ticket tips and tricks Glastonbury 2024 resales

[Fish Bulb]

13 hours ago, parsonjack said:

Yes I've carried out similar tests to @stuie in the past and determined similarly that there is a refresh rate limit that should be avoided, hypothesising at the time that it was IP based, and that there could be risks where shared connections were used on a single IP.  

@stuie's tests do seem to suggest now that the 'block' is not solely IP based, which would be sensible so as not to disadvantage customers on shared connections, and it's therefore possible that See's system inspects requests at a lower level than IP, MAC address that is, to ensure the 'block' is device based rather than solely IP.

The only fly in the ointment here is that there has been evidence in the past, admittedly some time back, of multiple devices all on the same connection getting a booking page once one device does. This, if the case and not just anecdotal, *could* disprove the above.

The crux of all this is that we don't know for sure how it all works, and load balancing could add an extra level of complexity to boot.  We have some good ideas, but ultimately unless we can get a See tech to give us a PowerPoint we remain, somewhat, in the dark.

Whilst we can't be fully certain my advice would therefore be to use 1 browser, 1 tab, 1 device, with multiple devices using separate connections eg 1 x WiFi, 1 x 4G etc.  All devices then stay within the 1/second F5 rate. This would minimise any risk of 'blocking' and simultaneously maximise request attempts and therefore chance.

 

Very thorough. I know this is partly theory but just so I understand the thinking right:

Are you saying that if I had a Chrome tab on a 2 second auto-refresh, and a separate Edge tab on the same device also on a 2 second auto-refresh - as these share an IP address they aren't really separate  and combined would trigger the (supposed) 1 second limit?

[t0paz]

Just now, Fish Bulb said:

Very thorough. I know this is partly theory but just so I understand the thinking right:

Are you saying that if I had a Chrome tab on a 2 second auto-refresh, and a separate Edge tab on the same device also on a 2 second auto-refresh - as these share an IP address they aren't really separate  and combined would trigger the (supposed) 1 second limit?

Yes. That’s exactly what I tested yesterday.

That’s not how I previously thought it worked so now realise I could have been shooting myself in the foot the last few sales!

[Neil]

the too many refreshes lock out is session based, so anything from the same browser on one machine, even if different windows or tabs.

[t0paz]

1 minute ago, Neil said:

the too many refreshes lock out is session based, so anything from the same browser on one machine, even if different windows or tabs.

I got locked out in all browsers at once on the same machine. So I think the suggestion it might be MAC based is quite likely correct.

[Fish Bulb]

3 minutes ago, t0paz said:

Yes. That’s exactly what I tested yesterday.

That’s not how I previously thought it worked so now realise I could have been shooting myself in the foot the last few sales!

Thanks. I did actually just test the same thing in the office with 3 browsers all doing 2 seconds and two of them did hit the "holding page" at about the same time. Interestingly though the 3rd one never did. Also I didn't switch off the auto-refresh and that actually bounced me back off the holding page on those two browsers (I didn't have to wait 20 seconds in holding jail).

This isn't getting any clearer is it!

[Neil]

Just now, t0paz said:

I got locked out in all browsers at once on the same machine. So I think the suggestion it might be MAC based is quite likely correct.

mac address isn't so easy to utilise. session is a breeze.

[t0paz]

2 minutes ago, Fish Bulb said:

This isn't getting any clearer is it!

No, it’s not. I’ve also seen the same thing with being randomly being let back in.

I’ve also noticed some refreshes the text is aligned differently which would suggest one or more servers aren’t configured exactly the same as the others.

Everything observed now should be taken with a massive pinch of salt

[BenG92]

Going for a PC on a VPN, laptop on wifi, phone on 5G, and just refreshing each manually in a row. Wouldn't trust auto refreshers as far as I can throw them personally. Just click browser, F5, click browser, F5, ad nauseum.

Edited October 20, 2022 by BenG92

[leath02]

8 minutes ago, BenG92 said:

Going for a PC on a VPN, laptop on wifi, phone on 5G, and just refreshing each manually in a row. Wouldn't trust auto refreshers as far as I can throw them personally. Just click browser, F5, click browser, F5, ad nauseum.

Have read somewhere (could be complete BS) that VPNs get booted

[Crazyfool01]

I’ve got my second internet connection going in today which is much faster than current . My first contract I’ve cancelled a few days after T day . Will likely make not much difference but at least I feel like I’m doing something different after failing to get anywhere in last 2 main sales 

[Neil]

5 minutes ago, leath02 said:

Have read somewhere (could be complete BS) that VPNs get booted

if the blocking is ip addressed-based, then a vpn is a bad idea as you're likely to share the vpn ip with another vpn user.

[Neil]

4 minutes ago, Crazyfool01 said:

I’ve got my second internet connection going in today which is much faster than current . My first contract I’ve cancelled a few days after T day . Will likely make not much difference but at least I feel like I’m doing something different after failing to get anywhere in last 2 main sales 

won't make any difference, i once tried four devices with a unique ip for each, i failed to get tickets in the sale that year.

