Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
What you talk about makes no sense, the latency of having to hop to the VPN from your location then the Apple server already puts you behind.

Just be like everyone else and preorder like everyone else does. It'll be faster that way in comparison. :)

Theoretically, perhaps. But in practice this has not been the case for me. I'm in CA and generally have had access to very fast TWC business class connection. In prior years, I couldn't access the website or app until 12:20... sometimes later. Yet last year one of my servers, colocated at One Wilshire, could connect much sooner. I think perhaps Apple's servers get overloaded and lump IP blocks into queues or something. All I know is, there is some voodoo going on about who can get in first. It is never the case that "the people with the fastest fingers" get their order in first. It is always "the people who can somehow connect."
 
Theoretically, perhaps. But in practice this has not been the case for me. I'm in CA and generally have had access to very fast TWC business class connection. In prior years, I couldn't access the website or app until 12:20... sometimes later. Yet last year one of my servers, colocated at One Wilshire, could connect much sooner. I think perhaps Apple's servers get overloaded and lump IP blocks into queues or something. All I know is, there is some voodoo going on about who can get in first. It is never the case that "the people with the fastest fingers" get their order in first. It is always "the people who can somehow connect."
Show me benchmarks or it didn't happen. Go ahead and try timing access to the Apple store without VPN, then time access using VPN to access the same site. If you get smaller times end to end, show results. Otherwise, just tell me I'm right (and post results if you wish). :D
 
Show me benchmarks or it didn't happen. Go ahead and try timing access to the Apple store without VPN, then time access using VPN to access the same site. If you get smaller times end to end, show results.

What you're saying is correct from a networking standpoint for a site without access issues. But Apple's servers get the equivalent of a DOS attack every year when preorder happens. I don't have metrics to show you, sorry. I'm saying I've had luck with different servers and trying to see if anyone else out there has too.
 
What you're saying is correct from a networking standpoint for a site without access issues. But Apple's servers get the equivalent of a DOS attack every year when preorder happens. I don't have metrics to show you, sorry. I'm saying I've had luck with different servers and trying to see if anyone else out there has too.
The Apple store is down right now, but the site is still up. You should be able to measure the end to end finish load time to this site using the cases I mentioned in my last post. Go ahead, do it. Challenge me.

AOS-CoverScreen-Desktop-SF-08.gif
 
The Apple store is down right now, but the site is still up. You should be able to measure the end to end finish load time to this site using the cases I mentioned in my last post. Go ahead, do it. Challenge me.

Fine. I'm using a VNC session to connect to the server and running traceroute in a terminal window.

from home:

$ traceroute store.apple.com
traceroute to e2850.a.akamaiedge.net (23.42.91.171), 64 hops max, 52 byte packets
1 <redacted> 0.444 ms 0.215 ms 0.209 ms
2 <redacted> 7.882 ms 5.813 ms 9.872 ms
3 <redacted> 10.054 ms 9.088 ms 10.372 ms
4 <redacted> 18.201 ms 18.434 ms 15.828 ms
5 <redacted> 18.741 ms 16.596 ms 15.943 ms
6 <redacted> 14.210 ms 14.670 ms 15.881 ms
7 <redacted> 7.388 ms
<redacted> 8.004 ms
<redacted> 13.193 ms
8 <redacted> 16.803 ms 17.048 ms 9.789 ms
9 <redacted> 12.342 ms 13.297 ms 7.955 ms

from colocated server:

$ traceroute store.apple.com
traceroute to e2850.a.akamaiedge.net (23.42.91.171), 64 hops max, 52 byte packets
1 <redacted> 0.575 ms 0.433 ms 0.428 ms
2 <redacted> 0.436 ms 1.265 ms 1.137 ms
3 <redacted> 0.752 ms 0.987 ms 0.951 ms
4 <redacted> 0.833 ms
<redacted> 0.683 ms
<redacted> 0.869 ms
5 <redacted> 0.998 ms 0.802 ms 0.775 ms
6 <redacted> 2.170 ms 17.468 ms 12.788 ms
7 <redacted> 1.286 ms 1.034 ms 1.148 ms
8 <redacted> 0.853 ms 0.796 ms 0.837 ms
 
