Author Topic: Forum speed issues (Read 13005 times)

EEVblog · « **on:** June 14, 2012, 04:38:22 am »

Apparently the forum speed in the last few days has been poor, and many requests not being accepted. I've seem this myself too.
Who else is seeing the issue?

I installed a wordpress plugin coincidentally around the same time, and have now just disabled this thinking that might be the cause (as it imports Youtube comments into wordpress every hour)
So if this seems to have have fixed it, please let me know.

Dave.

SeanB · « **Reply #1 on:** June 14, 2012, 04:41:11 am »

Had noticed the slowdown, with pages either not loading ( but not giving an error) unless refreshed, or loading slowly, such that I could see it rendering line by line.

Seems to be better now, thanks.

EEVblog · « **Reply #2 on:** June 14, 2012, 06:14:13 am »

My host is saying no issue.
I'm getting this ping screen grab.
Some issue with Telstra in Oz?

Dave.

amspire · « **Reply #3 on:** June 14, 2012, 06:32:52 am »

It is definitely getting much better. Pages are now loading quicker. Posting replies still a little slow. Perhaps it is just a whole backlog of junk in a TCP stack that just has to be cleaned out, and then it will be back to the way it was.

I noticed that closing my browser and reopening made the pages load much quicker, so it could be that half the problem was the TCP-IP stack on my machine trying to buffer the broken communications from this morning.

Hopefully if I reboot, everything will be back to normal.

Richard.

IanB · « **Reply #4 on:** June 14, 2012, 07:07:57 am »

A few hours ago I noticed that forum pages wouldn't load properly and were getting stuck when half rendered. But everything seems to be nice and snappy now.

EEVblog · « **Reply #5 on:** June 14, 2012, 07:40:38 am »

I've added a poll.
Have you had any forum issues in the last day or two?
And if you have, where are you from?
I suspect not everyone has been seeing it.

Thanks
Dave.

codeboy2k · « **Reply #6 on:** June 14, 2012, 09:08:22 am »

I had problems a few days back, mostly with posting previews being slow. Not any issues with page loads, those always seemed ok to me.

Today, posting previews seems snappy again.

EEVblog · « **Reply #7 on:** June 14, 2012, 09:15:48 am »

I got my server host to investigate and this is their response:

Quote

Hello,
I apologize for any inconvenience this may cause you, but you will need to contact your ISP in order to have this resolved. You can see that at hop 8 on the ping plotter results that there is 100% packet loss. Since this is out side of our network we are unable to resolve this for you. When contacting your ISP this Ping Plotter result will help them in solving this for you because it will let them know where the issue is occurring. If you have any other questions or comments please feel free to let us know.
Best regards,
Linux Systems Administrator
HostGator.com LLC
http://support.hostgator.com

I guess they don't get it's not just a problem for me, but lots of people.
So either it's say a problem only for us aussies using Telstra, or it's something else?

Dave.

joelby · « **Reply #8 on:** June 14, 2012, 10:05:21 am »

Quote from: EEVblog on June 14, 2012, 06:14:13 am

My host is saying no issue.
I'm getting this ping screen grab.
Some issue with Telstra in Oz?

If some hops are not responding to pings, but hops beyond them are responding without any loss, it usually just means that one router on the path isn't responding to pings or has a private (RFC1918) address. Nothing to be concerned about.

EEVblog · « **Reply #9 on:** June 14, 2012, 10:15:50 am »

Quote from: joelby on June 14, 2012, 10:05:21 am

Quote from: EEVblog on June 14, 2012, 06:14:13 am
My host is saying no issue.
I'm getting this ping screen grab.
Some issue with Telstra in Oz?

If some hops are not responding to pings, but hops beyond them are responding without any loss, it usually just means that one router on the path isn't responding to pings or has a private (RFC1918) address. Nothing to be concerned about.

I hope so. But it seems at least a quite a few people are having an issue, which would be the first time since moving to the dedicated server. So if it's just a rouge router, it must be limited to people geographically?

