It's not limited to Google, now I wonder if this is actually a Cloudflare/other CDN/International Infra issue or is it that so many services use Google Cloud stuff and they are gone too because of it.
I don't, for that I usually rely on search engine or autocorrect. I hate the fact that it is not pronounced as it is written. Because I always go with "Wenesday" on my own.
> ℹ Note: Multiple online services including Google Search and YouTube are currently experiencing international outages; incident not related to country-level internet disruptions or filtering #GoogleDown
Currently working fine for me in Pakistan. Btw, you can run https://api.webprobe.org/scan/ in a spare tab in the background and it will periodically check the status of such websites and update the NetBlocks status.
Can confirm, from Yerevan - Armenia.
Internet slow, then slows down to a halt in a matter of 10 seconds. Then it'll work in a burtst of another 10 seconds, and go back to nothing at all, and stay like that for minutes at a time.
Internet on a National level is fast. Any connection to the outside seems to go through a bottleneck.
the initial handshake of the connection in the affected countries is going through all the way obviously so the problem is not google services.
the issue is in the interim path
I guess either this is an internet exchange collocation, ISP tier 1 fuckup or some of the non-existent agencies were instructed to sabotage connections. Or they did some interception fuckup.
>Google CEO says over 25% of new Google code is generated by AI
>One of the largest outages in the history of Google follows
It's probably just a coincidence vibe-bros... :P
This quote gets thrown around a lot and I always think about what asdfman123 said here:: https://news.ycombinator.com/item?id=41991291
Vibe correlations
It's not limited to Google, now I wonder if this is actually a Cloudflare/other CDN/International Infra issue or is it that so many services use Google Cloud stuff and they are gone too because of it.
I think all affected websites run on google infra somehow, it's not a cloudflare problem .
(I love how we are using abreviated "infra" bcz we can't google how to write the full word)
It is funny because this is one of the very few words I learned while a kid to construct it from other parts.
So it is "infra" + "structure" but structure is common word so I need to remember how to write "infra" only.
The most bizarre one is the word "yesterday" which I memorized as "yes" + "ter" + "day" and don't ask me why "ter" because I don't remember.
How do you manage "Wednesday"?
I don't, for that I usually rely on search engine or autocorrect. I hate the fact that it is not pronounced as it is written. Because I always go with "Wenesday" on my own.
Ah, yet it's easy!
The day of Wednes. Wednes, old spelling of wedness, the state of being wed.
(So my comment is not totally useless, I'll link to the actual etymology for the ones who are now curious - my sibling comment already alluded to it)
https://en.wiktionary.org/wiki/Wednesday#Etymology
I'm generally for spelling reform, but I like Wednesday's reference to Odin/Woden. Maybe respell it Wedensday and pronounce it like that too.
I'd go all iso format when it comes to that. do you mean "Wed SEP 3"?
I feel like it's an issue in the Google infrastructure as all services which are down seem to be related to Google.
Another example is the HN API, which runs on Firebase.
Spotify as well. I believe they switched from on-premises to Google cloud in 2023.
https://x.com/netblocks/status/1963507222685573522
Currently working fine for me in Pakistan. Btw, you can run https://api.webprobe.org/scan/ in a spare tab in the background and it will periodically check the status of such websites and update the NetBlocks status.
Even connectivitycheck.gstatic.com doesn't seem to be responding, making Androids think they don't have internet access.
Our GCP VMs are also not responding (europe-west4-a and us-central1-b).
edit: Seems to be a network problem. We can't connect to them from Bulgaria, but we can connect to them from the US.
All down here (Turkey).
DNS maybe?
Hmmm maybe that is why www.youtube.com wasn't resolving, but youtube.com is.
Had to change my /etc/hosts to get mpv youtube working today.
No. It works for me.
64 bytes from lhr35s10-in-f14.1e100.net (216.58.206.46): icmp_seq=1 ttl=110 time=47.9 ms
I can ping it too but get 502 errors on google.com and youtube.com
Google and YT works for me just fine: https://www.youtube.com/watch?v=lo4bA4p3MBQ
Maybe depends on region
PING google.com (172.217.17.142): 56 data bytes
Request timeout for icmp_seq 0
Request timeout for icmp_seq 1
Request timeout for icmp_seq 2
Can confirm issues! I can't log in for a google meet call at all.
Everything Google related is down (Gmail, docs, youtube, google, etc.).
Is this the end?
Google Apps and cloud status page down in Romania.
mee too Edit: Firstly their ingresses are not respondable Followup, this impacts all of the traffic which is originated from balkans.
Probably the rest of the Balkans too.
Still up in Slovenia for now.
in Hungary: YouTube is down, other Google services (search, workspace) seem OK as of now.
Seems like it. Youtube as well
Yes, google, youtube, gmail
Armenia, down as well.
Can confirm, from Yerevan - Armenia. Internet slow, then slows down to a halt in a matter of 10 seconds. Then it'll work in a burtst of another 10 seconds, and go back to nothing at all, and stay like that for minutes at a time. Internet on a National level is fast. Any connection to the outside seems to go through a bottleneck.
Down in Bulgaria too.
Seems ok in Canada.
Same in Georgia
In particular the following are not accessible ajax.googleapis.com fonts.gstatic.com gmail.com google.com www.googletagmanager.com www.youtube.com
Yes...
Yes
the initial handshake of the connection in the affected countries is going through all the way obviously so the problem is not google services. the issue is in the interim path
I guess either this is an internet exchange collocation, ISP tier 1 fuckup or some of the non-existent agencies were instructed to sabotage connections. Or they did some interception fuckup.
similar conclusions here
same from Greece, ping works (intermittently), curl returns 502
[dead]