Fine. I'm using a VNC session to connect to the server and running traceroute in a terminal window.

from home:

$ traceroute store.apple.com
traceroute to e2850.a.akamaiedge.net (23.42.91.171), 64 hops max, 52 byte packets
1 <redacted> 0.444 ms 0.215 ms 0.209 ms
2 <redacted> 7.882 ms 5.813 ms 9.872 ms
3 <redacted> 10.054 ms 9.088 ms 10.372 ms
4 <redacted> 18.201 ms 18.434 ms 15.828 ms
5 <redacted> 18.741 ms 16.596 ms 15.943 ms
6 <redacted> 14.210 ms 14.670 ms 15.881 ms
7 <redacted> 7.388 ms
<redacted> 8.004 ms
<redacted> 13.193 ms
8 <redacted> 16.803 ms 17.048 ms 9.789 ms
9 <redacted> 12.342 ms 13.297 ms 7.955 ms

from colocated server:

$ traceroute store.apple.com
traceroute to e2850.a.akamaiedge.net (23.42.91.171), 64 hops max, 52 byte packets
1 <redacted> 0.575 ms 0.433 ms 0.428 ms
2 <redacted> 0.436 ms 1.265 ms 1.137 ms
3 <redacted> 0.752 ms 0.987 ms 0.951 ms
4 <redacted> 0.833 ms
<redacted> 0.683 ms
<redacted> 0.869 ms
5 <redacted> 0.998 ms 0.802 ms 0.775 ms
6 <redacted> 2.170 ms 17.468 ms 12.788 ms
7 <redacted> 1.286 ms 1.034 ms 1.148 ms
8 <redacted> 0.853 ms 0.796 ms 0.837 ms
Did you get one?
 
Fine. I'm using a VNC session to connect to the server and running traceroute in a terminal window.

from home:

$ traceroute store.apple.com
traceroute to e2850.a.akamaiedge.net (23.42.91.171), 64 hops max, 52 byte packets
1 <redacted> 0.444 ms 0.215 ms 0.209 ms
2 <redacted> 7.882 ms 5.813 ms 9.872 ms
3 <redacted> 10.054 ms 9.088 ms 10.372 ms
4 <redacted> 18.201 ms 18.434 ms 15.828 ms
5 <redacted> 18.741 ms 16.596 ms 15.943 ms
6 <redacted> 14.210 ms 14.670 ms 15.881 ms
7 <redacted> 7.388 ms
<redacted> 8.004 ms
<redacted> 13.193 ms
8 <redacted> 16.803 ms 17.048 ms 9.789 ms
9 <redacted> 12.342 ms 13.297 ms 7.955 ms

from colocated server:

$ traceroute store.apple.com
traceroute to e2850.a.akamaiedge.net (23.42.91.171), 64 hops max, 52 byte packets
1 <redacted> 0.575 ms 0.433 ms 0.428 ms
2 <redacted> 0.436 ms 1.265 ms 1.137 ms
3 <redacted> 0.752 ms 0.987 ms 0.951 ms
4 <redacted> 0.833 ms
<redacted> 0.683 ms
<redacted> 0.869 ms
5 <redacted> 0.998 ms 0.802 ms 0.775 ms
6 <redacted> 2.170 ms 17.468 ms 12.788 ms
7 <redacted> 1.286 ms 1.034 ms 1.148 ms
8 <redacted> 0.853 ms 0.796 ms 0.837 ms

That is not a true end to end test. Your running the traceroute from the colocated server? Your not running end to end starting from the location of the computer your on hopping to the colocated server to get to store.apple.com. If you add the extra time taken for you to get to that colocated server from you own machine, it will be higher.
 
What you talk about makes no sense, the latency of having to hop to the VPN from your location then the Apple server already puts you behind.

Just be like everyone else and preorder like everyone else does. It'll be faster that way in comparison. :)

He’s not talking about using a vpn. He’s talking about opening a session on a remote server via something such as vnc, possibly ina region that is networkologically closer / less congested than his local routing. Totally legit tactic.

Edit: sorry didn’t read the whole thread and see this was discussed immediately above.
 
He’s not talking about using a vpn. He’s talking about opening a session on a remote server via something such as vnc, possibly ina region that is networkologically closer / less congested than his local routing. Totally legit tactic.