Dave.

Psi · « **Reply #10 on:** June 14, 2012, 10:34:39 am »

Yes, i'm definitely experiencing this issue.
Even right now as i post this message.

I swapped a router around at my place a few days ago and i had just assumed the problem was at my end, but after some testing i found it's only the eevblog pages, so my new router is unrelated.

What happens is i will click on a eevblog forum page and nothing will load.
Firefox recognizes i have clicked something but just sits there saying "Waiting for www.eevblog.com.."

If i open another window and try again the same thing happens. After anywhere from 5 seconds to 2 minutes later all tabs i have open to eevblog instantly load at the same time.

Massive packet loss at ten-0-3-0-120.bdr01.akl02.akl.VOCUS.net.au

Host Loss% Snt Last Avg Best Wrst StDev
1. trafficshaper.local.####.co.nz 0.0% 62 0.3 0.2 0.1 0.6 0.1
2. akl-ip21-xdsl-gw.wxnz.net 0.0% 62 10.1 9.8 8.5 19.9 1.5
3. as17435.cust.bdr01.akl02.akl.VOCUS.net.au 0.0% 62 20.1 14.3 7.9 44.7 9.7
4. ten-0-3-0-120.bdr01.akl02.akl.VOCUS.net.au 47.5% 62 8.3 8.8 8.0 9.9 0.4
5. ten-0-2-0-400.bdr01.akl01.akl.VOCUS.net.au 47.5% 62 160.8 161.0 160.0 162.1 0.5
6. ip-88.202.31.114.VOCUS.net.au 47.5% 62 9.3 9.6 8.5 12.4 0.7
7. ten-1-0-0.bdr01.akl04.akl.VOCUS.net.au 47.5% 62 160.9 161.0 160.1 163.5 0.6
8. te2-6.bbr01.cs01.lax01.networklayer.com.any2i 46.7% 61 168.7 170.0 168.0 205.6 6.5
9. ae19.bbr01.eq01.dal03.networklayer.com 48.3% 61 201.7 197.4 195.6 205.7 2.5
10. ae0.bbr01.sr02.hou02.networklayer.com 46.7% 61 221.7 202.6 200.1 225.0 5.5
11. ae5.dar02.sr02.hou02.networklayer.com 45.0% 61 204.8 204.9 203.1 226.5 3.9
12. po2.fcr01.sr02.hou02.networklayer.com 45.0% 61 204.2 205.4 203.5 221.8 3.6
13. 184.173.12.165-static.reverse.softlayer.com 45.0% 61 207.4 207.2 206.0 215.5 1.6

metalphreak · « **Reply #11 on:** June 14, 2012, 11:10:13 am »

http://pastebin.ca/2161155

I'm seeing a fair amount of packet loss at their router in Dallas (I assume from the rDNS).

I would say that is the issue at the moment. Site loads fast for me with a bit of lag sometimes, which is consistent with a small amount of packet loss.

This would affect everyone differently depending on what route your ISP has to the datacentre.

Psi has similar but far worse packet loss: http://pastebin.ca/2161156

The issue is definitely with ae19.bbr01.eq01.dal03.networklayer.com - 16.67% packet loss in your graphs Dave.

JuKu · « **Reply #12 on:** June 14, 2012, 11:24:18 am »

Very long delays. I'm in Finland.

baljemmett · « **Reply #13 on:** June 14, 2012, 11:27:20 am »

I've seen seeing a few half-loaded pages and pages stuck on 'waiting for eevblog.com' -- it seems to happen in bursts rather than randomly, e.g. I happily browsed through all the unread posts just now without a problem but then when I came to vote in this poll the page refused to load for several minutes, along with a couple of others. Then suddenly everything started working again.

