@IsaacUI said bypassing didn’t help in their case:
Howdy @sdayman unfortunately we’ve been unable to verify @IsaacUI’s claim or tie it directly to any routing issues. We have been in direct communication with him and multiple other users regarding this issue. So far every other user who has disabled cloudlfare has experienced no issues loading their site or accessing it.
At this time we have still been unable to receive any legitimate response from Cloudflare aside from typical ‘check your server’ response. It appears their entry level techs are mis-understanding our ticket and simply copying and pasting from an internal knowledge base.
For customers continuing to have this issue, I recommend opening tickets and working directly with Cloudflare, or disabling cloudflare all together but as of yet their support has been useless.
I can confirm disabling cloudflare resolves these issues.
We’re also having similar issues, hosted on Known host.
We’ve just disabled cloudflare proxying and it seems to clear things up, but obviously this isn’t a permanent solution because now we’re getting hammered with http requests.
Hey there, Knownhost user also. I have 2 servers, 1 is just fine and the other is plagued with 520, 524 and other errors. Apache is showing hundreds of cloudflare IPs hitting it all in a “Reading Request” state and never completing. Knownhost support basically tells me the problem is on Cloudflare’s side.
This issue has been escalated internally and we’re investigating. If you encounter this issue, please let us know here & include your ticket number. If you don’t have a ticket, just let us know here if you are encountering this issue. We’ll post an update on this thread after further investigation.
Same issue, ticekt # 2151187
same issue here as well: #2151070
Same issue ticket 2150605 not responded in 12+ hours and escalated in theory to max priority. VERY poor from CF as we pay a very expensive plan.
Sorry for the issues you’re facing, did you receive the reply from a few hours ago? (If not, lmk and I’ll figure out why not). Were you able to/would you contact your hosting provider/server team and ask them to check why the upstream servers (your origin web servers) are prematurely closing the TCP connection with Cloudflare servers? More details in the reply from a few hours ago.
prematurely closing the TCP connection with Cloudflare servers
@cloonan Where is this coming from? It’s the first I’m hearing of such a symptom…though it’s consistent with something we did see earlier but was only happening from CF and not the rest of the internet.
I’m at a loss with your obsession on blaming the host and not providing any evidence that the issue was actually on the host’s end. “We saw 5xx, talk to your host”. Of course you saw 5xx errors, that’s what we’re telling you and also telling you that to see this across hundreds of servers there’s something bigger happening here than an issue on the server.
Hi there’s been NO REPLY to our ticket, I checked all day also in spam folder.
The provider (that I see replied) opened a ticket with you and had to shout on Twitter to be heard:
The simple thing is: if I disable Cloudflare everything works perfectly, so it definitely is hard to believe that the error is anything if not Cloudflare, unless you think a provider might “close tcp connection” only with CF servers, that doesn’t make any sense. Also because it’s you the one undergoing massive maintenance these months. With your chatbot automatically warning of possible problems because of that.
Absolutely appalling service, we pay a high monthly bill not to wait 12+ hours for a ticket reply, and instead of investigating the problem you clearly have no clue (or no will to investigate and return with a detailed explanation or solution for you or provider) and you keep bouncing back the problem with guesses on what could be wrong on anything but you.
Sorry for the issues, @thereturng, I dropped you a message with the details from 2150605.
Here is late night, I just checked and I received no update email to our ticket, nothing from the opening of the ticket 13 hours ago.
And because this issue is obviously affecting many others, and myself I have another server on same provider and with CF (no ticket opened with that) with same problem, I think would be also correct to publicly report here.
In terms of ticket response, nothing arrived to my mailbox.
I had to go on website to check ticket, and there’s a POOR REPLY from 4h ago, once again telling me all the errors I reported to you already (!!), and telling me it’s a problem on server, and COMPLETELY IGNORING the fact that I told already that is also on other servers (and I didn’t see all the ones reporting here as well!), and mostly that disabling Cloudflare magically everything works so it’s not a problem on the server.
Your ticket saying “suggests that something causes the connection to be reset unexpectedly. This could be a firewall, proxy, load balancer, webserver, or any other software or hardware.” is ludicrous. It’s missing you saying it could be the aliens, because you list every possible cause but something on your side.
So basically you first say it’s a problem of the server, then you read and understand when people tell you that the server works perfectly fine once disabling CF, and so you start accusing the next candidate, the provider. If the provider can show that they’re ok, are you going to point to a specific location of cables of the internet? What then?
Please stop saying you’re sorry for the issues, the reply of 4h ago (that never arrived to mailbox) is even more appalling than what is written here.
We’ll need to ask for accountability for this problem, and also for how is being dealt with amateur and late support, absolutely not worth the Business plan we pay.
Talking of lots of coincidences that happened around this problem…just two hours ago, the massive plan of data centres maintenance disappeared (it was anyway not updated as promised), only 3 remain: Planned maintenance for Cloudflare data centers – Cloudflare Help Center
It’s really hard to not make a series of connections. Yesterday you were working on data centres maintenance, KH reports 520 (and others) errors, Kinsta reports 522 (and others) errors.
You admit at some point the Kinsta problem with Miami data centre (in the list of maintenance), say you’re reverting something that was done in Miami…Kinsta is solved, ‘magically’ (because you insist is not related) also KH problems start disappearing.
And this morning the page of planned maintenance is modified deleting any report of finished maintenance (yesterday Ashburn was listed as done and Miami was in progress or soon to come if I remember correctly), and instead of dozens planned up to July, just 3 remain.
I hope at some point there will be an official explanation after the ‘post mortem investigation’ you promised to KH in their ticket.
Till then, it might be all a long series of coincidences but it’s hard to not see a chain of events all related to each other…and a far from honest or professional attitude dealing with it so far in terms of support to customers and hosting provider.
We’re experiencing same issue on both Pro and Free accounts. We have had to disable Cloudflare to restore uninterrupted service to our sites. Please Cloudflare address this ASAP. See ticket #2151006.
Hi @sales65 I see your ticket with Support and have added myself to it, 2152191
@sales65 Thank you for checking in with us. We have updated your ticket with a response.