet connections, you in effect have more bandwidth available to your network. You can also set it up to do failover to increase network up time and reliability. If you wanted to "combine" your connections and make them act as one, then you are looking at a bonding solution, not a load balancing one.

Bonding is rarely a viable option for most people to increase bandwidth. Usually to enact a bonding setup you need to own and control both ends of the bonded connection (in most situations this will be your local site and a remote site with a very high bandwidth connection). If both of your connections are from the same provider you can contact them and ask them if they support bonding, if they would be willing to set it up for you, and how much they would charge you for this, but this is not likely in most situations. I know of one device that claims to be able to dynamically bond TCP across multiple connections, but that hardware is very expensive.

That is most likely because they are using a download manager that opens multiple sessions to download one file. In order to get the maximum effect in that case you'll want to use both addresses and ports as the PCC matcher and a 3rd party program.

Plz any body take my code to try and tell me the result and take Screen Capture of the download