My workers started failing; I upgraded to business but I still am not able to access support.
#I upgraded to business but I still am not able to access support.
modern spindle
grave copper
My workers started failing; I upgraded to business but I still am not able to ac...
It can take a bit to update, logging out/in may help. If it's something we could potentially help with here feel free to ask as well. Worker Exceptions/1101s for example should be tailed to get more information out of, etc
modern spindle
i got in eventually, but this morning i'm seeing a cascade of errors with CF workers; when i try to initiate chat I get other system errors:
There was a problem opening your chat. Please refresh your browser and try again. If this problem persists, open a support ticket or please report it in the Cloudflare Community.
In this last 24 hours I've seen at least FOUR CF service fail with internal system errors. Is the company aware of this? Are there urgent steps being taken? How paying business users depending on this for production services get support?
grave copper
i got in eventually, but this morning i'm seeing a cascade of errors with CF wor...
there's currently a dashboard issue: https://www.cloudflarestatus.com/incidents/x7q6v6rtgyq2
There's no edge impact though, everything's running fine on that front. What services are you seeing fail?
modern spindle
hi. i saw that. since last night my workers are failing sporadically; routes are getting automatically disabled; the dashboard shows frequent 500 and 504 errors; my clients are getting 1042 errors; requests to MailChannels are returning 500 errors; the dashboard shows redline system failures, and the Support Chat process consistently shows: T"here was a problem opening your chat. Please refresh your browser and try again. If this problem persists, open a support ticket or please report it in the Cloudflare Community."
it feels like a complete meltdown
everything working fine yesterday morning.
i have a case open https://support.cloudflare.com/hc/en-us/requests/3236221 but I don't think the tech supporting the issue understand the problem since it was diagnosed as due to rate limiting.
mystic hinge
requests to MailChannels are returning 500 errors
thats not a cloudflare issue
grave copper
hi. i saw that. since last night my workers are failing sporadically; routes are...
Routes are getting automatically disabled
be careful that you're not modifying them in the dashboard and then using wrangler cli to push a new version (or wrangler github action), it'd override them all
the dashboard shows frequent 500 and 504 errors
this should be fixed now
my clients are getting 1042 errors
1042 is this: "Worker tried to fetch from another Worker on the same zone, which is unsupported."
https://developers.cloudflare.com/workers/observability/errors/
requests to MailChannels are returning 500 errors
external
(I'm not a cf employee, but can try to help)
I haven't seen any issues on my deployed workers today nor seen any other mass reported ones
modern spindle
ok -- thanks -- this is really helpful already
re 1042 -- how then can a Pages Funciton call a Worker?
grave copper
All the pages functions run on pages.dev zone, so they could call workers on other zones (like workers.dev, or one on a custom domain)
they run on the pages.dev even if they are called through a custom domain
If you want to call a worker with a worker, service bindings are recommended and even faster/lower latency/less potential errors
https://developers.cloudflare.com/workers/runtime-apis/bindings/service-bindings/
modern spindle
ok thanks -- that very likely accounts for one of my issues then. i'll take that out of the loop to start.
grave copper
the zone executing thing also applies to mailchannels. It should also return a more exact error like (this zone is not authorized), but for the mailchannels cfid stuff, the cfid is off the CF-Worker header Workers automatically send, and CF-Worker header is set off the following:
Worker: If it's ran via your workers.dev, it's yourworkersubdomain.workers.dev
Worker: If it's ran via route or custom domain, it's the zone/website name (may be a subdomain if using partial setup)
Function: If it's a Pages Function, it's always the projectname.pages.dev even if called via a Custom Domain
Worker: If it's something that runs without an http request, like crons, queues, etc, it's your workers.dev
could be why it's 500ing, would be made more clear if you have the body of the failed requests
modern spindle
ok that's really helpful again. thank you.
modern spindle
hi again. things have settled down. the dashboard errors yesterday have gone away and your explanation re calling workers from other workers and pages was really helpful.
i'm still getting 500s from MailChannels and I assume this is due to a lockdown misconfiguraiton now that i'm using a custom domain for my production worker.
my DNS looks right to me though
dig @1.1.1.1 _mailchannels.dxos.org TXT
_mailchannels.dxos.org. 300 IN TXT "v=mc1 cfid=hub.dxos.network"
(my worker's custom domain is hub.dxos.network) and it's sending mail from dxos.org -- hence the DNS record above.
there's no additional info in the response from MC when I log that. appreciate any suggestiong you might have about that.
grave copper
(my worker's custom domain is hub.dxos.network) and it's sending mail from dxos....
it's based off the cf-worker header, which is your zone/website
essentially: What's the name of the website under websites? It's probably dxos.network, which is what that should be
and just to verify, that is a worker yea? Not a Pages/Pages function, in any way?
modern spindle
yes it's a worker deployed to hub.dxos.network
i thought it would need the subdomain.
ok that worked. i've been moving things around a lot and haven't quite internalized how all of this works.
thanks again for all of your help.
vital umbra
Would anyone please help me? IVE been hacked and it's nasty, I cannot go on any longer and am desperate for some help.