Cloudflare Developers•2y ago

I'm looking for something that gets replicated around the globe (use case:- link shortner)

ⱤⱤØ₥₳₦I'm looking for something that gets replicated around the globe (use case:- link...

Hello, I’m Allie!•2/17/24, 1:33 PM

Just to check what is your ping to 1.1.1.1?

Hello, I’m Allie!•2/17/24, 1:36 PM

That high of latency definitely sounds like a routing issue

Hello, I’m Allie!•2/17/24, 1:36 PM

Also, what is your query speed returned in the metadata?

HHello, I’m Allie!Just to check what is your ping to 1.1.1.1?

Ɽ

ⱤØ₥₳₦OP•2/17/24, 2:04 PM

HHello, I’m Allie!Also, what is your query speed returned in the metadata?

Ɽ

ⱤØ₥₳₦OP•2/17/24, 2:04 PM

Ɽ

ⱤØ₥₳₦OP•2/17/24, 2:17 PM

What does Cf-Placement:- remote-LHR means?

ⱤⱤØ₥₳₦What does Cf-Placement:- remote-LHR means?

Isaac McFadyen•2/17/24, 2:19 PM

That indicates that you have Smart Placement enabled, and that your Worker was placed in the LHR colo.

Isaac McFadyen•2/17/24, 2:19 PM

Colos are named by their nearest airport, so LHR is London

ⱤⱤØ₥₳₦Click to see attachment

Isaac McFadyen•2/17/24, 2:20 PM

So this query latency is very low.

Isaac McFadyen•2/17/24, 2:20 PM

That indicates that the issue definitely isn't with the D1 database itself. Therefore there are two options.

IIsaac McFadyen So this query latency is very low.

Ɽ

ⱤØ₥₳₦OP•2/17/24, 2:20 PM

look at network tab

Isaac McFadyen•2/17/24, 2:21 PM

1. There is high network congestion from Worker -> D1
2. There is high network congestion from You -> Worker.

ⱤⱤØ₥₳₦look at network tab

Isaac McFadyen•2/17/24, 2:21 PM

Yes, but that time is including the time from You -> Worker -> D1.

Isaac McFadyen•2/17/24, 2:21 PM

The durationduration is the actual time D1 took to execute your query, which is very low so indicates the problem isn't with D1 itself.

Ɽ

ⱤØ₥₳₦OP•2/17/24, 2:22 PM

I get it

Ɽ

ⱤØ₥₳₦OP•2/17/24, 2:24 PM

yes

Ɽ

ⱤØ₥₳₦OP•2/17/24, 2:25 PM

that is insane

Ɽ

ⱤØ₥₳₦OP•2/17/24, 2:25 PM

I was thinking that too

Ɽ

ⱤØ₥₳₦OP•2/17/24, 2:25 PM

Isaac McFadyen•2/17/24, 2:26 PM

Hmm, that is quite odd.

Isaac McFadyen•2/17/24, 2:26 PM

Are you making any other requests in the Worker? Perhaps hitting an external API via fetchfetch or similar?

Isaac McFadyen•2/17/24, 2:26 PM

Also that SSL Handshake time is quite high... indicates there might be congestion between you and Cloudflare.

IIsaac McFadyen Are you making any other requests in the Worker? Perhaps hitting an external API...

Ɽ

ⱤØ₥₳₦OP•2/17/24, 2:27 PM

nope, just d1 thing

Isaac McFadyen•2/17/24, 2:27 PM

Even if the Worker is placed in London your connection will still be made to your nearest datacenter...

Isaac McFadyen•2/17/24, 2:27 PM

Can you go to https://cloudflare.com/cdn-cgi/trace and tell us the colo=colo= field?

Ɽ

ⱤØ₥₳₦OP•2/17/24, 2:27 PM

const tpvchk = await env.DB.prepare("SELECT * FROM posts WHERE post_id = ?").bind(post_id).all();
return new Response(JSON.stringify([tpvchk]));

worker has only 2 line of code

IIsaac McFadyen Can you go to https://cloudflare.com/cdn-cgi/trace and tell us the `colo=` field...

Ɽ

ⱤØ₥₳₦OP•2/17/24, 2:28 PM

colo=AMD

Isaac McFadyen•2/17/24, 2:30 PM

Interesting. So that's in India but it doesn't host Durable Objects (which D1 uses). However, the SIN colo does. So I'd expect D1 to be spawned in SIN, unless D1 is disabled in that colo for some reason.

Isaac McFadyen•2/17/24, 2:30 PM

Can you try disabling Smart Placement?

Ɽ

ⱤØ₥₳₦OP•2/17/24, 2:30 PM

D1 is in singapore

Isaac McFadyen•2/17/24, 2:30 PM

That might line up with what you are seeing then.

Isaac McFadyen•2/17/24, 2:31 PM

If D1 is in SIN then that's 300ms away from AMD

Isaac McFadyen•2/17/24, 2:31 PM

https://where.durableobjects.live/colo/AMD/

Ɽ

ⱤØ₥₳₦OP•2/17/24, 2:32 PM

But still that doesn't explain why it is smart placed in london

IIsaac McFadyen Can you try disabling Smart Placement?

Ɽ

ⱤØ₥₳₦OP•2/17/24, 2:38 PM

I disabled smart placement, what's even intriguing is that in USA & EU i'm getting 150ms max, that is super strange

ⱤⱤØ₥₳₦I disabled smart placement, what's even intriguing is that in USA & EU i'm getti...

Isaac McFadyen•2/17/24, 2:38 PM

That would make sense if D1 is in Singapore.

Isaac McFadyen•2/17/24, 2:38 PM

D1 doesn't move.

IIsaac McFadyen That would make sense if D1 is in Singapore.

Ɽ

ⱤØ₥₳₦OP•2/17/24, 2:41 PM

nope, the average latency between most US cities and singapore is 210-250ms (raw latency doesn't include server response time and bla bla), so it still doesn't make sense why is USA i'm getting 150ms

Ɽ

ⱤØ₥₳₦OP•2/17/24, 2:42 PM

and not where i want 150ms (way closer to sin)

ⱤⱤØ₥₳₦nope, the average latency between most US cities and singapore is 210-250ms (raw...

Isaac McFadyen•2/17/24, 2:42 PM

Yes, but is that via Cloudflare's private backbone?

Nicolas4677•2/17/24, 3:34 PM

Looking for some help here, anything would be appreciated.

So, I'm trying to self-host some services without opening any ports in my router. I've been trying to do this through cloudflare's tunnels but no luck yet. I keep encountering this error when I enter my website through my subdomain (see image 1).

The tunnel's hostname setup is the following (see image 2). The reason why I've chosen localhost:8080 is because that's where my service is running at the moment, although I don't know if that's correct. When I open localhost:8080 I'm actually able to see my YouTrack service running.

simplenotezy•2/17/24, 4:26 PM

I am asked to help an owner of an old Battlefield Bad Company 2 server who's a victim on ongoing and recurring DDoS attacks. I recommend him using Cloudflare, but I am not sure how this could be setup. Could anybody point me in the right direction?

soos•2/17/24, 5:04 PM

general-helpWebsocket HLS stream can't be received

would streaming like this be against cloudflare TOS? its just LHLS tunnel for a security camera that wouldn't be on 24/7

Ssimplenotezy I am asked to help an owner of an old Battlefield Bad Company 2 server who's a v...