Dual Bridge Bypass
When mapping your circuit to multiple bridges there are two scenarios to consider: backhauling traffic through your Exinda appliance and link aggregation through your Exinda appliance. The dual bridge bypass feature ensures that traffic is processed properly in these scenarios. Note that "dual" really means "multi" as traffic can pass through two or more bridges. There are separate dual bridge bypass settings for acceleration and for monitoring. Note that by default both acceleration and monitor for dual bridge bypass are enabled.
Backhauled traffic flows between distributed sites and the Internet via a centralized backbone, which is typically at headquarters. This means that traffic may go through your Exinda appliance at headquarters twice: from the source through the Exinda appliance, turning around at a router, back through the Exinda appliance, and on to the destination. This is problematic for accelerated traffic because you do not want to re-accelerate the traffic. The dual bridge bypass feature allows each bridge to treat traffic differently, so that the traffic is accelerated on one bridge on the way in and bypasses the acceleration handling from the second bridge on the way out.
Link aggregation maps multiple bridges to act as a single link. You may have a single router that is passing data through multiple bridges, where some packets may go over one bridge and some over the other. Or you may have multiple upstream service providers with asymmetric routes, where traffic may go out one bridge and return via another bridge. In this case, dual bridge bypass needs to be disabled so that traffic is handled the same regardless of which bridge it arrives on.
For acceleration, dual bridge bypass controls where packets are intercepted and passed to acceleration and edge cache.
- When enabled, all acceleration processing is performed on one bridge only. This is desirable for backhaul settings but will cause traffic problems if there is asymmetric routing or link aggregation in use. Consider traffic passing through the appliance from the WANWide Area Network to the Internet as: WAN <-> br10 <-> router <-> br20 <-> internet
- For an incoming accelerated connection, the acceleration processing happens on the first WAN interface to see the SYN. In this case, where the connection is going to the internet, the accelerated connection would be processed on br10. That same traffic when seen on br20 would then be passed though untouched (the traffic 'bypasses' br20).
- For an outgoing accelerated connection, the acceleration processing happens on the first bridge with a matching acceleration policy. In this case, where the connection is from the Internet, if there is an acceleration policy on br20, the acceleration policy happens on br20 and the traffic would be untouched on br10. Or if the acceleration policy was on br10, then the traffic would be passed through br20 untouched and accelerated on br10.
- When disabled, acceleration processing is performed on whatever bridge it arrives on.
This is desirable for asymmetric routing or link aggregation, where you want traffic to switch bridges. However, in an accelerated backhauled scenario, the accelerated traffic cannot switch bridges and still function properly. The SYN from the client must be processed on the same bridge as the SYN/ACK from the server.
For monitoring, dual bridge bypass controls how flows are tracked and how flows appear in the real-time monitor.
- When enabled, the same flowthe network traffic between network objects is tracked separately on each bridge. This allows you to see the different policies and transfer rates that are being applied on each bridge in the real-time monitor.
This may be desirable when you have dual bridge bypass for acceleration enabled so that you can see all the policies applied to the flows. If traffic is both accelerated and edge cached you see the two separate flows in real-time, but they will not be coloured yellow and blue to indicate acceleration and edge cache. They will both be coloured the one of the two colours.
If you see the (A) symbol in the real-time monitor and you get asymmetric route warnings, either the appliance is not seeing half of the flow or the other half of the flow is on another bridge, where the flow is bypassing acceleration processing on the second bridge and therefore you need to disable dual bridge bypass for acceleration for it to work properly.
If you have two circuits configured on the Exinda such that traffic between the headquarters appliance and the branch appliance goes through one circuit and the traffic between the headquarters appliance and the internet goes through a second circuit, then the traffic for a single flow will be counted in the appropriate virtual circuitlogical definitions that partition a a physical network circuit and used to determine what traffic passes through it and how much for each circuit.
- When disabled, the same flow through two different bridges will be shown as one merged flow. The real-time monitor will only show the policy of the last packet processed.
When doing load balancing, failover, link bonding, or seeing asymmetric routes (either locally or in a HA cluster), you will want to turn off the monitor dual bridge bypass function to get back to a merged view of all flows.
Modifying the Dual Bridge Bypass settings
To modify acceleration settings:
- Go to Configuration > System > Optimization > TCPTransmission Control Protocol.
- Modify the Dual-bridge bypass setting. If checked, it is enabled.
To modify monitoring settings:
- Go to Configuration > System > Setup > Monitoring.
- In the Monitoring Options section, modify the Dual-bridge bypass setting. If checked, it is enabled.