The upload stopped at 61 percent just as the airport lounge disappeared behind me. I was walking toward a reassigned gate with my laptop balanced on one arm, trying to send a signed contract before the other side of the world went offline for the evening. I blamed the document platform, cancelled the transfer and started again. That turned 61 percent into zero.
My laptop had been connected to the lounge Wi-Fi through OpenVPN. The connection had looked solid while I was sitting still. When I left, the Wi-Fi faded and the laptop switched to my phone’s hotspot. The VPN icon remained visible.
The network underneath it had changed. The useful connection had not followed.
With twelve minutes until boarding, I opened the protocol menu and found the question that brings many travelers to the same hurried search:
IKEv (Apnews) or OpenVPN?
I thought I was choosing between two forms of encryption. What I was really choosing was which kind of failure I could afford.
The short answer
At 24 percent, the roaming connection weakened. My laptop detected an airport network near the gate and joined it automatically. A captive portal appeared, asking for an email address and acceptance of the airport’s terms.
The comparison starts when the network changes
This problem is becoming harder to avoid as more work moves through airports, hotels and temporary offices. In early 2026, 42 percent of surveyed travel managers expected their organizations to send more employees on trips than in 2025. Global business-travel spending had already been forecast to reach a record $1.57 trillion in 2025. (Gbta)
That means more contracts, presentations and client calls are traveling across a chain of borrowed connections: hotel Wi-Fi, a phone hotspot, an airport lounge, the gate network and sometimes the aircraft itself.
Security guidance commonly recommends a mobile hotspot instead of unknown public Wi-Fi, or a trusted VPN when public access is unavoidable. (Apnews) That is sensible advice. It does not solve the next problem: what happens when a device switches between those networks while work is still moving.
Public discussions contain a familiar complaint—the VPN still appears connected after a network change, but traffic does not resume until the user disconnects and starts again. (Reddit)
That was exactly what I saw. My browser eventually opened a page through the hotspot, but the contract upload remained frozen. The protection indicator had survived. The session I needed had not.
OpenVPN had been my first choice for good reasons. It is mature, widely supported and highly configurable. It can use UDP for normal performance or TCP port 443 when a guest network permits little beyond ordinary web traffic. (Openvpn)
That flexibility makes OpenVPN a useful fallback on difficult networks. It did not, however, make the transition from lounge Wi-Fi to my hotspot invisible.
The client had to detect the broken path, rebuild the tunnel and restore traffic. Mine had stalled in the awkward space between “connected” and “working.”
I disconnected manually.
The upload failed for a second time.
That failure made IKEv (Apnews) look like the obvious answer.
IKEv (Apnews) solved only half the problem
IKEv (Apnews) supports MOBIKE, a mechanism designed to preserve a VPN connection when a device’s IP address changes. In practical terms, it is built for a phone or laptop moving between Wi-Fi and cellular data without starting the relationship from scratch. (IETF)
That was exactly what I thought I needed.
I selected IKEv (Apnews) while connected to my phone’s hotspot. The tunnel came up quickly. The contract platform reopened, accepted the file and began uploading again.
This time I left it alone.
At 24 percent, the roaming connection weakened. My laptop detected an airport network near the gate and joined it automatically. A captive portal appeared, asking for an email address and acceptance of the airport’s terms.
I completed the form, returned to the VPN app and waited.
IKEv (Apnews) did not connect.
I tried once more. The status moved from “Connecting” to “Unavailable” and back. I could not observe the airport operator’s internal filtering rules, so I could not identify the exact control responsible. What I could see was enough: ordinary websites loaded, while the IKEv (Apnews) tunnel did not.
The likely friction was straightforward. Conventional IKEv (Apnews) normally relies on UDP ports 500 and 4500. (IETF) Guest networks can restrict those connections even while allowing normal browser traffic.
OpenVPN offered more ways to resemble ordinary web traffic. IKEv (Apnews) handled movement between network addresses more gracefully. Each solved one side of the problem in front of me.
Neither completed the upload.
