iPhone 11 Pro running iOS 13.3.
I got dinged last week for exceeding my monthly data usage on my corporate plan, so I did a reset on my cellular usage on 12/13 and this morning checked it. In 3 days, I had used 1.9GB of cellular data and 1GB of that was 1.1.1.1. I switched cellular usage for 1.1.1.1 off, but that defeats the purpose of having it, shielding my DNS from my provider (Verizon in my case.)
Is this even remotely normal? At 1GB every 3 days, that sucks down 10GB every month just for 1.1.1.1, which is not tenable even if everything did go through it and only it. I work a lot from home and have my iPhone set to default to my home network which can take 10GB of traffic from my phone in a month, but not if I’m working on the road.
Are you referring to the DoT service or the Warp+ VPN service? The latter could easily rack up such an amount of traffic (of course that depends on your traffic behaviour).
Warp+ VPN service. Yes, I can see that happening, the amount not recorded going to the app is very close to what did, but that makes me wonder why it’s going out twice? The service is advertised as designed for mobile devices, which says to me that by utilizing it will double my monthly bandwidth and associated costs, it’s not truly ready for a mobile device.
My VPN on my laptop for work uses a little bit more than an unencrypted connection, but definitely not double. Honestly, if it’s working as intended, I can’t use it.
It’s very, very unlikely that the traffic resulted from Warp itself. It’s far more likely that something else on your phone used significantly more data this month, or perhaps was incompatible with Warp+ and failed in such a way that caused it to send exorbitant amounts of data. (In theory, this could happen if a service blocks requests from Warp+, and an app makes repeated requests to that service, retrying immediately and indefinitely when those requests fail.)
“The amount not recorded”? Not sure what you mean by that?
As you are referring to “twice”, do you mean you actually did use 500 MB in these three days?
Also, where did you take the traffic figures from? From your phone or from your provider? If it was the former, there is some chance that might be skewed due to the VPN nature.
Sorry if I wasn’t clear, this is a capture from my iPhone this morning, the Cellular Data usage. The current period started on 12/13/19 running through today, 12/17, so 4 days.
From 12/13-16, three days, 1.9GB of cellular data was used with 1.1.1.1 using 1GB, everything else used .9GB. That’s my reasoning behind 1.1.1.1 doubling the data utilized. The top 8 applications, including System Services, is 883 MB, just a little bit away from .9GB or 921 MB. The difference between when I had 1.1.1.1 enabled for Cellular and today is .1GB or 104 MB, less than one-sixth of what had been consumed in 3 days (104 vs 633 MB).
Since VPN happens at the data link or network layers (2 & 3 depending on if it’s Linux or Winblowz) and data comes in at the presentation or application layer (6 & 7), nothing should leave the device without going through VPN first. That’s the whole point, nothing leaves the machine without going through the VPN first. I can easily see .9GB becoming 1GB passing through the VPN, but not simply doubling.
I am far more familiar with the OSI model for open systems and I don’t know how they translate to mobile, but iOS and Android both are built around a very modified a linux kernal, so there shouldn’t be too much difference.
I am not quite sure to be honest and I only vaguely remember having read something related here on the forum, but could that be an issue of how iOS counts traffic?
Assuming you had the VPN on, all data should have gone through Cloudflare and hence the other applications should be at zero. Considering they are not I am wondering if the listed data simply is what went from the application to the VPN application (internally) and the one gigabyte is the total the VPN application actually sent as traffic over the wire.
Hence my earlier question whether you can check this against your provider’s statistics.
iOS is Unix-like, but it’s BSD-based, not Linux-based, and its networking is pretty different even from other BSDs. I doubt that’s the issue here, though.
I believe the cellular data usage you’re seeing isn’t double-counting usage: that is, if data is passing through the VPN, it only shows under 1.1.1.1, whereas if it’s not passing through the VPN, it shows under the app from which the data is actually originating. And it’s almost certainly not showing Warp overhead separate from traffic used by other apps that passes through Warp–it doesn’t really have a great way to calculate that reliably.
If I’m right, that’d mean half your traffic passed through Warp and half didn’t. The 1.0 GiB you’re seeing for Warp actually came from other apps, and the traffic you see directly associated with other apps never passed through Warp.
After running the fast.com test a second time, just to be sure:
Safari: 309 MB
1.1.1.1: 213 MB
iOS appears to be double-counting bandwidth: that is, if you use 100 MB while on Warp, it’s going to attribute 100 MB to whatever app initiated the traffic as well as 100 MB to Warp. However, assuming your carrier isn’t doing something scummy, they should only bill you for 100 MB.
This is just a theory, though. To be certain, I need to check my data usage. As of about an hour ago, T-Mobile was reporting 988.7 MB–that’s over a period of about 3 weeks. I’ve run the test a bunch more times, with Safari now at 449 MB and 1.1.1.1 at 371 MB. I’d expect to see my billable usage jump by about 360 MB if iOS is double-counting or by about 700 MB if it’s not (or if T-Mobile is doing something scummy). I’m currently near a T-Mobile store; if my usage doesn’t update automatically within the next few hours, I’ll head over there and ask them. I’m also near an Apple store, so I can ask there for clarification about usage.
