Technical Details on Today's Outage
crowdstrike.com
crowdstrike.com
CrowdStrike Update: Windows Bluescreen and Boot Loops - https://news.ycombinator.com/item?id=41002195 - July 2024 (3590 comments)
Putting the actual blast radius aside, this whole thing seems a bit amateurish for a "security company" that pulls the contracts they do.
The company that lobbied the hardest and paid the most in bribes got the contracts.
This seems very amateurish for companies who regularly talk professionally to win said contracts , whether the best product or not.
My guess is C-suite, crisis consultants and lawyers are involved heavily so the actual engineering folks have little voice now in any communication and we get stuff like this.
You notice how this outage affected hospitals and airlines? There is a strong tendency in software sales for industries to align around one or two leaders. Oh, American chose Crowdstrike? Maybe we at Delta should just do what they did. Or literally Delta hires the VP from American to be their CISO and he just does what he did before.
Vendor selection is hard and buyer's remorse is frequently hard to deal with once you've sunk cost into a migration.
- They don't do enough or the right kind of smoke tests.
- They don't do exponential-canary deployments with an ability to rollback, and instead just YOLO it.
- They don't appear to have a customer-side security / client platform team update approval gating change control process for software updates or for definitions (or whatever they use).
This is fundamentally laziness and/or incompetency.
From management - that couldn't care less about low visibility, low impact projects
Everyone has a buggy release at some point, but impacting global customers at this level is damn near unforgivable.
Heads need to roll for this oversight.
And it sounds like they shipped some malformed channel file and the software that interprets it can't handle malformed inputs and ate shit. That software happened to be kernel mode, and also marked as boot-critical, so it if falls over, it causes a BSOD and inability to boot.
and it's kind of understandable that channel files might seem safe to update constantly without oversight, but that's just assuming that the file that interprets the channel file isn't a bunch of dogshit code.
On the scale of "no one bothered to put error handling or validation in" to "a subtle problem exists for this given input"; you and I lack the information to make a judgement.
Think about this a little harder: what do you know about the number of customers affected? We do actually have enough information to make a judgement - bricking millions of critical systems, a very high percentage of their total Windows customer population, tells us that they don’t have progressive rollouts, don’t fail into a safe mode, and that if they do have tests those tests are catastrophically unlike anything their customers run – all they had to do was launch an EC2 instance and see if it kept running.
I'm guessing they contain some combination of pattern/regexp type stuff, and interpreted code/scripting with trigger criteria, etc. that all gets loaded into the "engine" that actually runs the threat detection.
But no, saying "channel files aren't kernel code" is just hilarious, considering the channel files define how the actual kernel code is supposed to behave, so it might as well be kernel code. Especially when the bad behavior in question is triggered by bad channel files!!
If it controls the behavior of a computer, then it's code.
> and it's kind of understandable that channel files might seem safe to update constantly without oversight
Yeah, no, it's not. They pushed an update that crashed the majority of their Windows installed base in a way that couldn't be fixed remotely. It doesn't matter what the update was to. It needed to be tested. There is no way that any deployment pipeline that could fail to catch something that blatant could possibly be "understandable".
... and that kernel mode code shouldn't have been parsing anything with any complexity to begin with. And should have been tested into oblivion, and possibly formally verified.
This is amateur-hour nonsense. Which is what you expect from most of these "Enterprise Cyber Security(TM)" vendors.
... AND the users shouldn't have just gone and shoved that kind of thing into every critical path they could think of.
https://sre.google/workbook/canarying-releases/
Which starts with "a majority of incidents are triggered by binary or configuration pushes". The stats for config related failures is one link away at
https://sre.google/workbook/postmortem-analysis/
Where it says 31% of outages in 2010-2017 are caused by "configuration push".
https://x.com/perpetualmaniac/status/1814376668095754753?s=4...
https://x.com/ananayarora/status/1814269058088304760
The authors explain the coding error and coredump well, but I'm lost: Is the buggy code that they're describing the channel file, or some kernel code that consumes the channel file? Is there a way to tell?
As for detection, Zach mentions that modern tooling could've been used to find this, so I'm assuming Valgrind can find this: https://x.com/Perpetualmaniac/status/1814376690958868979
Hope this helps!
