No longer possible to opt-out of transparent HTTP cache via My Account?

Forums › 2degrees (including Slingshot, Orcon, Flip, Stuff Fibre, MyRepublic, 2talk and Vocus) › No longer possible to opt-out of transparent HTTP cache via My Account?

Detruire

1767 posts

Uber Geek

#151688 1-Sep-2014 22:40

I'm not sure how long this has been the case for (as I just noticed this tonight), but the cache settings available via the "My Account" section of the Slingshot website no longer has any effect (for my connection at least.) Regardless of the status on that page, the cache definitely still seems to be enabled.

I tried the following, to no effect:

toggling the setting "on" then "off" again, waiting 10 minutes, then rebooting my router
changing it to "on", waiting 10 minutes, then rebooting my router
changing it back to "off", waiting 20+ minutes, then rebooting my router
disabling Global Mode in case that forced the cache on (or itself acted like one)

The "Internet Cache Settings" page currently states "You currently have caching OFF", but this cache test indicates a cache is still present, as does tcping:

> tcping bbc.co.uk

Probing 212.58.246.103:80/tcp - Port is open - time=20.411ms
Probing 212.58.246.103:80/tcp - Port is open - time=20.625ms
Probing 212.58.246.103:80/tcp - Port is open - time=19.503ms
Probing 212.58.246.103:80/tcp - Port is open - time=96.258ms

Ping statistics for 212.58.246.103:80
     4 probes sent.
     4 successful, 0 failed.
Approximate trip times in milli-seconds:
     Minimum = 19.503ms, Maximum = 96.258ms, Average = 39.199ms

> tcping bbc.co.uk 443

Probing 212.58.246.103:443/tcp - Port is open - time=327.677ms
Probing 212.58.246.103:443/tcp - Port is open - time=288.066ms
Probing 212.58.246.103:443/tcp - Port is open - time=288.123ms
Probing 212.58.246.103:443/tcp - Port is open - time=286.196ms

Ping statistics for 212.58.246.103:443
     4 probes sent.
     4 successful, 0 failed.
Approximate trip times in milli-seconds:
     Minimum = 286.196ms, Maximum = 327.677ms, Average = 297.516ms

>tcping internode.on.net

Probing 150.101.140.197:80/tcp - Port is open - time=28.053ms
Probing 150.101.140.197:80/tcp - Port is open - time=19.405ms
Probing 150.101.140.197:80/tcp - Port is open - time=19.715ms
Probing 150.101.140.197:80/tcp - Port is open - time=19.551ms

Ping statistics for 150.101.140.197:80
     4 probes sent.
     4 successful, 0 failed.
Approximate trip times in milli-seconds:
     Minimum = 19.405ms, Maximum = 28.053ms, Average = 21.681ms

>tcping internode.on.net 443

Probing 150.101.140.197:443/tcp - Port is open - time=68.325ms
Probing 150.101.140.197:443/tcp - Port is open - time=68.501ms
Probing 150.101.140.197:443/tcp - Port is open - time=68.166ms
Probing 150.101.140.197:443/tcp - Port is open - time=66.203ms

Ping statistics for 150.101.140.197:443
     4 probes sent.
     4 successful, 0 failed.
Approximate trip times in milli-seconds:
     Minimum = 66.203ms, Maximum = 68.501ms, Average = 67.799ms

Is anyone else here seeing the same thing? If so (or if not), what plan/connection type are you on?

rm *

Slingshot

1004 posts

Uber Geek

Trusted

Slingshot

#1120346 2-Sep-2014 11:00

Hey Detruire, thanks for the heads up. Looks like a bug with UFB customers. The guys are working on it now so should hopefully have a solution soon.

www.slingshot.co.nz

www.twitter.com/SlingshotNZ

https://www.facebook.com/SlingshotNZ

Lenovo

Shop now for Lenovo laptops and other devices (affiliate link).

eXDee

4032 posts

Uber Geek

Trusted

#1120633 2-Sep-2014 17:47

Time to fix the iRules then? :D

Brumfondl

1186 posts

Uber Geek

Trusted

#1120663 2-Sep-2014 18:09

Very interesting. I had this issue a while ago when trying to get WIldstar to patch (eventually fixed by Nebbie bypassing the cache for everyone to get to the Wildstar patch server adresses).