VNC is even worse. VNC is streaming images... IMAGES.... of a remote machine to your local machine. The delay it takes for you to start that VNC session, have images sent to your screen of that remote server and pass keystrokes/actions to it... needs to be added on top of that server's own time to get to the store.apple.com page.
 
VNC is streaming images... IMAGES.... of a remote machine to your local machine. The delay it takes for you to start that VNC session, have images sent to your screen of that remote server and pass keystrokes/actions to it... needs to be added on top of that server's own time to get to the store.apple.com page.

Yep. Lots of overhead with VNC. But if you happen to be in a region that’s routed to a cluster that’s not scaling traffic well, it could still be a viable option.
 
Yep. Lots of overhead with VNC. But if you happen to be in a region that’s routed to a cluster that’s not scaling traffic well, it could still be a viable option.

Can you clarify what do you mean? VNC's got a lot of overhead, if your routed to a cluster that's not scaling traffic well, you wouldn't even get there? What is it that your trying to say? I'm a tech geek, geek out.
 
Yep. Lots of overhead with VNC. But if you happen to be in a region that’s routed to a cluster that’s not scaling traffic well, it could still be a viable option.

Exactly my point. This is NOT an average network traffic scenario. The real question is, how are Apple's network engineers load balancing? How do these crippling delays play out? In my experience different regions, or perhaps IP block ranges, can get in ahead of others. This is a network congestion issue.

Of course I understand that a VNC session is sending images over the network. I'm not talking about trying to load CNN faster, I'm talking about a strategy for accessing a site under crippling load.
[doublepost=1509077063][/doublepost]
Can you clarify what do you mean? VNC's got a lot of overhead, if your routed to a cluster that's not scaling traffic well, you wouldn't even get there? What is it that your trying to say? I'm a tech geek, geek out.

Let's say you're in Atlanta, GA. There's quite a lot of people in Atlanta, GA trying to access store.apple.com. From your home connection, you cannot load the site due to this congestion. If you can access a server on a backbone however, this congestion is not an issue. Is this making sense now?
 
Exactly my point. This is NOT an average network traffic scenario. The real question is, how are Apple's network engineers load balancing? How do these crippling delays play out? In my experience different regions, or perhaps IP block ranges, can get in ahead of others. This is a network congestion issue.

Of course I understand that a VNC session is sending images over the network. I'm not talking about trying to load CNN faster, I'm talking about a strategy for accessing a site under crippling load.
[doublepost=1509077063][/doublepost]

Let's say you're in Atlanta, GA. There's quite a lot of people in Atlanta, GA trying to access store.apple.com. From your home connection, you cannot load the site due to this congestion. If you can access a server on a backbone however, this congestion is not an issue. Is this making sense now?

If Apple was smart enough, they would have probably already used a CDN or service like cloudfront to distribute the load so something like that wouldn't happen anyways. I'd say that's still a moot point unless historically the website has been hammered to that point in the past.
 
My buddy lives near Tim Cook and guessed his WiFi password. We’re gonna preorder on his network.
 
  • Like
Reactions: Starfyre
My buddy lives near Tim Cook and guessed his WiFi password. We’re gonna preorder on his network.

I wonder how many Apple employees will just go into work and order phones for their friends and family from the Apple corporate network.. :p

Were probably going to be competing with Apple employees waiting for pre-orders to start folks.
 
Exactly my point. This is NOT an average network traffic scenario. The real question is, how are Apple's network engineers load balancing? How do these crippling delays play out? In my experience different regions, or perhaps IP block ranges, can get in ahead of others. This is a network congestion issue.

Of course I understand that a VNC session is sending images over the network. I'm not talking about trying to load CNN faster, I'm talking about a strategy for accessing a site under crippling load.
[doublepost=1509077063][/doublepost]

Let's say you're in Atlanta, GA. There's quite a lot of people in Atlanta, GA trying to access store.apple.com. From your home connection, you cannot load the site due to this congestion. If you can access a server on a backbone however, this congestion is not an issue. Is this making sense now?
Did you get one yet?
 
  • Like
Reactions: ssrij and mcdj
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.