6 ms·

this isnt even the first time its happened. Crowdstrike has killed an OS every month for the past four months. At this point they are a threat actor. if you

this isnt even the first time its happened. Crowdstrike has killed an OS every month for the past four months. At this point they are a threat actor. if you havent kicked their amateur-hour software out of your infrastructure by now, chances are good senior management and engineering have at least considered it formally. https://en.wikipedia.org/wiki/CrowdStrike#Severe_outage_incidents https://en.wikipedia.org/wiki/CrowdStrike#Severe_outage_inci...

metadat 2y ago
That incident list is damning. Is senior leadership asleep at the wheel, or how can this many incidents possibly happen every 30 days for months on end? If leadership really cared, they'd make sure post-mortems and other best practices are in place to reduce the frequency. Unfortunately, the executive disconnect isn't new. It's actually uncommon that they care about the reality for end users and customers (which is antithical to my entire ethos, hence why I get paid the medium bucks). Why bother waking up and going to work everyday unless you are contributing in some way to sustaining a better future for everyone? It's actually great for marketing and it's already going to be a tough 100+ years from today for our children, even with our collective care. P.s. People can be so selfish, it kind of breaks my brain but not really. Have you seen the CO2 emissions visualization from NASA this week? It was a wakeup call for me. 'Tremendous' NASA Video Shows CO2 Spewing from US into Earth's Atmosphere https://www.newsweek.com/nasa-video-carbon-dioxide-co2-emissions-1929622 https://www.newsweek.com/nasa-video-carbon-dioxide-co2-emiss... It's concerning.. and caught no traction.. http://news.ycombinator.com/item?id=41064029 http://news.ycombinator.com/item?id=41064029
- swasheck 2y ago
  here’s a fun connection: https://x.com/anshelsag/status/1814426186933776846 https://x.com/anshelsag/status/1814426186933776846 “ For those who don't remember, in 2010, McAfee had a colossal glitch with Windows XP that took down a good part of the internet. The man who was McAfee's CTO at that time is now the CEO of Crowdstrike. The McAfee incident cost the company so much they ended up selling to Intel.” so yeah, “leadership” (and that’s a loose term) doesn’t seem supremely concerned about much more than earnings
  - valicord 2y ago
    Not to worry, McAfee CTO was not actually in charge of technology https://archive.is/20240724213623/https://www.barrons.com/articles/crowdstrike-week-reckoning-stock-incident-ed00a543#selection-541.93-541.188:~:text=George%20was%20there%20as%20a%20sales%2Dfacing%20CTO%2C%20not%20in%20charge%20of%20engineering%2C%20technology%2C%20or%20operations https://archive.is/20240724213623/https://www.barrons.com/ar...
    - hinkley 2y ago
      The fish rots from the head. Also what the fuck is a sales-facing CTO??
      - cratermoon 2y ago
        I'm suspicious of CrowdStrike now. If we rip the cover off would we find that it's little more than a reskin of McAfee?
        Natsu 2y ago
        McAfee the company or the person? Because John McAfee was pretty out there... https://www.businessinsider.com/john-mcafee-tweet-said-his-suicide-would-him-being-whackd-2021-6?op=1 https://www.businessinsider.com/john-mcafee-tweet-said-his-s...
        cratermoon 2y ago
        McAfee antivirus the product
        graycat 2y ago
        The Internet is able to transmit odors of rotting flesh???? Recently ordered an HP laptop for some light work (not my startup), and when placing the order said don't include McAfee, that "I don't trust them", all just from some odor! CloudStrike runs in kernel mode? No wonder there are problems; kernel mode sounds like more of a threat than a protection. Sooooo, for my Web server(s), McAfee and CloudStrike are issues I get to ignore. Problems avoided and time, money, energy saved!! Simple.
- throwaway984393 2y ago
  [dead]
- Wytwwww 2y ago
