hello, Please guide me on how to configure my Pepwave MAX Transit (MAX-TST-DUO-LTEA-W. firmware version: 8.4.1) for a hot fail over for the purpose of video conference and remote server ssh connection without any downtime interruption. I want to set my ISP1 (RADIUS) as my primary and ISP2 (PLDT) is my secondary standby hot fail over.
current setup
ISP1 = connection to WAN PORT (RADIUS)
ISP2 = connected to LAN PORT (reserved VLAN10) (PLDT)
What about these configs
WAN smoothing?
Forward Error Correction?
Receive buffer?
Policy?
Congestion Latency Level?
Packet Jitter buffer?
I am a systems Engineer and I always perform microsoft Remote Desktop connection, SSH connection and Teams Video Conference.
Hello.
Do you have another Peplink device for WAN Smoothing?
FEC, also, need another Peplink devide at your network/topology.
Hello, I do NOT have other peplink device , I only have Pepwave MAX Transit (MAX-TST-DUO-LTEA-W)
what will be your recommendation and suggestion ?
Just using a single Peplink device will not give you seamless failover.
To start… the health checks need N failures before the mark the system down, so if you require 3 failures and check every 5 seconds that is a minimum of 15 seconds to detect the failure of the main link.
Then, the data will now be sent from the IP address of the second WAN… That will reset every TCP connection. Therefore every SSH, and RDP sessions will all reset, web sessions and teams will recover fairly well, but you have asked for “without any downtime interruptions”.
For regular failover you would use the outbound policy and select “Priority” and put the 2 Wans in the priority desired.
For Seamless 100% WAN redundancy you need to use a SDWAN technology. This requires either SpeedFusion Cloud , or hosting your own FusionHub either at your business endpoing, or in a cloud VM provider (many of us use VULTR, but AWS, google, azure all work. ).
You then establish a Pepvpn/SpeefFusion tunnel using FEC and Wan smoothing. Wan smoothing sends packets via both WAN links… and uses the 1st one that arrives… Therefore it responds and corrects WAN link errors and outages in real time. Also the long lived TCP sessions for VPN, SSH, RDP etc never shift IP addresses, since the remote endpoint sees the cloud IP which doesn’t change.
You can search for tutorials on using Fusionhubs and Speedfusion Cloud, and how to use outbound policies to only send the critical protocols and traffic across the Speedfusion tunnel, where your regular web and streaming traffic can live with regular failover.
Hello Paul,
ok I am now using the speedfusion Cloud with outbound rule as the fusionhub as my priority rule.
Do you know what will be the best settings for it?
I use smoothing: Medium and FEC: Low.
You also need the second WAN to be priority1 as well, or the system isn’t sending packets to both at the same time,
Hi Paul, Is this correct?
what about policy?, dynamic weighted bonding is correct?
You will have to look for discussions of the two bonding weightings… It really depends on the speeds/load differences between the 2 WANs. Test them both with live traffic and real world latency.
So now you should test it by pulling out the 2 network cables and seeing that your requirements are satisfied.
