Multiple jump-host problem
We're having trouble configuring multiple jump hosts in the program.
We've added two IP addresses—for two jump hosts—to the VPN/tunnel/gateway - SSH gateway entry.
(Currently, due to the aforementioned issues, we only have one entry)
It seems that the connection is established starting from the top of the list, and if the first jump host isn't available, the connection doesn't switch to the second jump host. 
I'm not sure, but the same thing happens if the “Detect reachable host in wait” option is selected.
The question is: did we configure something incorrectly, is this how it's supposed to work, or is it a software bug?
13545098-53f0-400c-b04a-d509aab7a571.png
66bf3da6-0864-4e88-9a7e-5a0efb5bd6a6.png
All Comments (2)
Hello,
Thank you for reaching out regarding this matter.
I am currently investigating this with the assistance of our development team. I will follow up with you as soon as I have additional information regarding the VPN/Tunnel/Gateway entry type with the SSH Gateway.
Thank you for your patience.
Best regards,
Jacob Lafrenière
Hello,
Thank you for your patience,
I wanted to follow up with a clarification on how the SSH Gateway entry works when multiple hosts are listed.
Multiple hosts = chained jump hosts, not failover. The list of hosts in an SSH Gateway entry maps directly to SSH's -J (ProxyJump) parameter:ssh -J user@jump1,user@jump2 user@target
This means RDM will connect to host 1 first, then tunnel through host 1 to reach host 2, and so on, each hop is a sequential step, not an alternative. If the first host isn't reachable, the chain stops there; RDM will not attempt the second host independently.
"Detect reachable host in wait" is a separate option that probes the final target to avoid waiting for the full VPN timeout, it doesn't affect which jump host is tried.
If your two IPs are two independent paths to your targets (i.e., two separate bastions, not two sequential hops), the correct setup would be:
- Create one SSH Gateway entry per jump host
- Create one session entry per target, each linked to the appropriate gateway
There is currently no built-in failover mechanism where RDM automatically tries a second jump host if the first is unavailable, this is not supported by the underlying SSH -J parameter either.
If this is a feature you'd find valuable, I'd encourage you to submit a feature request on our forum so our product team can evaluate it. If you'd like, I can also convert this thread into a feature request on your behalf, just let me know.
Best regards,
Jacob Lafrenière