  > Is senior leadership asleep at the wheel, or how can this many incidents possibly happen every 30 days for months on end? Presumably it doesn't matter that much and isn't worth spending money/manpower on? If the usefulness/quality of their software has no influence on their potential customers decision making process. why bother? It would make much more sense to allocate any excess resources to the departments that do actually matter like sales and marketing.
  - maerF0x0 2y ago
    > Presumably it doesn't matter that much and isn't worth spending money/manpower on? Well, if they think any of the $20B of shareholder value lost recently has to do with the quality issues... Then perhaps they should reconsider. (keep in mind marketcap also represents their ability to raise capital in the future with more/less dillution)
hinkley 2y ago
Staffing problems? Management often sees, “I have a dozen people on this.” When in fact the bus number was three, you laid one off, another quit and the third is sick or having life struggles.
- kermatt 2y ago
  "I have a dozen people on a dozen different things."
wannacboatmovie 2y ago
From your linked article: > A Hacker News user claimed that Nice to see Wikipedia has devolved even further into a dumpster fire in that they are now citing random HN posts as authoritative sources of facts.
- majewsky 2y ago
  Wikipedia is not an individual actor or a hivemind, so there is no capital-T "They". It's a system of multiple people each acting on their own accord. For a developing news story like this, I find this type of sourcing acceptable, especially because it is cited as "some person on the internet claims", not as "it is true that". If you disagree with this choice of source, you can flag this part as needing better sources. The simplest way to do so is to just leave a comment on the talk page.
surfingdino 2y ago
Never assume malice where incompetence will suffice. I have worked on teams where we could not get the basics like a test or integration environments signed off for months yet the managers expected us to go to production. Suffice to say production was also not signed off for half a yer and we had to improvise. I wonder is something similar was at play at CS?
- gregw2 2y ago
  Never assume incompetence when greed will suffice.
  - 7bit 2y ago
    It still is incompetence. That manager was incompetent.
    - fsflover 2y ago
      Or they did what they wanted in the name of profit.
jgalt212 2y ago
> this isnt even the first time its happened. Crowdstrike has killed an OS every month for the past four months. Yeah, but doesn't MS have to sign every kernel mode driver? They've allowed Crowdstrike's foot gun to continue to live in the kernel.
whiplash451 2y ago
Or maybe crowdstrike is dealing with the hardest threats and hence ends up having to rollout stuff rapidly against zero-days? Not a CS fanboy, but just wanted to suggest an alternative to sheer incompetence
- stoperaticless 2y ago
  Definitely incompetence. Maybe not personal, but as a collective - definitely. 1. They run inside a kernel. 2. They have many clients. (Incl. hospitals?) This implies serious severity and impact of potential incidents, so in turn it should imply certain level of quality assurance process. As we witnessed, it has glaring holes.
- ffgjgf1 2y ago
  How could the couple of minutes they saved by not testing their changes somehow made any difference?
- 7bit 2y ago
  I like the positivity that led to your suggestion. But I find it hard to follow the argumentation. They know that mistakes can take out thousands and thousands of devices, therefore it is imperative they prioritize stability over rollout speed. They have more direct access to devices, than any 0-day would ever have, therefore there is a significant risk that they do more damage with an update than any 0-day ever could. You have to remember, a 0-day could come to existence that threatens every system, but that usually happens once every couple of years (last one of that category was probably Blaster/I LOVE YOU). But CS risks of damaging the system every couple of hours with an update. Therefore, it should be tested to make absolutely sure it doesn't cause crashes. IMHO, it was sheer negligence and incompetence.

