Thursday, April 9th 2026

Apple MacBooks Lose Network Connectivity After 49.7 Days Due to macOS "Time Bomb" Bug

Apple's MacBook Neo, MacBook Air, and the latest MacBook Neo series of laptops are now susceptible to a significant performance drop if they remain powered on for 49.7 days without a reboot. Specifically, after exactly 49 days, 17 hours, 2 minutes, and 47.296 seconds of continuous operation, the entire TCP/IP networking stack shuts down and stops establishing new connections. This unusual bug stems from a kernel-level issue in Apple's XNU kernel involving tcp_now, a 32-bit unsigned integer that tracks elapsed milliseconds since boot for the TCP stack. When it reaches its maximum value of 4,294,967,295, the counter stalls instead of resetting, disrupting the arithmetic macOS uses to expire and discard closed connections.

These inactive connection sessions accumulate quickly, increasing CPU usage as the system struggles to handle an ever-growing backlog. Once the available port pool, typically around 16,384 ports, is fully used, no new connections can be opened. Applications that rely on network access begin to fail, although already-established sessions continue to function, and the machine still responds to pings, albeit much more slowly. This initially delayed the diagnosis of the bug. A reboot resets the counter and restores normal operation, but the issue will recur on the same schedule after another 49.7 days of continuous use. There are potential solutions to fix the bug by using larger 64-bit integers, but this would require more extensive code changes, performance testing, and additional work compared to a simple fix.
The Photon team that discovered this issue notes that anyone managing a long-running MacBook should keep in mind the 49.7-day mandatory reboot until a fix is delivered. This is particularly important in enterprise environments, where managing machines is crucial for ensuring smooth operation without disrupting the workflow. The Photon team is also preparing a targeted workaround with the tcp_now that will address the connection freeze without requiring a restart, likely in the form of a live kernel patch or something similar. We need to wait and see how this will be implemented before the patch is delivered.
Sources: Photon, via Tom's Hardware
Add your own comment

28 Comments on Apple MacBooks Lose Network Connectivity After 49.7 Days Due to macOS "Time Bomb" Bug

#1
Vayra86
Bwhaahaha this is a 1995 level fail

Well done Apple, maybe start vibe coding like MS so you can make modern day fuck ups instead?
Posted on Reply
#2
Shihab
32bits to track time in 2026.
It's not like 99% of the industry already know why that is a bad idea...
Posted on Reply
#3
Vayra86
Shihab32bits to track time in 2026.
It's not like 99% of the industry already know why that is a bad idea...
Get this, in the SaaS application I work with, they had an integer value for customers' policy numbers. You can guess what happened. The first answer on the ticket 'But none of our customers have so many policies in the system'... totally forgetting a policy number doesn't simply count up, but is a combination of pre/suffix numbering depending on for example the conversion out of an old system.
Posted on Reply
#4
HLB
Coincidentally, this is the max number of credits you can have in No Man's Sky 🤣 You'd think Apple would know better, eh?
Posted on Reply
#5
igormp
I wonder how old this bug is, because I've had issues with wifi suddenly dying and not working at all (not even being able to scan networks) until a reboot a couple years ago, but I never had that again after I went from a M2 to my current M3, so go figure (or maybe I've been rebooting earlier).
Posted on Reply
#6
Nostras
You know, a 32-bit counter isn't that weird. The counter not resetting is surprising however.
Posted on Reply
#7
windwhirl
Shihab32bits to track time in 2026.
It's not like 99% of the industry already know why that is a bad idea...
True, but it seems it's only affecting the network stack. Most likely whoever wrote the code never thought anyone would have a machine online for 50 days non-stop, specially since Xserve has been dead and buried for well over a decade now
Posted on Reply
#8
MTS-C
igormpI wonder how old this bug is, because I've had issues with wifi suddenly dying and not working at all (not even being able to scan networks) until a reboot a couple years ago, but I never had that again after I went from a M2 to my current M3, so go figure (or maybe I've been rebooting earlier).
I'd say pretty old, but I haven't had this issue with my Macbook either.
Posted on Reply
#9
igormp
windwhirlTrue, but it seems it's only affecting the network stack. Most likely whoever wrote the code never thought anyone would have a machine online for 50 days non-stop, specially since Xserve has been dead and buried for well over a decade now
I usually have uptimes in the 30~90 days range, and so do most of my friends and coworkers.
Posted on Reply
#10
TheinsanegamerN
igormpI usually have uptimes in the 30~90 days range, and so do most of my friends and coworkers.
Going 3 months without installing security updates is generally not a good idea.
Posted on Reply
#11
igormp
TheinsanegamerNGoing 3 months without installing security updates is generally not a good idea.
Not all security updates are day-0 issues, nor are they often a problem since my devices are not fully exposed to the internet (unlike a proper server, where said updates are way more relevant). Ofc, that's not to say that you should never update, I reboot exactly to apply updates, otherwise I wouldn't even care about rebooting at all.
Keeping my browser up to date is way more relevant, security-wise.
Posted on Reply
#12
Grimling
Anyone remembers this year 2000 Bug?
This is the apple version of it.
Posted on Reply
#13
_roman_
i see it not that worse. when you choose to big variables for every feature your code gets too big.

not rebooting or using cold boots in 50 days for a portable device?
samsung tablet fake android even annoyed me with the pop up message when i intend to let it automatically reboot? at 2am was the default.
windows will quite often crash and therefore most often never hit 50 days. or will annoy you with forced reboot because of a update.

