Measuring…
The verdict for this game appears once the first full second of packets is in.
Comparing with other tests to this server…
Measuring…
The verdict for this game appears once the first full second of packets is in.
Comparing with other tests to this server…
Are you sure you want to delete all results?
| Date & Time | Duration | Avg Ping | Avg Jitter | Packet Loss | Actions |
|---|
Host migration every other mission? Enemies teleporting while your own frame moves fine?
Warframe missions are not run on Digital Extremes' servers. One player in the squad is the host and their machine simulates the mission for everyone else, which means the connection that matters is the one between you and another player - and if you are the host, the connection that matters to three other people is yours. That single design decision explains most of what players find confusing about the game's networking, including why the same connection feels excellent in one squad and unplayable in the next.
What this test sends: 30 packets per second of 150 bytes, matching Warframe rather than a generic ping. estimated
Peer-hosted missions - the host machine sets the rate. Digital Extremes publishes no figure.
A note on the rate: because missions are peer-hosted, there is no single authoritative rate to simulate - the host machine sets it. This test uses 30 Hz as a representative figure and labels it an estimate. The loss and jitter it reports are properties of your line and hold regardless of who ends up hosting.
Run the test above for at least 60 seconds. Here is what each number means for this game specifically.
When the host's connection fails or they leave, the squad has to elect a new host and rebuild the mission state - that is the host migration screen. Loss on your own line causes it when you are the host, and causes you to be dropped from the squad when you are not. Red on this chart is the direct predictor of both.
Your own movement is predicted locally and feels smooth regardless. Everything else - enemies, other players, projectiles - comes from the host, so uneven delivery shows up as everyone except you stuttering. That asymmetry is the clearest sign that the problem is the link to the host rather than your machine.
There is no server to be near. Your effective latency is to whichever squadmate is hosting, which is why it changes mission to mission with no change on your end. The game gives you a lever for this in the matchmaking options, and it is the most useful setting in the menu for anyone with an unstable line.
What you see in game, and which number on this page explains it.
| What you see in Warframe | Look at | What it means |
|---|---|---|
| Host migration screen mid-mission | loss | The host's connection failed - yours, if you were hosting. |
| Enemies stutter, your own frame does not | jitter | Updates from the host arrive unevenly; your movement is predicted locally. |
| Ping varies wildly between missions | ping | You are measuring the distance to a different squadmate each time. |
| Frequent failed joins, never hosting | server | A restrictive NAT, not a line quality problem. |
In a client-server game there is one authoritative machine, it is professionally hosted, and everyone's experience depends on their own path to it. Warframe distributes that role instead: for the duration of a mission, one player's home connection is the authority for the squad. This keeps the game cheap to run at an enormous scale, and it makes the quality of a session depend on somebody else's router.
The immediate consequence is that latency is not a property of your setup at all. You can run two missions back to back on an identical line and have one feel instant and the next feel sluggish, because the host changed. Players routinely interpret this as the game being inconsistent or their connection being unstable, when in fact it is working exactly as designed.
The second consequence is the one that catches people out: when you are the host, your problems become collective. A short burst of loss on your line is not a hitch you absorb privately - it is a hitch for three other players, and if it goes on long enough it ends the mission for all of them in a host migration. Being a good host is a real responsibility in this game, and it is one you cannot discharge without measuring your line.
This is also why the diagnostic asymmetry is so useful here. Your own frame's movement is predicted locally and will look smooth even on a badly broken connection, while everything the host owns will stutter. If your Warframe glides and the entire rest of the world jerks, you have already localised the fault to the network path, and the only remaining question is which end of it - which is what the numbers on this page answer.
If the chart above shows packet loss or high jitter, work through these in order.
The Symptom: Some squads play fine, others are unplayable.
Warframe lets you cap the ping of squads you will be matched into. Setting it to a value you can actually play at trades a little matchmaking time for consistency, and it is far more effective than any local tuning because it removes the distant hosts entirely rather than trying to compensate for them.
The Symptom: Your squads suffer host migrations.
When you are the host, your line's loss becomes everyone else's problem, and a burst of wireless retries can end the mission for four people. A cable is the difference between a connection that is merely yours and one that is four players' shared infrastructure.
The Symptom: You rarely host and often fail to join.
Peer-to-peer matchmaking depends on both machines being reachable. A restrictive NAT prevents connections from being established at all, which shows up as failed joins and being matched only with a subset of players. Enabling UPnP or forwarding the game's ports is the fix, and it is unrelated to anything this chart measures.
The Symptom: The squad complains when your household is busy.
Hosting means uploading state to three other clients continuously. A saturated upload queue delays all of it, and the squad experiences your congestion as the mission stuttering. Router QoS that prioritises the machine keeps a hosted session healthy.
Because the player hosting the mission lost their connection or left, and the squad has to rebuild the session around a new host. If it happens in squads you host, your own line is the cause and this chart will show the loss behind it.
Missions are peer-hosted: one squad member's machine simulates the mission for everyone. That is why your latency changes between missions without anything changing on your end, and why there is no server location to choose.
Low enough to exclude hosts you cannot play against, high enough that matchmaking still finds squads. Run this test first: if your own line is unstable, tightening the limit will not help, because the problem is not the distance to the host.
Your own movement is predicted on your machine, so it stays smooth whatever the connection does. Enemies come from the host. When only they stutter, the link to the host is the problem, not your computer.