When I first noticed it I put it down to some changes I've made on my network at home, but I'm in the office at the moment so it obviously isn't that -- unfortunately the people who run my office building block all pings etc. so I can't get a traceroute or check for packet loss :-/

Rerouter · « **Reply #14 on:** June 14, 2012, 11:32:50 am »

same here for me, lags heavily but only for seemingly random periods of time, i have had my adblocker block everything other than local to your forum scripts and images, and still seeing the issue (is the yahoo plugin new?)

also the download speed from the site is crippled, your ping image is transfering at less than 1KB/s yet the ping of the site isnt varying,

i would partly suspect something may be overloading your server,

metalphreak · « **Reply #15 on:** June 14, 2012, 11:47:48 am »

It's not his server. It's one of the upstream network connectivity suppliers to the datacentre that is having issues. I'm sure they are aware of the issue, and there isn't much HostGator can do about it if its outside their network.

ModemHead · « **Reply #16 on:** June 14, 2012, 12:25:55 pm »

I keep a tab open on this site and do a "show unread posts" fairly often while I'm near the computer. And I haven't seen a single problem in the past few days. Everything loads pretty much instantly. Ping times to the host are around 100-150ms. I'm in NC, USA.

Bored@Work · « **Reply #17 on:** June 14, 2012, 12:43:58 pm »

If it is an issue with traffic from Europe, RIPE has some massive rDNS problems the last few days. Is there anything in SM, or an SM plugin doing rDNS?

EEVblog · « **Reply #18 on:** June 14, 2012, 01:09:35 pm »

Quote from: Bored@Work on June 14, 2012, 12:43:58 pm

If it is an issue with traffic from Europe, RIPE has some massive rDNS problems the last few days. Is there anything in SM, or an SM plugin doing rDNS?

I have no idea what that means, but if so, then Wordpress must do the same things. Because I have the exact same issue with the blog site and the SM forum.

Had to refresh this page toget this edit box to load. Damn this is annoying...

Dave.

EEVblog · « **Reply #19 on:** June 14, 2012, 01:12:09 pm »

And I thought this was a different issue, but now I think it's related. My wordpress blog when it does work, sometimes throws up and old cache image not showing the latest posts. have never seen that happen before yesterday.

Dave.

EEVblog · « **Reply #20 on:** June 14, 2012, 01:24:07 pm »

And I thought this was a different issue, but now I think it's related. My wordpress blog when it does work, sometimes throws up and old cache image not showing the latest posts. have never seen that happen before yesterday.

Dave.

amspire · « **Reply #21 on:** June 14, 2012, 01:51:58 pm »

I suspect cache problems are somehow involved in the ongoing Eevblog Forum problems. If I switch to a different browser, the forum seems to work a lot better. I did clear my Firefox cache, but that did not solve the problem.

It could even be the cache of a downstream ISP that has a lot of bad packets stored. It could also be that if people did not us the forum much during the peak of the problems, then there may be no bad packets cached.

If it is a cache problem, then I would expect things to get back to normal in a day or two.

If the meantime, try using a different browser.

Richard.

caroper · « **Reply #22 on:** June 14, 2012, 02:43:00 pm »

Hardly usable from South Africa, it has taken 8 Hours to post this reply.
I have made several attempts but this is the first time all day that the Reply Editor window has opened.

Cheers
Chris

Zad · « **Reply #23 on:** June 14, 2012, 03:10:05 pm »

It seems to work for a burst of a few minutes then disappears for a fair time.

SeanB · « **Reply #24 on:** June 14, 2012, 03:23:34 pm »

Took 20 seconds to load this reply window, generally it is much faster ( 2 weeks ago or so) so definitely slower. However I am in SA and using a non Hellkom ISP, via SEACOM cable so that is generally a different route which has a direct link via a fibre loop to Australia, the other is via Telkom where it goes via EU or USA.