Will be interesting to see what this affects since I have never been back in to turn the cache back on on may account.

As an aside, does anyone else get DNS issues on Slingshot? I often have to flush my DNS cache because sites stop working properly.

Brumfondl

Detruire

1767 posts

Uber Geek

#1120824 2-Sep-2014 21:01

Brumfondl: Very interesting. I had this issue a while ago when trying to get WIldstar to patch (eventually fixed by Nebbie bypassing the cache for everyone to get to the Wildstar patch server adresses).

I've had it disabled on my account (and my parents' before that) since it was possible to do so. Back then I used to run into sites that wouldn't update, downloads that would crawl, and a bunch of other weird annoying problems that all went away without the cache.

The fact that it took me almost a month to notice (rather than a couple of days last time it had been left on) goes to show that there's been an improvement in that area.

Brumfondl: As an aside, does anyone else get DNS issues on Slingshot? I often have to flush my DNS cache because sites stop working properly.

Yes. I assumed it was something up with the UFB router under load, but I haven't looked into it.

rm *

Detruire

1767 posts

Uber Geek

#1130507 17-Sep-2014 11:00

Any update on this?

rm *

OmniouS

423 posts

Ultimate Geek

Trusted

Lifetime subscriber

#1154005 14-Oct-2014 22:40

Yes, please provide an update on this. Turning off the cache via the management portal doesn't have any effect. We'd like to see if turning it off fixes the problems we've been having.

On all PCs in the house, pages are not loading intermittently. More often than not, it is CDN hosted content that won't load and makes sites like eBay and stuff completely unusable. e.g.*.ebaystatic.com, staticx.stuff.co.nz. Sometimes Google searches will fail with connection reset errors

I might install Fiddler or similar to try and work out what's going on. We're a new UFB customer using Slingshot supplied router. These issues didn't happen with naked ADSL

Edit (from work PC via VPN) - Each Geekzone page is now taking ~10 secs to load on my home connection as pages are blocked waiting on resources from Google to load

OmniouS

423 posts

Ultimate Geek

Trusted

Lifetime subscriber

#1154026 14-Oct-2014 23:11

Possibly related - We can ping machines on the local network with an MTU of 1472. Anything behind the router fragments above 1372. I can't see where to change MTU settings on the device..

Compare this to my working connection @ the office - we can ping anything local or on the Internet with an MTU of 1472

Edit: And something else - ping to the first hop after the router is highly variable. Fair enough if they deprioritise it but the range is higher than what I'd expect:

Detruire

1767 posts

Uber Geek

#1161815 24-Oct-2014 22:46

Any news on this front?

rm *

OmniouS

423 posts

Ultimate Geek

Trusted

Lifetime subscriber

#1162928 27-Oct-2014 13:05

+1

If no clear resolution in sight, we'd like to exit our contract without penalty

Slingshot

1004 posts

Uber Geek

Trusted

Slingshot

#1163315 28-Oct-2014 09:24

Hey guys,

I know this is a pain, I was hoping this would be sorted a while ago. Can everyone in here having issues send me a message with your account numbers? If I've got a bunch of examples I can try and bump it up the urgency list.

Cheers,
Chris

www.slingshot.co.nz

www.twitter.com/SlingshotNZ

https://www.facebook.com/SlingshotNZ

Detruire

1767 posts

Uber Geek

#1217802 19-Jan-2015 19:02

I see this still hasn't been fixed.

rm *

pushkar90

16 posts

Geek

#1219174 21-Jan-2015 14:50

starting to completely lose faith in slingshot