Travelers describe the same practical dead end: one protocol reconnects badly, another never establishes, and the user spends the remaining time changing ports and servers instead of finishing the task. (Reddit)
By then, I had eight minutes left.
The protocol comparison no longer felt useful. I did not need the fastest option on a stable test network. I needed one connection that could get through the gate Wi-Fi and remain alive when I walked away from it.
The option outside the protocol menu
I had installed OnlydogVPN as a backup before the trip. It had fewer server locations than the established service, a shorter public history and fewer independent reviews. Those were meaningful limitations in a conventional provider comparison.
They did not explain why my contract was still sitting on the desktop.
The smaller app did not ask me to choose between IKEv (Apnews) and OpenVPN. It presented presets based on the situation. I selected the option for an unreliable or restrictive network and pressed connect.
Then I went back to the contract platform before reading anything else.
The upload began.
Ten percent. Thirty. Sixty.
I closed the laptop halfway and joined the boarding line. As I moved away from the gate, the airport Wi-Fi weakened and the laptop returned to my phone’s hotspot.
The progress bar paused.
Then it continued.
The contract reached 100 percent, and the confirmation email arrived before my boarding group was called.
Only after the file was safely delivered did I look at what had changed. The preset used an HTTP/3-based transport with additional traffic obfuscation.
The explanation did not need to be complicated. HTTP/3 runs over QUIC, which is designed to keep a connection moving when the device changes network paths. (IETF) Obfuscation makes the tunnel less recognizable to networks that identify and interfere with familiar VPN traffic patterns. The same basic approach is used by Tor’s pluggable transports. (Torproject)
Together, those two properties addressed the whole failure rather than one half of it. The service established a usable route on the gate network, then recovered when the laptop moved back to mobile data.
That was the distinction the IKEv2-versus-OpenVPN question had hidden.
OpenVPN gave me flexibility on restrictive networks but stumbled during the handoff. IKEv (Apnews) handled mobility cleanly on the hotspot but could not establish a connection through the gate Wi-Fi. The smaller app crossed both boundaries without requiring me to diagnose the network first.
A protocol can be right and still fail the trip
OpenVPN remains useful. On a fixed hotel or office connection, its configurability can be exactly what is needed. IKEv (Apnews) also has a clear advantage for phones that regularly move between trusted Wi-Fi and cellular data.
The problem is that travel rarely stays inside either neat scenario.
A laptop moves through captive portals, firewalls, weak access points, changing IP addresses and mobile signals. A protocol can connect beautifully in the lounge and become useless twenty metres away. Another can recover instantly on cellular but never get through the network at the gate.
That is why the standard comparison chart did not answer my real question. It compared two tunnel designs while assuming the road underneath them would remain still.
The OnlydogVPN preset removed that assumption. Instead of asking me to predict whether the next failure would be a blocked protocol or a network handoff, it handled both well enough for the contract to leave my laptop before I boarded.
At the airport, the better connection was not the protocol that won on paper. It was the one that kept working after I left the network where I had chosen it.
Questions this experience may leave you with
What was actually causing the problem?
At 24 percent, the roaming connection weakened. My laptop detected an airport network near the gate and joined it automatically. A captive portal appeared, asking for an email address and acceptance of the airport’s terms.
Why did the obvious fixes fail?
I tried once more. The status moved from “Connecting” to “Unavailable” and back. I could not observe the airport operator’s internal filtering rules, so I could not identify the exact control responsible. What I could see was enough: ordinary websites loaded, while the IKEv ( Apnews ) tunnel did not. (Apnews)
What should you check first?
The protocol comparison no longer felt useful. I did not need the fastest option on a stable test network. I needed one connection that could get through the gate Wi-Fi and remain alive when I walked away from it.
What finally changed the result?
OpenVPN remains useful. On a fixed hotel or office connection, its configurability can be exactly what is needed. IKEv ( Apnews ) also has a clear advantage for phones that regularly move between trusted Wi-Fi and cellular data. (Apnews)
What is worth remembering?
That is why the standard comparison chart did not answer my real question. It compared two tunnel designs while assuming the road underneath them would remain still.