a small issue compared to afaik rockchip having a 30th february in their date chip of their processors.
Posted on Reply
#14
FatherOBlivion
Interesting little bug. Haven't noticed it personally. I wonder if it's only the latest Neo/Air models impacted? I only ever reboot for system updates (M2 Air personal, M4 Air work). Currently sitting at 33 days uptime. Assuming no security updates are released in the next 16 days, I'll be hitting that threshold and will be curious if impacted.
Posted on Reply
#15
Visible Noise
FatherOBlivionInteresting little bug. Haven't noticed it personally. I wonder if it's only the latest Neo/Air models impacted? I only ever reboot for system updates (M2 Air personal, M4 Air work). Currently sitting at 33 days uptime. Assuming no security updates are released in the next 16 days, I'll be hitting that threshold and will be curious if impacted.
macOS 26.4.1 was released today.
Posted on Reply
#16
FatherOBlivion
Visible NoisemacOS 26.4.1 was released today.
Ha! What timing. Just took a look at the release notes: "This update introduces 8 new emoji..." Looks like I have no choice but to update. Oh well, won't be hitting that 49.7 day threshold this time.

(To be fair, there are security updates as well. Just the first thing mentioned in the release notes is "8 new emoji").
Posted on Reply
#17
lexluthermiester
Vayra86Bwhaahaha this is a 1995 level fail

Well done Apple, maybe start vibe coding like MS so you can make modern day fuck ups instead?
Let's be honest, Apple doesn't have issues very often. It's gonna happen, but nothing compared to the constant microsoft issues..

To be fair, it's minor glitch. It's a not a security issue, or stability problem. Save whatever needs to be saved, reboot and one is up and running fine.

While this is amusing, I think the Apple reputation for a rock solid MacOS experience is safe..
Posted on Reply
#18
Vayra86
lexluthermiesterLet's be honest, Apple doesn't have issues very often. It's gonna happen, but nothing compared to the constant microsoft issues..

To be fair, it's minor glitch. It's a not a security issue, or stability problem. Save whatever needs to be saved, reboot and one is up and running fine.

While this is amusing, I think the Apple reputation for a rock solid MacOS experience is safe..
Sure, Apple also barely updates anything. Might be related :D
Posted on Reply
#19
FatherOBlivion
Vayra86Sure, Apple also barely updates anything. Might be related :D
Multiple major architectural changes, the most recent from Intel/x86 to in-house designed ARM64 based, OS/apps rebuild, etc. So seamless end users barely noticed. Other than the (massive) performance improvements.

I can't think of a consumer tech company that performs more major updates than Apple.

How's that Windows for ARM coming along? After what, 15 years or so now? :roll:
Posted on Reply
#21
lexluthermiester
meski.ozAn epoch fail
Nice play on words! Accurate too!

Welcome to TPU!
Posted on Reply
#22
Vayra86
FatherOBlivionMultiple major architectural changes, the most recent from Intel/x86 to in-house designed ARM64 based, OS/apps rebuild, etc. So seamless end users barely noticed. Other than the (massive) performance improvements.

I can't think of a consumer tech company that performs more major updates than Apple.

How's that Windows for ARM coming along? After what, 15 years or so now? :roll:
I'll rephrase that, yeah. 'At least Apple has some common sense in what it updates and how'.

Windows on ARM? I think that already went the way of the Windows Phone tbh. But its not unlike MS to revive that zombie a half dozen times in the near future. I can't wait for them to reintroduce, say, 'Zune' for music, now called CoZune or some stupid shit because it can detect what song you want to hear by taking all your data. And then, again wonder why it won't fly.
Posted on Reply
#23
unwind-protect
I experienced this networking hang several times. Good to see it is now a reproducible issue.
Posted on Reply
#24
DaemonForce
TheinsanegamerNGoing 3 months without installing security updates is generally not a good idea.
This take can eat me and the squirrely Athlon 64 eMachines uptime.

Maybe don't listen to every single "possible" threat in the shadows when you're not even a target. If my Server Core and Windows 10 ProWD builds are fine, we're good.
We've gone from WITH-LOVE-FROM-AMERICA.txt appearing on desktop from Unity Hub mid-2022 and getting to CPUID downloads getting pwned for a day in April 2026.
There are definitely threat actors targeting many of us based on the interests we have that fall under wrongthink (we own our hardware) and it does appear to be getting worse.
At the end of the day, only stupid people get catastrophically pwned and the only real threats that make their way onto hardened systems are through supply chain disruptions.

It's funny that it took this long for an Apple bug of this nature to make itself known because there are Apple servers and I'm sure many of them are running from some Macbook.
All this thread tells me is that almost none of them make it to 50 days of uptime and either that's something these customers find okay or they actually use computers like normal.
Posted on Reply
#25
tpa-pr
This is quite the hilarious bug because macOS (besides the mobile OSes) is THE OS that doesn't get rebooted/shut down. A lot of Apple users just close their MacBook lid and let their device go to sleep! When I was in support the amount of devices I'd see with ridiculous uptime is staggering :)

Can anyone elaborate on who the "Photon" team are? I assume security researchers?
Posted on Reply
Add your own comment
Nov 22nd, 2026 17:16 CST change timezone

New Forum Posts

Popular Reviews

TPU on YouTube

Controversial News Posts