So if I put this all together:
a) The driver (sensor) csagent.sys includes code that hasn't checked with a tool like Valgrind or ASan or something and so includes some kind of memory management bug.
b) Since n, n-1 and n-2 versions of the sensor all died equally spectacularly, that bug as been around for at least three versions of csagent.sys.
c) The bug can be triggered by getting the csagent.sys to swallow a shitty channel file and since csagent runs in kernel mode, when it crashes it BSOD's the system.
d) Someone at Crowdstrike uploaded a shitty channel file as part of an update process that apparently happens many times a day.
Am I on the right track so far? If so, there's no/inadequate memory management checks in the csagent driver, and either:
1)There were also no checks before the borked channel file was uploaded because of a failure to follow process, or because there was no process, but whatever the case it was an accident.
or
2) Someone uploaded on purpose, not by accident, the borked channel file intending for a nasty outcome (probably not BSOD)
I can't believe that there are not a million checks and balances in place to let (1) happen, but as my grandma used to say, "Don't assume malice where stupidity will do" :-)
FWIW, at least Microsoft still "dogfoods" (and it's what coined that term), and even if the results of that aren't great, I'm sure they would've caught something of this severity... but then again, maybe not[1].
Push update to machines, observe, power cycle them, observe...
I could understand error in some rarer setup, but this was so common that it should have been obvious error.
I actually don't think it's outrageous that these files are rolled out globally, simultaneously. I'm guessing they're updated frequently and _should_ be largely benign.
What stands out to me is the fact that a bad config file can crash the system. No rollback mechanism. No safety checks. No safe, failure mode. Just BSOD.
Given the fix is simply deleting the broken file, it's astounding to me that the system's behavior is BSOD. To me, that's more damning that a bad "software update". These files seem to change often and frequently. Given they're critical path, they shouldn't have the ability to completely crash the system.
In the past, I've worked around this by validating the configuration of a file before attempting to run it. You bail out in a safe way during validation, but still allow a hard error during run time.
Doesn't prevent all misconfigured files, but prevents the stuff like.
I agree with the amateur hour observation. But then most things seem to be.
The real takeaway is that we need to design systems so this kind of issue is less possible. Put less code in the kernel, use tools that prevent these kinds of issues, design computers that can roll back the system if they crash.
Anyone competent that manages software at scale should generally hold the opposite opinion to this.
The lesson of "multiple parser implementations for the same thing seems bad" and "sanity checks to prevent breaking things are hard heuristics to define" such that further changes were deferred.
All that to say that I can appreciate circumstances in which satisfying "don't crash the system" in response to configuration data can actually be fairly hard to realize. It can very significantly depend on the design of the pieces in question. But I also agree that it's pretty damning.
I can see your perspective, but you should consider this: They protect these many companies, industries and even countries at such a global scale and you haven't even heard of them in the last 15 years of their operation until this one outage.
You can't take days testing gradual roll outs for this type of content, because that's how long customers are left unprotected by that content. Although the root cause is on the channel files, I feel like the driver that processes them should have been able to handle the "logic bug" in question so we'll find out more over time I guess.
For example, with windows defender which runs on virtually all windows systems, the signature updates on billions of devices are pushed immediately (with exception to enterprise systems, but even then there is usually not much testing on signature files themselves, if at all). As far as the devops process Crowdstrike uses to test the channel files, I think it's best to leave commentary on that to actual insiders but these updates happen several times a day sometimes and get pushed to every Crowdstrike customer.
If you can't take days to do it then do a gradual rollout in hours. It's not a high bar.
"Developers can use frameworks such as DriverKit and NetworkExtension to write USB and human interface drivers, endpoint security tools (like data loss prevention or other endpoint agents), and VPN and network tools, all without needing to write kexts. Third-party security agents should be used only if they take advantage of these APIs or have a robust road map to transition to them and away from kernel extensions."
Specifically the 2nd sentence above says security software should use the APIs, not Apple's kernel extensions.
I certainly don't want to know (through disaster news) about the construction company that built the bridge I drive through everyday, not for another 15 years, not ever!
This kind of software simply should not fail, with such a massive install base on so many sensitive industries. We're better than that, the software industry is starting to mature and there are simple and widely-known procedures that could have been used to prevent it.
I have no idea how CrowdStrike stock has only dropped 10% to the values of 2 months ago. Actually, if the financial troubles you get into are only these, take back what I said, software should be failing a lot (why spend money on robustness when you don't lose money on bugs?)