Chris, are you using Telkom Internet or a Seacom provider? The local loop is still the same to the edge router, but the Seacom links are noticeably lower latency and much less bursty.

metalphreak · « **Reply #25 on:** June 14, 2012, 03:31:45 pm »

From a UK based server
Packets Pings
Host Loss% Snt Last Avg Best Wrst StDev
1. ################## 0.0% 42 4.0 1.6 0.6 9.9 2.0
2. 87.117.211.198 0.0% 42 0.9 1.1 0.7 11.9 1.7
3. 365.net1.north.dc5.as20860.net 0.0% 42 0.8 0.7 0.5 1.1 0.2
4. 593.core1.thn.as20860.net 0.0% 42 1.4 6.7 1.3 212.2 32.5
5. 202.core1.hex.as20860.net 0.0% 42 1.6 2.1 1.4 16.7 2.3
6. bbr02.lon01.networklayer.com 0.0% 42 1.6 5.0 1.6 39.8 8.1
7. ae1.bbr02.tl01.nyc01.networklayer.com 0.0% 41 69.4 70.9 69.0 95.9 5.3
8. ae1.bbr01.eq01.chi01.networklayer.com 4.9% 41 118.3 95.0 90.9 118.3 7.5
9. ae20.bbr01.eq01.dal03.networklayer.com 0.0% 41 108.2 108.6 108.1 115.1 1.3
10. ae0.bbr01.sr02.hou02.networklayer.com 0.0% 41 112.7 113.1 112.5 118.9 1.3
11. ae5.dar01.sr02.hou02.networklayer.com 0.0% 41 114.0 115.5 113.6 135.3 4.7
12. po1.fcr01.sr02.hou02.networklayer.com 0.0% 41 114.4 119.2 113.9 156.2 11.4
13. 184.173.12.165-static.reverse.softlayer.com 0.0% 41 113.0 113.3 112.8 115.9 0.7

From a Western Australian Server

Packets Pings
Host Loss% Snt Last Avg Best Wrst StDev
1. ############### 0.0% 20 0.9 0.6 0.2 1.1 0.3
2. 218.100.54.109 0.0% 20 2.2 2.0 1.5 4.0 0.5
3. ge-0-0-501.bdr01.per01.wa.VOCUS.net.au 0.0% 20 1.8 6.1 1.6 87.7 19.2
4. ten-0-3-0-400.bdr02.per02.wa.VOCUS.net.au 0.0% 20 206.9 207.3 206.1 219.2 2.8
5. ge-0-1-6.cor02.syd03.nsw.VOCUS.net.au 0.0% 20 50.7 51.3 50.6 56.8 1.3
6. pos-1-2-2.bdr02.sjc01.ca.VOCUS.net.au 0.0% 20 207.0 206.3 205.7 207.0 0.4
7. te2-6.bbr01.cs01.lax01.networklayer.com.any2ix.coresite.com 5.0% 20 213.8 214.1 213.6 215.0 0.4
8. ae19.bbr01.eq01.dal03.networklayer.com 0.0% 20 242.3 242.9 241.6 257.8 3.6
9. ae0.bbr01.sr02.hou02.networklayer.com 0.0% 20 249.2 249.8 249.0 252.9 0.8
10. ae5.dar02.sr02.hou02.networklayer.com 0.0% 19 257.5 263.8 246.5 296.3 16.5
11. po2.fcr01.sr02.hou02.networklayer.com 0.0% 19 249.6 254.1 249.4 305.5 13.3
12. 184.173.12.165-static.reverse.softlayer.com 0.0% 19 247.3 251.6 246.8 294.0 11.5

Clearly an issue with softlayer's network here.

You should reply back to hostgator support with these as examples, to show that it's a widespread issue (and not limited to just you on Telstra).

I assume your wordpress site is on the same server. Your browser is probably caching bits and pieces when it can't load them properly (many browsers actually have multiple threads requesting page elements simultaneously, some but not all of these may be failing)

caroper · « **Reply #26 on:** June 14, 2012, 03:37:54 pm »

I am with Afrihost Sean so Internet Solutions -> Seacom.
When the problem started I assumed it was another of the many Seacom outages so was pleased to see I was not the only one with problems.

