It took 48 hours, but the mystery of the mass Asus router outage is solved | Hacker News Reader

It took 48 hours, but the mystery of the mass Asus router outage is solved

arstechnica.com

71 pointsmfcc6422 comments

Did an llm write this? I can imagine the prompt. "write a snappy sounding article about how a tech mystery got solved"

In reality, nothing is solved.

> Why were routers affected even when they had been configured to not automatically update and no manual update had been performed? Asus has yet to address this,

Why indeed

From the article:

> “On the 16th, Asus pushed a corrupted definition file for ASD, a built-in security daemon present in a wide range of their routers,” one person wrote. “As routers automatically updated and fetched the corrupted definition file, they started running out of filesystem space and memory and crashing.”

I'd say that resolves most of the overall mystery. It doesn't explain your specific question, but it certainly explains more than "nothing".

---

I also think the answer to your specific question is relatively obvious, even if Asus didn't confirm it explicitly: Asus routers download new security definitions automatically, even when automatic updates are disabled. You're free to dislike that (I don't know how I feel about it), but it is clearly what's happening.

It's not like this is an uncommon practice. Windows Defender also downloads new malware definitions even when automatic Windows updates are disabled (which is hard to do, but possible on Pro editions of Windows via e.g. setting registry keys).

The malware signature db is presumably updated separately from the firmware image, since they're on completely different release cadences.

My ASUS router with the malware filter disabled wasn't affected, FWIW.

May as well run the normal parser on an update file to print its version info and maybe some metrics in the manual update UI?

If tech youtube is to be believed ASUS is quite the sinking ship at the moment, which is sad considering they have been such a cool company for many years.

They've been going down like this for a long while now.

I remember when toggling a VRAM option within their EFI BIOS caused it to brick after one of their firmware updates on many of their motherboards where their support gave you the run around or refused to honor the warranty/issue an RMA until you threatened legal action.

They also have failed to properly follow specifications for ACPI, and did translated IO while claiming to support PCIE Atomics (which are mutually exclusive options).

The market is so concentrated, and they are failing so hard, its just a matter of time before a new entity entering the market with a slightly less comparable board could be financially viable.

Lost their magic once they split Asus into three companies (to satisfy Apple from what I understand), thus became Asus, Pegatron and Unihan to remove conflict of interest for their ODM side of the business

I'd wager most people aren't thinking 13-15 years back when discussing the sudden decline of ASUS. They'd done a lot of cool stuff since then.

They have, I agree. Perhaps it is my rose tinted glasses. I suspected the split would lead to pressure to create more 'hit' products and eventually impact quality. It's hard to express how rock solid Asus motherboards were perceived back then.

Apple? How are they involved?

Asus was very successful and at one point had captured a large portion of the motherboard market due to their extremely reliable motherboards at that time. (I recall reading many years ago as many as 1 in 4 computers were using an Asus mobo).

Asus started to branch into making other equipment. Their notebooks were good quality and their premium notebooks were innovative and risked eating into markets of other manufacturers. Complaints were raised that Asus had a conflict of interest as they were an ODM making equipment for companies like Apple but then able to leverage that to assist their own business. Asus was then split to remove conflict of interests.

Meanwhile Samsung is totally free to make parts for Apple and directly compete in a hundred other categories…times have changed?

Maybe Samsung just has a stronger legal department.

I read somewhere else in this thread that Asus was actually making the entire MacBook plastic laptop which I guess is very different from just provider components.

Huh? Where can we read more?

"In 2007, Apple significantly dialed down its MacBook manufacturing business with Asus and the company had to respond. That same year, they announced that they would hive off their original design manufacturing business into a 100% owned subsidiary called Pegatron. Management hoped this would reassure their customers that improper information cannot be shared across lines.

But the conflicts of interest persisted. And even got worse when Asus released its hit Eee laptop product. The first so-called netbook, the low-end PC sold 5 million units at the start and presented a real threat to Pegatron's customers. They were eating in other companies' market share. "

https://asianometry.substack.com/p/asus-explained-a-rare-tai...

Article is a bit big -- I'll read it but in the meantime, what did Apple do exactly? They "dialed down its MacBook manufacturing business with Asus"? What does that mean? Asus would have been their hardware supplier, or?

Asus made products for Apple and others. An example from memory is the sleek plastic Apple macbook (white and black) from 2006 was made by Asus. At the same time, Asus were making laptops under their own brand which showed a lot of potential e.g Asus W7J which had magnesium body.

The idea was Asus could use knowledge they obtained making products for others to make their own products. Not only that, the Asus brand was becoming more known, the risk was the consumer might buy Asus products (they made quality products for those other big brands right? Why not go direct).

Apple was rumored to pull back from using Asus as a manufacturing partner for this reason.

Oh I see. So Apple wanted ASUS to only produce laptops for them.

Nasty but also not surprising, sadly.

I was expecting it to be something more interesting given that it took 2 days

It raised a lot of interesting questions in the original thread about automatic updates and other issues that will plague us for decades to come.

And for all the talk of OpenWRT, which is a great addition to the wireless ecosystem, I'm still concerned about the staggering burden maintaining all those ports must be, and how closely the source trees are reviewed.

Just because it's open source doesn't mean it's a panacea.

Did this affect routers with the Merlin version of installed? I don't run stock asus on my routers and had no such incident.

Reply on news.ycombinator.com