[stuie]

35 minutes ago, Neil said:

the too many refreshes lock out is session based, so anything from the same browser on one machine, even if different windows or tabs.

That's what I found.  Will check again sometime before tday though.

Of course, it might change on the day when loads are balanced.

33 minutes ago, t0paz said:

I got locked out in all browsers at once on the same machine. So I think the suggestion it might be MAC based is quite likely correct.

That's not what happened in my tests yesterday. 

Chrome locked out and Edge and ie were fine. 

[Havors]

13 hours ago, parsonjack said:

Yes I've carried out similar tests to @stuie in the past and determined similarly that there is a refresh rate limit that should be avoided, hypothesising at the time that it was IP based, and that there could be risks where shared connections were used on a single IP.  

@stuie's tests do seem to suggest now that the 'block' is not solely IP based, which would be sensible so as not to disadvantage customers on shared connections, and it's therefore possible that See's system inspects requests at a lower level than IP, MAC address that is, to ensure the 'block' is device based rather than solely IP.

The only fly in the ointment here is that there has been evidence in the past, admittedly some time back, of multiple devices all on the same connection getting a booking page once one device does. This, if the case and not just anecdotal, *could* disprove the above.

The crux of all this is that we don't know for sure how it all works, and load balancing could add an extra level of complexity to boot.  We have some good ideas, but ultimately unless we can get a See tech to give us a PowerPoint we remain, somewhat, in the dark.

Whilst we can't be fully certain my advice would therefore be to use 1 browser, 1 tab, 1 device, with multiple devices using separate connections eg 1 x WiFi, 1 x 4G etc.  All devices then stay within the 1/second F5 rate. This would minimise any risk of 'blocking' and simultaneously maximise request attempts and therefore chance.

 

Maybe session blocking using tokens? Has anyone tried being blocked then completely restarting with a clear cache? 

[scatteredscreens]

We need someone on the inside...

[Neil]

1 minute ago, scatteredscreens said:

We need someone on the inside...

they won't give the details as then we'd know how to beat their system.

[scatteredscreens]

1 minute ago, Neil said:

they won't give the details as then we'd know how to beat their system.

Not with enough interrogation..

[Havors]

3 minutes ago, Neil said:

they won't give the details as then we'd know how to beat their system.

Someone knows haha! A friend of mine and his group he goes with have been successful every year for the past 15 years. He said one of the guys gets their tickets and knows the system. Gets 12 of them tickets every year without fail. Wont divulge how haha... wouldnt surprise me if it was illegal haha 

[Neil]

1 minute ago, Havors said:

Someone knows haha! A friend of mine and his group he goes with have been successful every year for the past 15 years. He said one of the guys gets their tickets and knows the system. Gets 12 of them tickets every year without fail. Wont divulge how haha... wouldnt surprise me if it was illegal haha 

if their mate works for see he can sort their tickets.

[Havors]

1 minute ago, Neil said:

if their mate works for see he can sort their tickets.

Apparently he doesnt he is an infrastructure engineer by trade haha 

I say hes just a hacker myself 😄 

[scatteredscreens]

6 minutes ago, Havors said:

Apparently he doesnt he is an infrastructure engineer by trade haha 

I say hes just a hacker myself 😄 

Maybe he's got an ethernet cable running from the SeeTickets data centre straight to his laptop

[leath02]

https://www.reddit.com/r/glastonbury_festival/comments/ddgggb/a_web_developers_advice_on_getting_tickets/

 

[incident]

1 hour ago, Neil said:

the too many refreshes lock out is session based, so anything from the same browser on one machine, even if different windows or tabs.

It's more complicated than that. There's at least some other factors at play beyond the session - some form of browser/device fingerprinting I believe.

I did some experiments of my own a few years ago, got results similar to those reported above (ie multiple devices "locked out" at once), but couldn't figure out exactly what triggers what so just structured my setup in such a way that avoids the question entirely.

[gfa]

1 hour ago, BenG92 said:

Going for a PC on a VPN, laptop on wifi, phone on 5G, and just refreshing each manually in a row. Wouldn't trust auto refreshers as far as I can throw them personally. Just click browser, F5, click browser, F5, ad nauseum.

 

1 hour ago, leath02 said:

Have read somewhere (could be complete BS) that VPNs get booted

Ticket websites will likely not favour VPNs because these are what people use to exceed ticket limits with etc. Likely theres a universal block list of all VPNs i'd have thought.

36 minutes ago, Havors said:

Someone knows haha! A friend of mine and his group he goes with have been successful every year for the past 15 years. He said one of the guys gets their tickets and knows the system. Gets 12 of them tickets every year without fail. Wont divulge how haha... wouldnt surprise me if it was illegal haha 

There are ways to get a ridiculous amount of tabs open and all but guarantee one getting in or probably bypasses to just get straight through

[parsonjack]

45 minutes ago, stuie said:

That's not what happened in my tests yesterday. 

Chrome locked out and Edge and ie were fine. 

Yup this is where the load balancing could be part of the story....the request from Chrome is directed in such a manner that it's seen as breaching the limit, but the Edge request is directed elsewhere and seen as within the limit.