Jump to content
SAU Community

Recommended Posts

I've been having a few problems accessing this site - it sometimes hangs and won't respond. Every time it does it is trying to access

ad.yieldmanager.com

I assuming you're using that to monitor or generate adds on each page.

Can you check to see if it is occasionally hanging or taking a long time to respond? Do you get any stats from it? Is there any way your code can bypass it if it is not responding within a few seconds?

Ian

Link to comment
https://www.sau.com.au/forums/topic/264307-adyieldmanagercom-causing-problems/
Share on other sites

What broswer? Internet speed?

I cant say I've ever had a problem and im on a fair bit :thumbsup:

ADSL2 - close to the exchange ie I actually get 17mbits/sec download (not just line sync). Firefox, latest version. Windows XP. Pretty sure it isn't my system.

And it comes and goes - ie today not at all, yesterday a lot...

Ian

I sometimes have the same issue, it shouldn't matter, ad providers are sometimes down, or slow,

or unreachable, _except_ it blocks the page rendering!

Also, and off this topic, it takes 8 seconds to fully render an average topic, 7 of those are waiting

and loading the first and second and third rows.. before getting to any meat you want to read..

most of this seems like CPU overload, but it doesn't help much that, for me, the site is a full 27 hops away..

a ridiculous trip from telstra thru to LA, down the coast, across the states, before ending up in detroit!

and about as awful a ping time as you can imagine, 220ms

Are hosting fees in australia still so out of wack that it must be hosted on the other side of the world?

Japan would probably be as cheap.. but half the ping.. and closer to the heartland of rb26 than detroit, haha.

& if you hosted locally you'd double page views per day, because we'd all spend our allocated SAU time

reading twice as much, and over the long term, increase membership as well..

2 172.18.113.141 (172.18.113.141) 9.265 ms 7.703 ms 7.292 ms

3 172.18.66.150 (172.18.66.150) 7.254 ms 7.048 ms 7.221 ms

4 172.18.239.161 (172.18.239.161) 7.391 ms 7.428 ms 7.692 ms

5 tengigabitethernet4-2.chw45.sydney.telstra.net (203.45.3.37) 6.859 ms 7.128 ms 7.471 ms

6 tengige0-1-0-0.chw-core2.sydney.telstra.net (203.50.20.129) 7.220 ms 7.024 ms 7.215 ms

7 bundle-ether1.oxf-gw2.sydney.telstra.net (203.50.6.90) 7.458 ms 7.099 ms 7.093 ms

8 tengigabitethernetx-0.syd-core04.sydney.reach.com (203.50.13.18) 7.265 ms 7.275 ms 7.679 ms

9 i-4-1.sydp-core02.net.reach.com (202.84.144.249) 10.815 ms 7.958 ms 7.850 ms

10 i-0-0.wil-core02.net.reach.com (202.84.144.101) 189.240 ms 189.576 ms 189.582 ms

11 i-7-8.tlot03.net.reach.com (202.84.251.238) 151.368 ms 152.728 ms 151.792 ms

12 level3.tlot03.net.reach.com (134.159.63.30) 190.326 ms ge-6-21.car3.losangeles1.level3.net (4.68.111.137) 189.740 ms ge-6-20.car3.losangeles1.level3.net (4.68.111.133) 191.287 ms

13 vlan89.csw3.losangeles1.level3.net (4.68.20.190) 191.405 ms 190.105 ms 198.691 ms

14 ae-83-83.ebr3.losangeles1.level3.net (4.69.137.41) 195.792 ms 191.793 ms 198.217 ms

15 ae-2.ebr3.sanjose1.level3.net (4.69.132.9) 197.714 ms 207.101 ms 199.445 ms

16 ae-73-73.csw2.sanjose1.level3.net (4.69.134.230) 208.788 ms 198.976 ms 198.203 ms

17 ae-72-72.ebr2.sanjose1.level3.net (4.69.134.213) 160.760 ms 171.906 ms 161.640 ms

18 ae-3.ebr1.denver1.level3.net (4.69.132.58) 235.646 ms 227.547 ms 234.647 ms

19 ae-1-100.ebr2.denver1.level3.net (4.69.132.38) 228.585 ms 228.271 ms 234.139 ms