metadat 2y ago

That incident list is damning. Is senior leadership asleep at the wheel, or how can this many incidents possibly happen every 30 days for months on end? If leadership really cared, they'd make sure post-mortems and other best practices are in place to reduce the frequency. Unfortunately, the executive disconnect isn't new. It's actually uncommon that they care about the reality for end users and customers (which is antithical to my entire ethos, hence why I get paid the medium bucks). Why bother waking up and going to work everyday unless you are contributing in some way to sustaining a better future for everyone? It's actually great for marketing and it's already going to be a tough 100+ years from today for our children, even with our collective care. P.s. People can be so selfish, it kind of breaks my brain but not really. Have you seen the CO2 emissions visualization from NASA this week? It was a wakeup call for me. 'Tremendous' NASA Video Shows CO2 Spewing from US into Earth's Atmosphere https://www.newsweek.com/nasa-video-carbon-dioxide-co2-emissions-1929622 https://www.newsweek.com/nasa-video-carbon-dioxide-co2-emiss... It's concerning.. and caught no traction.. http://news.ycombinator.com/item?id=41064029 http://news.ycombinator.com/item?id=41064029

swasheck 2y ago
here’s a fun connection: https://x.com/anshelsag/status/1814426186933776846 https://x.com/anshelsag/status/1814426186933776846 “ For those who don't remember, in 2010, McAfee had a colossal glitch with Windows XP that took down a good part of the internet. The man who was McAfee's CTO at that time is now the CEO of Crowdstrike. The McAfee incident cost the company so much they ended up selling to Intel.” so yeah, “leadership” (and that’s a loose term) doesn’t seem supremely concerned about much more than earnings
- valicord 2y ago
  Not to worry, McAfee CTO was not actually in charge of technology https://archive.is/20240724213623/https://www.barrons.com/articles/crowdstrike-week-reckoning-stock-incident-ed00a543#selection-541.93-541.188:~:text=George%20was%20there%20as%20a%20sales%2Dfacing%20CTO%2C%20not%20in%20charge%20of%20engineering%2C%20technology%2C%20or%20operations https://archive.is/20240724213623/https://www.barrons.com/ar...
  - hinkley 2y ago
    The fish rots from the head. Also what the fuck is a sales-facing CTO??
    - cratermoon 2y ago
      I'm suspicious of CrowdStrike now. If we rip the cover off would we find that it's little more than a reskin of McAfee?
      - Natsu 2y ago
        McAfee the company or the person? Because John McAfee was pretty out there... https://www.businessinsider.com/john-mcafee-tweet-said-his-suicide-would-him-being-whackd-2021-6?op=1 https://www.businessinsider.com/john-mcafee-tweet-said-his-s...
        cratermoon 2y ago
        McAfee antivirus the product
      - graycat 2y ago
        The Internet is able to transmit odors of rotting flesh???? Recently ordered an HP laptop for some light work (not my startup), and when placing the order said don't include McAfee, that "I don't trust them", all just from some odor! CloudStrike runs in kernel mode? No wonder there are problems; kernel mode sounds like more of a threat than a protection. Sooooo, for my Web server(s), McAfee and CloudStrike are issues I get to ignore. Problems avoided and time, money, energy saved!! Simple.
throwaway984393 2y ago
[dead]
Wytwwww 2y ago
> Is senior leadership asleep at the wheel, or how can this many incidents possibly happen every 30 days for months on end? Presumably it doesn't matter that much and isn't worth spending money/manpower on? If the usefulness/quality of their software has no influence on their potential customers decision making process. why bother? It would make much more sense to allocate any excess resources to the departments that do actually matter like sales and marketing.
- maerF0x0 2y ago
  > Presumably it doesn't matter that much and isn't worth spending money/manpower on? Well, if they think any of the $20B of shareholder value lost recently has to do with the quality issues... Then perhaps they should reconsider. (keep in mind marketcap also represents their ability to raise capital in the future with more/less dillution)

hinkley 2y ago

Staffing problems? Management often sees, “I have a dozen people on this.” When in fact the bus number was three, you laid one off, another quit and the third is sick or having life struggles.

kermatt 2y ago
"I have a dozen people on a dozen different things."

wannacboatmovie 2y ago

From your linked article: > A Hacker News user claimed that Nice to see Wikipedia has devolved even further into a dumpster fire in that they are now citing random HN posts as authoritative sources of facts.

majewsky 2y ago
Wikipedia is not an individual actor or a hivemind, so there is no capital-T "They". It's a system of multiple people each acting on their own accord. For a developing news story like this, I find this type of sourcing acceptable, especially because it is cited as "some person on the internet claims", not as "it is true that". If you disagree with this choice of source, you can flag this part as needing better sources. The simplest way to do so is to just leave a comment on the talk page.

surfingdino 2y ago

Never assume malice where incompetence will suffice. I have worked on teams where we could not get the basics like a test or integration environments signed off for months yet the managers expected us to go to production. Suffice to say production was also not signed off for half a yer and we had to improvise. I wonder is something similar was at play at CS?

gregw2 2y ago
Never assume incompetence when greed will suffice.
- 7bit 2y ago
  It still is incompetence. That manager was incompetent.
  - fsflover 2y ago
    Or they did what they wanted in the name of profit.

jgalt212 2y ago

> this isnt even the first time its happened. Crowdstrike has killed an OS every month for the past four months. Yeah, but doesn't MS have to sign every kernel mode driver? They've allowed Crowdstrike's foot gun to continue to live in the kernel.

whiplash451 2y ago

Or maybe crowdstrike is dealing with the hardest threats and hence ends up having to rollout stuff rapidly against zero-days? Not a CS fanboy, but just wanted to suggest an alternative to sheer incompetence

stoperaticless 2y ago
Definitely incompetence. Maybe not personal, but as a collective - definitely. 1. They run inside a kernel. 2. They have many clients. (Incl. hospitals?) This implies serious severity and impact of potential incidents, so in turn it should imply certain level of quality assurance process. As we witnessed, it has glaring holes.
ffgjgf1 2y ago
How could the couple of minutes they saved by not testing their changes somehow made any difference?
7bit 2y ago
I like the positivity that led to your suggestion. But I find it hard to follow the argumentation. They know that mistakes can take out thousands and thousands of devices, therefore it is imperative they prioritize stability over rollout speed. They have more direct access to devices, than any 0-day would ever have, therefore there is a significant risk that they do more damage with an update than any 0-day ever could. You have to remember, a 0-day could come to existence that threatens every system, but that usually happens once every couple of years (last one of that category was probably Blaster/I LOVE YOU). But CS risks of damaging the system every couple of hours with an update. Therefore, it should be tested to make absolutely sure it doesn't cause crashes. IMHO, it was sheer negligence and incompetence.