Connection lag experienced by users in Australia

Really hoping someone can help with this as it’s been a problem for at least a year now. Many visitors to our site ( )who are based in Australia tell us that our site freezes and doesn’t load properly, typically they also get a page unresponsive message. This problem appears to be unique to Australia and seems to affect most but not all customers located in that country. We have also had one report of the same problem from a user in New Zealand too. They all confirm that their devices perform perfectly when visiting other sites so this is definitely nothing to do with their broadband connection.

We had an development team look at the issue but after running some tests they advise us to speak to netlify support and ask how many websites are hosted on the same server as they suspect that could be the problem?


Hi @andrewharmer :wave:t6: welcome to the forums! I appreciate you raising this concern to us. I have escalated this to our helpdesk so our support engineers can dive deeper into what’s going on. Please stay tuned to your email for an update.

I replied to the support ticket and I just sent a second reply there because I discovered something else.

The traffic from Cloudflare isn’t going to all US CDNs node. It is only going to the or IP addresses (which I cannot say for certain without more research but this clue alone helps).

I really do think if you modify Cloudflare to use the <site subdomain> format the performance will improve. If it does not, please reply here to let us know.

Hi Luke,

Thanks for the series of replies. Just to give you a bit more background on the issue, the problem has been ongoing for over a year and a couple of weeks ago a guy who has been trying to solve the problem for me (cc’d on this email) suggested we try using cloudflare. He set it up but it made no difference and the problem continued. He then suggested I contact you guys for help. I therefore think cloudflare is a red herring and would prefer to switch it off as it has only added to the ’noise’ around the real problem.

I am not that technical so please "reply all” so Tomasz can also see your recommended course of action

Thanks and best regards


Andrew Harmer
Managing Director

t: +44 (0) 1273 410550

Transp-blue-1200 copy.png

Hi Luke,

Did you get a chance to check into this further following my last email?

Foe your info we have stopped using cloudflare, which we only used for a week or two anyway. As I said in my last email this problem has been around for over a year. Our tech team suggested trying cloudflare as a solution but it hasn’t worked so we have reverted to our former settings. I would be grateful if someone could take another look at our traffic from Australia to see if the real cause of the issue can be identified?

Thanks and regards


Andrew Harmer
Managing Director

t: +44 (0) 1273 410550

Transp-blue-1200 copy.png

Hi @andrewharmer :wave:t6: nice to see you again. I’m escalating this back to our help desk to see if our team has an update. (:

Hi, @andrewharmer. With the current DNS configuration, everyone in the world goes to San Francisco and no CDN is used at all. This is the current DNS configuration for this site:		300	IN	A

For the www subdomain no DNS records exist:

$ dig

; <<>> DiG 9.10.6 <<>>
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NXDOMAIN, id: 53691
;; flags: qr rd ra; QUERY: 1, ANSWER: 0, AUTHORITY: 1, ADDITIONAL: 1

; EDNS: version: 0, flags:; udp: 512
;		IN	A

;; AUTHORITY SECTION:		900	IN	SOA 1 7200 900 1209600 86400

;; Query time: 66 msec
;; WHEN: Mon Jan 02 19:23:04 PST 2023
;; MSG SIZE  rcvd: 131

That shows the start of authority (SOA) record which indicates the request record (for for not exist.

This means no CDN is being used. All site traffic goes to a single IP address near San Francisco.

To resolve this, two changes are required.

1. The DNS record below must be created (while replacing <site subdomain here> with the correct subdomain):		1800	IN	CNAME	<site subdomain here>

So, if the site’s subdomain is example (meaning, then the CNAME type DNS record would be this:		1800	IN	CNAME

2. The site’s primary custom domain (found under Site Name > Settings > Domain management > Custom domains) would need to be changed from to

To summarize, first the CNAME record for the www subdomain should be created and ten the site’s primary custom domain must be changed from to If both of those changes are made, the site performance will improve drastically as people will be directed to their closest CDN node geographically. This means that people in Australia will be directed Australian CDN nodes, people in the EU will be directed to EU CDN nodes, etc.

If that doesn’t resolve the issue or if there are questions, please let us know.

Hi Luke,

My tech team are telling me they made the changes requested in your last email but our website is still not working properly for customers in Australia.

This is very concerning as these customers are now turning to competitors of ours so I need to resolve this somehow.

Are you able to check that all changes made were correct and let me know if we need to do anything else?

Thanks and best regards


Andrew Harmer
Managing Director

t: +44 (0) 1273 410550

Transp-blue-1200 copy.png

Could you share a HAR file recording showing slow loads?

H there,

We think the attached HAR file (recorded in Australia) demonstrates the issue but please let me know if it doesn’t and we will try again.

Best regards

Andy (2.26 MB)

When sorted with time taken descending, I can see, this is the Netlify request that took the maximum time:

The ones above that are not served by Netlify. 390ms seems reasonable and expected time to us. Am I missing something?