Support Open Hours/Availability


#1

Hi folks,

I’m a paying Pro customer and submitted a support request hours ago, does anybody know if support are open currently? Very long wait time…

Best,
Jim


#2

I’ve had to pause Cloudflare due to the error and our site is currently working now. Obviously I dont want to ‘resume’ the site until we know Cloudflare issues are resolved. Still nothing from support, coming up to 5 hours now :frowning:


#3

Support is staffed all day, every day.

Pro plan “mean response time” is 2 hours.

Some issues take longer to respond to than others. But post the ticket # here and one of the mods will probably check on the ticket status.


#4

The ticket # is 1532662
I’ve had a response now, but not actually helpful - pretty much confirmation of what I had sent them already showing the error. Almost 6 hours for a response which hasn’t actually given any information as the support person doesn’t seem to have actually read what was sent over.


#5

I would have to agree, a MRT of 2h is way lower than reality (it was way better, though)… my last support request (the one that you might know about @sdayman) took an average of 8h, 14h once! (granted it was a Sunday, but still middle of the day in the USA and it was only for the first response).


#6

That’s pretty typical based on my experiences. But often when you keep going in the conversation the replies become more insightful, better, and more helpful. :slight_smile:

I occasionally have the idea that the first response is just a default message that applies to a situation and others like it.


#7

It must be the case, but after 5 - 6 hours that is not acceptable. Essentially, I sent an email showing 522 error within a browser from Australia (all customers were getting this in Australia). I also sent an image from UK to the same site, which routes to the same place worldwide, on the same server, working totally fine. All customers other than Australia could connect to our site, no latency issues or time out. Only Australia were showing this message… the response:

“I have checked the error logs for your domain X on Cloudflare for the last 12 Hours and I can see that 522 errors are occurring which means we are unable to connect to your origin”

^ I know this, thats why I emailed support. You should be able to connect to my host, everywhere else in the world can, to the same host! Once I ‘paused’ cloudflare and only used DNS rather than the ‘advanced stuff’, Australia resumed like normal.

This is clearly an issue from within the CDN’s network as when off, it works fine. Shocking support.


#8

That certainly sounds like an issue limited to the Australian POP(s). From the Support document, their Blocked Cloudflare Requests or (less likely) Faulty Network Routing seem to be the likely culprits.


#9

With all due respect you only included the first sentence of the response. The support engineer also included diagnostic information from multiple sources (with screen captures). That information included confirmation that we see the error occurring along with tests from a remote testing tool indicating that connections were currently going through. It also included common cause troubleshooting questions and an offer to follow-up with additional troubleshooting depending on the answers.

Now to be fair I’ve certainly sent replies to support tickets without reading them correctly and I’ve replied to questions here in the forum without clearly understanding what is going on. But if someone were to have kicked his ticket response to me for review (not that they would since I’m not in support) I would have given the response two thumbs up.

And my colleague was correct, this may be a difficult issue to diagnose, intermittent network issues often are.


#10

Hi Cscharff,

The helpdesk agent confirmed that they were getting 522 in the logs and the reason was that you could not connect to my origin. I knew that hence reporting the issue in the first place. I do not need confirmation that the issue I raised… was the issue I raised. That firstly was of no help and very frustrating after 5 hours. I get the origin was not responding from Australia, as that was my initial query to support, but it should have been working as the same origin is responding everywhere else worldwide perfectly fine.

Secondly a tracert to my site that was successful from your helpdesk agent. I gave two tracert (one UK, one Australia) both showing as successful, yet Australia could not connect (533) and UK/other could. All that shows me is that the thing that worked (tracert) when I opened the call still shows as working - aka, no actual help as thats always shown as ok anyway.

Next the tcp traffic test… I had already confirmed that the site was back up, so of course tcp will work now as it is back up. Telling me its all working now, where I had already said its working, offers no help. I know its working.

Then the questions… the agent essentially asked to check the server was ok and check if the line provider were having any traffic issues. <- pointless questions as with Cloudflare ‘paused’ Australia resumed like normal where you were only used as DNS.

Clearly this is a CDN network issue, telling me after 5 hours that you couldn’t contact my origin and to check the server and lines is not helpful.


#11

I’ve now had this response:

“Hi there,
Thanks for coming back to us here and apologies you’re still experiencing issues here.
As previously advised as this is a very intermittent problem, it makes it difficult to track down but we can only go on what we see in the logs. As you can observe from the screenshot I provided most of the 522 errors were seen from our Melbourne and Sydney PoPs which correlates to what you’re said.
At the moment we’re not able to replicate but these errors typically indicate we were unable to connect to your origin server.
I would recommend to check the logs at your origin for any issues with packet loss or if your server was under load at the time the errors happened and make sure that all Cloudflare IP ranges are fully whitelisted from any rate limits.
Sorry but as I said since the issue is intermittent we can only go by what we’re seeing in the logs.
Let me know if I can be of any further assistance here and I’ll be glad to help.”

The issue cannot be replicated currently as it is now working.
You could connect to my server from EVERYWHERE ELSE! Its not my fault that you cannot from Australia all of a sudden. Thats your issue.

‘Check my logs’ - so trying to pass the blame for the outage obviously.

No comment on why pausing the site and not using the more advanced CDN features then resolved the issue (hmm, because that would be accepting fault?)

Not happy.


#12

Did you actually check logs and settings? While they may not be the cause in this instance they are common causes (in fact @sdayman linked to articles with the same information).


#13

Yes, this was done at around 7:40am whilst the outage was ongoing and where the original request was made. The server, leased line, firewall were working perfectly and responding as expected.


#14

None of those questions answer why pausing the site resolved all issues. If I understand correctly pausing means Cloudflare is then only used as DNS and that traffic would transverse to the UK over the Internet rather than being forced through Cloudflare networks?

If so, pausing resolving the issue means the CF network had a fault ‘somewhere’. Otherwise pausing would have done nothing to alleviate the issue. (if it were my lines or server)

Or am I missing something here?


#15

This is certainly frustrating. My theory is there’s a firewall that’s being triggered by Cloudflare’s Australia POP requests. Maybe the requests are slow, or from an IP address the firewall doesn’t like, so it rate limits or blocks it for a bit.


#16

Cloudflare routes traffic from it’s POPs over the public internet to the origin, so while there may have been an issue connecting it could have been due to an issue anywhere along that path. So it does take a s different path potentially, but those paths may converge at some point which is part of the reason support asked to make sure Cloudflare IPs were whitelisted.

It can be difficult to determine at what point on a network path a connection attempt might be failing when you don’t control all of the hops/ devices in place hence the indicated difficulty in troubleshooting when the issue is intermittent… one has to test while the problem is occurring and hope that corresponding diagnostic information points to a cause.


#17

Its been fine now since midday UK time and before today had actually been of for about two years. It’s just very frustrating as I cannot say to Australian customers to stay down for 5 hours or so hoping for Clousflare support to respond at some future time. All I can do is pause the site, which resumes connectivity, then guess when it will be working via cloudflare and enable again (if support replied during that time at all).


#18

Yes, this makes sense. Thanks cscharff. Either way during the outage all other locations worldwide were performant and route to the same endpoint… that too points away from the issue being at my end. Being told to check my stuff, as I am sure you would agree, is very frustrating. If it were here, all customers would be affected because they all come through to my same server on the same line.


#19

Absolutely not. 100% not this.


#20

Why?

If my line were dropping traffic, or the server was having a performance issue, that would affect all customers… my server and line could not decide to only drop packets or performance from Australian customers. It would affect all if performance, or none.