20 ae-3.ebr1.chicago2.level3.net (4.69.132.62) 258.012 ms 246.968 ms 247.146 ms

21 ae-6.ebr1.chicago1.level3.net (4.69.140.189) 247.789 ms ae-1-100.ebr2.chicago2.level3.net (4.69.132.114) 248.378 ms ae-6.ebr1.chicago1.level3.net (4.69.140.189) 247.606 ms

22 ae-5.ebr2.chicago1.level3.net (4.69.140.193) 255.883 ms 247.559 ms 251.463 ms

23 ae-8-8.car1.detroit1.level3.net (4.69.133.241) 253.563 ms 256.157 ms 252.912 ms

24 ae-11-11.car2.detroit1.level3.net (4.69.133.246) 215.656 ms 216.143 ms 216.210 ms

25 4.53.74.58 (4.53.74.58) 254.588 ms 254.412 ms 253.918 ms

26 208.86.248.2 (208.86.248.2) 256.133 ms 253.974 ms 254.540 ms

27 rb26.skylinesaustralia.com (208.69.126.31) 216.143 ms 216.245 ms 216.717 ms

Hosting the site here would cost around 5-10x as much here in AUS.

Being (roughly) its $1000/month to host in the US... So moving it to AUS is totally out of the question. Anywhere else for that matter thats closer, is more expensive.

We dont run to profit, infact the forum until recetly was at a loss :thumbsup:

I dont know about 8 seconds though, i dont think it takes me even half that.

Ill check when i get home (my work connection is slower than home, so useless using it as an example)

I mean, if everyone donated $10 a year, then we could host in AUS easily, but thats just how it is.

People have to live with the ad's slowing page loads, US server slowing page loads... and so on. Part and parcel of using the site that runs to exist and doesnt charge users to post/use.

When you think about it too, SAU offers pretty good img hosting for free... when infact its costing the forum.

So while we might be able to double this, increase that - it'll never happen without serious financial backing

I pressed reload on this topic after you posted your reply.

The screen went white immediately, but it took 12 seconds until the loading indicator stopped.

At about second 8 or so, the bulk of the page _after_ the "> skylines australia" navigation

strip had appeared to start reading and scrolling, so I'm not fussed about the last 4 seconds

but the first 7 seconds is the first block, the second block, the logged in block and the

navigation block all painfully loading then appearing... one by one...

I'm on an mac with latest firefox and adsl2. Another forum I use also located in the states, actually a

bit further away (in new jersey) takes just 2 seconds to display a topic that is longer than

a page. It doesn't use any third party ad components, so that helps, but just pointing that

it isn't purely a function of 220ms latency. It can be 2 or 3 seconds, 24 hours 7 days,

when tuned right.

I believe yahoo has an excellent tool for analyzing page delay, called YSlow, part of firebug

which is the firefox developer free addon.

You can load it up, and load a page and it'll show you where the biggest time consumers are

and give tips. You can also use the basic one built into firebug that shows all the latencies

and delays of each bit of the page..

I reckon half the delay is the result of third party ad services that are inefficient (not like google

adsense which is very efficient) and half is what seems like a slow host. Considering just how

mega powerful an average data center rent-a-box can be now, there isn't much excuse for that..

if you are paying $1000/mo you can get a box that includes more cpu and memory than you

can use.. you can for nothing install the best php accelerator, if thats the problem.. if it is

database, it can be tuned to be instant.. (they are often not, out of the box).

anyway without testing more deeply, these are guesses only. But if I were your SA I'd be

trying to aim for a faster page load. And this is nothing to do with the problem that led the

topic: obviously adyieldmanager isn't even near 100% reliable and does block rendering

perhaps because the html requires its javascript to be remotely loaded right up front.. (a common

problem with services like that).

FYI - I just loaded this topic, over 64kb connection (because I'm capped) in 6 seconds..

At the end of the day, its Christians site, he runs it the best way he see's fit and in the spare time that he has to do so :(

If Ad's are slowing it, so be it simply because the site would cease to exist if they were not there.

if you look at my original post, I wasn't complaining about the adds, or them slowing it down by a few seconds - I was commenting that I'm running into times when that site kills it entirely...

So what I was hopping was that you could have a fall back position, ie if Ad.yieldmanager.com doesn't respond in a certain time (say 10sec?) you bring the page up anyway...