I actually expected their stock to drop a lot more than this, but goes to show you how valuable they are. investors know that any dip is only temporary because no one is getting rid of crowdstrike.
Think of the security landscape as early 90's new york city at night and crowdstrike as the big bulky guy with lots of guns who protects you for a fee, if he makes a mistakes and hurts you, you will be mad but in the end your need for protection does not suddenly go away and it was a one time mistake.
You're trying to hand-wave away the inexcusable. The outage is a symptom. The problem is the lack of even the most basic testing.
Clearly these files are sent out without even a minimal sanity check. That is a problem, and it's not something that can be hand-waved away.
This is inexcusable, point blank. “A couple of hours is a long time” is not a valid excuse when the alternative, as clearly evidenced, is millions of computers and critical systems simultaneously failing hard.
This might have been different if it was a small subset of computers, but this clearly could have been caught in minutes with any sort of sensible testing or canary rollout practices.
If problem encountered, don't send it out to everyone else.
They certainly run their software on those many customers' systems, but but based on my experience with them, "protect" isn't a descriptor I'm willing to grant them.
We don't have the counter-factual where Crowdstrike doesn't exist, but I'm not convinced that they've been a net economic or security benefit to the world over the span of their existence.
Sure, that applies to my company, but the counter-factual isn't "nothing is done and we keep getting pwned", the counter-factual is that instead of the resources spent on crowdstrike and their various problems (which have been regular since we adopted them, the recent mess was just the biggest), those resources are spent on improving security infrastructure without crowdstrike.
I don't understand CrowdStrike's rollout system, but given that people started seeing trouble earlier in the day, surely by that time they could have shut down the servers that were serving the updates, or something??
He also told me that soon after that the street outside the bank (another bank across the street, a hospital several blocks down) was lined with police who started barring entry to the buildings unless people had bank cards. By the time I woke up this morning technical people already knew basically what was going on, but I really underestimated how freaked out the average person must have been today.
The obvious joke here is CS runs the malicious C2 framework. So the system worked as designed: it prevented further execution and quarantined the affected machines.
But given they say that’s just a configuration file (then why the hell is it suffixed with .sys?), it’s actually plausible. A smart attacker could disguise themselves and use the same facilities as the CS. CS will try to block them and blocks itself in the process?
Given that this incident has now happened twice in the space of months (first on Linux, then on Windows), and that as stated in this very post the root cause analysis is not yet complete, I find that statement of “NO RISK” very hard to believe.
I’d like more information on how these Channel Files are created, tested, and deployed. What’s the minimum number of people that can do it? How fast can the process go?
> Although Channel Files end with the SYS extension, they are not kernel drivers.
OK, but I'm pretty sure usermode software can't cause a BSOD. Clearly something running in kernel mode ate shit and that brought the system down. Just because a channel file not in kernel mode ate shit doesn't mean your kernel mode software isn't culpable. This just seems like a sleezy dodge.
If you have the means to insert an AV config file update in between the config servers and the user's host then you probably can PWN the system pretty easily as well.
What this probably does mean is that Crowdstrike will be receiving some attention from hackers of both hat colors. Here's the bug bounty page ... https://hackerone.com/crowdstrike?type=team
See "It rather involved being on the other side of this airtight hatchway" https://devblogs.microsoft.com/oldnewthing/20200420-00/?p=10...
I can pwn my own desktop, yes, all I have to do is say "run as administrator". But the point of the security boundary is to make it impossible for software to get these privileges without me actively giving it to them.
If you're shifting the goalposts and imagining the computer does not belong to me, but to an organisation that I'm a mere employee of, they'll be using AD Group Policy to control what I can and can't do, and Bitlocker to encrypt the boot drive. I cannot boot into safe mode without having the tech support department give me a special code to unlock the computer. Again, that's how you get on the other side of the airtight hatch.
(Maybe there's some subtext that I'm missing, but I don't see how saying "these aren't kernel drivers" makes them look any better, and I do see why they might say it to be informative, so it looks like to me like they're doing the latter.)
It absolutely reads like this. They are getting blasted online for shipping kernel mode driver updates without proper QA and release engineering. Which just from face value just seems like some insano style engineering. They are saying "it's not actually a kernel mode value" to deflect blame.
I mean, I really don't understand why they would make this statement otherwise. If they are innocently just trying to say "this is just a channel file", there are other ways to say this, and it really isn't relevant enough to underline and emphasize.