To avoid messing with the metrics too much, I’ve switched to Wi-Fi and will try to remain on it until my billable usage updates.
Edit: After switching to Wi-Fi, Safari is now at 510 MB and 1.1.1.1 is at 457 MB. I suspect fast.com continued running in the background and some other applications made some requests through 1.1.1.1, increasing both numbers and closing the gap between them.
That’s interesting… I just did the same, resetting the usage and FAST pumped out 48 MB while 1.1.1.1 only used 26 MB. BUT with 1.1.1.1 enabled, it clocked at 14Mb/s while without it was 3.8Mb/s. It’s difficult to get exact figures since I’m on a corporate plan and we only find out we’re over after it’s all done, but #3282 (Verizon) shows I’m at 2GB for the period ending on 1/11.
Apple could very well be double-dipping, showing all application cellular usage even if it doesn’t end up going directly to cellular.
sigh
Now if only they enabled the same application data logging for WiFi only iPads that they do for Cellular. Again, logging is tied to layer 6, well above where any cellular or wifi antenna is, they should be able to do it but don’t.
That is my understanding from aforementioned vague memory
But if your provider statistics show a different picture then it might be worth looking into it. You said you only find out whether you exceeded your data allowance at the end of the billing period, yet say you are already at 2 gigabytes. How come?
I’ve got 6.7 GB of cellular data usage so far, since the 29th of November… normally, I only have about 500mb per month. I’m seeing an issue too with the iOS app…
@wsuschmitt, the current theory is that the traffic isn’t actually from Warp, but rather other apps on your device; when Warp is enabled, iOS reports the data usage twice: once for the app initiating the traffic, and once for Warp, since that’s the VPN. In that case, Warp isn’t actually responsible for the increase in data usage.
You mentioned that your normal usage is 500 MB/mo. However, for your current billing period, Facebook alone is already showing almost twice that. Oddly, the data usage for 1.1.1.1, 6.7 GB, exceeds the usage reported by T-Mobile, which is 4 GB; something doesn’t add up there.
Another consideration: Every byte that enters the 1.1.1.1 app leaves it too, depending on how iOS is accounting traffic it could be including data in both directions.
I haven’t tested this, but it is a definite possibility to explain why the app consumes more data than the carrier can account for.
One final thought, if a carrier offers any zero-rated content (e.g. their own streaming service), Warp would effectively bypass this and you would now be billed for said bytes. Similarly, any carrier silliness like per-site throttling to try and force streaming video sites to 480p or lower will also fail to identify such traffic that passes through Warp.
@Zenexer: you caught on to why this is a mess: 6.7 GB of “usage” by 1.1.1.1 when I said it only normally uses about 500 MB/mo… This is new behavior. In the past, I’ve seen about 500MB/mo on 1.1.1.1 when my total is about 4 GB on cellular.
My total cellular usage of all my other apps don’t even come close to the 6.7 GB of data usage by 1.1.1.1… it suddenly spiked and this is not normal behavior. About 500MB/mo is normal behavior. I was searching the web for other’s issues with data usage. Under normal circumstances, I’ve got a wifi signal to pull from while at work or home… I need only about 4 GB/mo for cellular data.
This month is out of norm for me, and 1.1.1.1 is the reason, this time. Hasn’t been in the past…
I’m starting to get the impression that iOS is to blame here. I don’t understand how it can show 6.7 GB for a single app but 4 GB total unless iOS is doing something wonky.
I’m not sure I actually trust those metrics. iOS is reporting 68% more for a single app than T-Mobile is reporting for its total. I’ve never known a carrier to under-meter–usually it’s the opposite.
That is the point and from this perspective it is logical the operating system displays the traffic for the originating application as well as for the VPN application, however the overall traffic still didnt double, simply because half of it was only internally routed.
Again, only Facebook used about twice that much. So either you exceeded your normal behaviour this month or Facebook transferred much more.
I am not not really an iOS user, but based on the thread I referred to earlier, this was my assumption all along.
It still would be important that the OP checks it against his provider’s statistics but I’d be surprised if they showed that traffic too.
When I was testing, this didn’t seem to be the case. It was counting once for each app (e.g., once for Safari + once for 1.1.1.1, not once for Safari + twice for 1.1.1.1).
This is possible, but I’m not aware of any that do that for Warp. After using about 500 MB on Warp, though, my T-Mobile reported usage went up about 10 MB. Either the updates are delayed, or you’re right about that.
Judging by the second post, I think @wsuschmitt is saying 500 MB/mo for 1.1.1.1, not 500 MB/mo total.
Well, double makes sense–that I could see iOS doing, since it’s just counting once for each app. That 6.7 GB metric doesn’t make any sense unless it’s counting both ingress and egress for 1.1.1.1, but in my tests that’s not what it did; it only counted egress.