Ian

  • 2 weeks later...
Also, and off this topic, it takes 8 seconds to fully render an average topic, 7 of those are waiting

and loading the first and second and third rows.. before getting to any meat you want to read..

most of this seems like CPU overload, but it doesn't help much that, for me, the site is a full 27 hops away..

a ridiculous trip from telstra thru to LA, down the coast, across the states, before ending up in detroit!

and about as awful a ping time as you can imagine, 220ms

Dude why care about 220ms out of 8 seconds? Its just under 3% contribution to the time...

Most of the internet is in the states... so think of it more as "normal" instead of "ridiculous" :-)

6pm = most users online (usually 6pm-8pm)

Cant be avoided far as i know unless, every user donates and we get a whiz bang server over here in AUS (as i said)

Nah just get a whiz bang server in the states - the 220ms is not worth worrying about.

I have been looking at AU hosting again lately, trying to weigh up the advantages as the exchange rate is hurting us atm.

I'm sorry the site seems slow at times, we are on three servers and there should definitely be no busy-period slow down. Especially as we are busy when the rest of the states' servers are idle so it wont be network related.

For the record, we are caching pages, we have a fairly tuned db (on a hefty box) and have done many other performance enhancements. Lately we removed our locally served ads also from the pages so there is less to load now.

Hope that clears this up a little bit.

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now


  • Similar Content

  • Latest Posts

    • If the gases flowing in those two tracts had the same properties, you could maybe use such broscience. But the exhaust has a different composition, different normal density, different actual density (because of different normal density, and mostly because of the massively higher temperature), and different viscosity (again because of much higher temperature). Consequently, all of the fluid dynamics parameters that matter, that you calculate from these inputs, such as the Reynolds number, friction factors (for wall friction) and so on, are all incomparable.
    • And we shall have to presume that Canada is the same?
    • In the US almost everything is E10. It can't exceed 10% by much or fuel systems have trouble adapting. At the same time because MTBE, MMT, and TEL are all banned they need as much ethanol in it as possible to boost octane.
    • I was mostly jesting. In my experience (and probably only my experience) the R34 GTT physical airbox space is actually too small to flow the amount of power it wanted. By sealing the box, I made it so it could only be fed by the ducts themselves. So you can seal it up and get nice cold air which IS good, but at a certain crossover point: More Hot Air > Less Cold Air I don't think you're at this point. In my case merely ducting the hot air intake with a very focused set of ducts counteracts the fact it's in a V8 engine bay. More cold air obviously best. The solution looks great.
    • Nah, the OEM CAI pipe is still installed behind the bumper, it is about 5" x 3" oval at the engine side, tapering down to a 3" pipe behind the bumper where it gets all the ambient air it needs Engine side of radiator support OEM intake pipe "oval hole" that is right in front of the filter My OEM NC1 CAI pipe: From NC2 onwards, below pic, they come slightly smaller at 2.75" diameter with corrugations and a resonance chamber to reduce intake noise, lucky for me my NC1 has the bigger noisy one, LOL   Basically, the "sealed" airbox will just get ambient air from a 3" pre filter intake tube that is the same size, 3" as the rest of the intake pipe post filter, and if a 3" intake isn't big enough to flow enough air for 150 killerwasps then there are other issues The whole intake is basically the same length as OEM, but it is now about 30% bigger from the airbox back through to the new intake plenum than OEM, and the intake plenum is port matched to the head And the intake is now about 30% bigger than my 2.5" exhaust, so the suck, squeeze, bang and blow black magic should be fine, well, to my uneducated understanding of fluid dynamics anyway Talking the the guys at MX5 Mania, it may even make a few more killerwasps as the intake isn't sucking hot air, especially off idle or when in slow traffic when it would be sucking hot air  As for the difference in IAT, I haven't logged IAT yet, as I don't currently have a OBD2 reader, but I will have a play with my thermal lazer thingie next time I take the car out to sèe how hot stuff gets under the bonnet near the intake filter prior to installing the air box, my "assumption is it has to be much better after the air box is in and sealed up compared to what it is now The aftermarket "performance" CAI elephant in the room: Aftermarket CAI intakes typically have the air filter tucked up behind the bumper, with a 2.5" intake tube (the OEM intake pipe is actually about 30% bigger than the fancy pants "aftermarket" version.....WTF), and you need to remove the bumper to service the filter, which is a PITA Like dis:    
×
×
  • Create New...