I think the story they're telling now, which so far as I know is the truth, looks worse for them, because it requires them to have screwed up their QC twice. Once when they made a product that do such bad things, and once when they pushed the data file to millions of PCs without checking what it did.
So I still don't see how "this particular file happens not to be kernel-mode code" makes them look any better, and therefore I don't see why they'd be saying it "to deflect blame". It doesn't deflect blame; they look just as bad either way.
Make no mistake, this RCA was not published for technical folks. The only reason it’s even published is to make their customers feel more secure. You and I are not their customers; high level management and executives are.
That caught his eye, because to him it sounded like madness. Apparently deleting random driver files is a fairly well known way to screw a Windows system up even more than it already was.
This statement from CS must have gone through legal and PR review, so we have to assume every word and statement has been carefully vetted from a cover-your-backside perspective. It is light on information content, but there must be reason for them to so forcefully telegraph that the files being deployed (and removed) are not themselves drivers.
> We understand how this issue occurred and we are doing a thorough root cause analysis to determine how this logic flaw occurred.
There's always going to be flaws in the logic of the code, the trick is to not have single errors be so catastrophic.
How a common bug was rolled out globally with no controls, testing, or rollback strategy is the right question
And then absolutely the release process.
Rollback is hard I guess once your OS can't boot.
This is why the client needs have enough error handling to realise it's latest update has now caused unsuccessful boot and roll that update back locally to the last known good configuration (or completely back to factory and pull all updates again).
That's going to find a cause: a programmer made an error. That's not the root of the problem. The root of the problem is allowing such an error to be released (especially obvious because of its widespread impact).
Linux provides eBPF and macOS provides system extensions.
I'll also add that Windows itself heavily prioritizes backwards-compatibility over security, which leads companies to seek out third-party solutions for stopping malware instead of design-based mitigations being built into Windows.
And I'm not sure epbf actually allows you to do a lot of the stuff crowdstrike-like software does. I know they use it on Linux though so maybe eBPF has evolved a lot since I last looked at it.
Must be corrected to "the issue is not the result of or related to a cyberattack by external agents".
Very weak and over corporate level of ass covering. And it doesn't even come close to doing that.
They should just let the EM of the team involved provide a public detailed response that I'm sure is floating around internally. Just own the problem and address the questions rather than trying to play at politics, quite poorly.
https://www.nathanhandy.blog/images/blog/OSI%20Model%20in%20...
If the initial root cause analysis is correct, Crowdstrike has pushed out a bug that could have been easily stopped had software engineering best practices been followed: Unit Testing, Code Coverage, Integration Testing, Definition of Done.
If I ever get a sales pitch from these shit brains, they will get immediately shut down.
Also fuck MS and their awful operating system that then spawned this god awful product/company known as “CrowdStike Falcon”
It also wouldn't happen on Linux: they use eBPF there which was designed by grownups and validates its inputs.
https://developer.apple.com/support/kernel-extensions/
They’ve added system extension mechanisms for the most common needs trying to balance the various things people use kexts for against the impact on security, performance, and reliability many kexts had.
They’re just not using it. They could have not used it for Linux too. The presence of the feature is not enough to guarantee this would’ve never happened in a hypothetical.
Similarly, Microsoft clearly sees the benefits but note that they themselves say that’s not production ready yet. I’m certain that this incident will cause people to consider migrating as soon as that changes.
Let’s say Linux is the leading OS around the world. How can we be sure that they would actually use eBPF if this was the case?
They would likely choose the fastest option in order to support the platform as quickly as possible. Perhaps eBPF didn’t even exist if they prioritized Linux support and implemented that first, since Falcon was first released in 2013 and eBPF in 2014.
Switching from kernel mode to eBPF would be quite a lift, so if it wasn’t baked in from the start it likely wouldn’t have been added in after the fact.
A decade worth of changes is a lot to confidently say what would have happened. If Linux and MacOS were more popular than Windows, it could have been completely different.
This doesn’t even touch on the massive Debian incident CS had earlier this year, which is not a hypothetical.
Has this changed?
By the way, Falcon can be and is deployed to Linux and MacOS hosts in these organisations too it's just that this particular incident only affected Windows.
1. critical infrastructure around the globe seemed to depend on CrowdStrike
2. "If I ever get a sales pitch from..." suggested you are in an environment that is far from critical infrastructure.