SeanB · « **Reply #27 on:** June 14, 2012, 03:45:50 pm »

SNAP.............If I try Hellkom I can go and get a cuppa with many sites, as well as many that do not load at all.

Neilm · « **Reply #28 on:** June 14, 2012, 06:49:30 pm »

Currently I am having no problems viewing or posting. Yesterday (13th), I had a lot of problems at about 21:00. I checked it wasn't my connection by Skypeing my brother who lives the other side of London and he reported it was slow as well.

Neil

alm · « **Reply #29 on:** June 14, 2012, 07:47:38 pm »

Quote from: metalphreak on June 14, 2012, 03:31:45 pm

7. ae1.bbr02.tl01.nyc01.networklayer.com 0.0% 41 69.4 70.9 69.0 95.9 5.3
8. ae1.bbr01.eq01.chi01.networklayer.com 4.9% 41 118.3 95.0 90.9 118.3 7.5
9. ae20.bbr01.eq01.dal03.networklayer.com 0.0% 41 108.2 108.6 108.1 115.1 1.3

6. pos-1-2-2.bdr02.sjc01.ca.VOCUS.net.au 0.0% 20 207.0 206.3 205.7 207.0 0.4
7. te2-6.bbr01.cs01.lax01.networklayer.com.any2ix.coresite.com 5.0% 20 213.8 214.1 213.6 215.0 0.4
8. ae19.bbr01.eq01.dal03.networklayer.com 0.0% 20 242.3 242.9 241.6 257.8 3.6

Clearly an issue with softlayer's network here.

Nah, the hops before and after it are fine. Just a busy router assigning a low priority to ICMP traffic destined to this router, as is common on core routers (who cares about core router ICMP traffic?).

amspire · « **Reply #30 on:** June 14, 2012, 09:10:49 pm »

The forum speed is returning to normal for me. I still think that downstream caches got junk in them after the problems of a day ago. This caused the requests that did get back to the forum to confuse the forum software totally, particularly with "reply" type pages.

As the junk in the caches is clearing, everything starts to work again. No real evidence, but that is my guess.

Whatever that plugin did that caused this problem, it was pretty bad. If it was my site, I would even change the admin forum and WordPress passwords in case the plugin included some malicious software.

Richard.

EEVblog · « **Reply #31 on:** June 15, 2012, 12:01:27 am »

It does seem better this morning, I haven't had a missed request yet.
It would be interesting to re-enable that wordpress plugin (it's not an unpopular one) and see what happens.

Dave.

amspire · « **Reply #32 on:** June 15, 2012, 12:44:20 am »

Quote from: EEVblog on June 15, 2012, 12:01:27 am

It does seem better this morning, I haven't had a missed request yet.
It would be interesting to re-enable that wordpress plugin (it's not an unpopular one) and see what happens.

Dave.

The fact it messed up the Simple Machine's forum which I gather is a totally separate program makes me wonder if part of the problem is that this Wordpress plugin was redirecting requests to a third party site. I think that you do a redirect to get from Wordpress to the SMF, and I think the Youtube comment Wordpress plugin was somehow interfering with this redirect. Somehow the connection was eventually made, but perhaps via another 3rd party address. The problem part of the code was possibly ending up in downstream caches, and not immediately timing out.

Otherwise, I am not sure how a Wordpress plugin affects a Simple Machine forum.

Edit: Today, if I go to www.eevblog.com/forum in Chrome, I end up with this in the address bar:

Quote

https://www.eevblog.com/forum/

Yesterday when the forum was really bad, and the pages were stalling, I got this in the address bar instead:

Quote

https://www.eevblog.com/forum/index.php?PHPSESSID=a864ac13390b3ada2a7ecb531b3394be;wwwRedirect

There is a PHP session ID attack that allows a hacker to gain access to a site:

http://www.acunetix.com/vulnerabilities/PHPSESSID-session-fixatio.htm

Richard.

EEVblog · « **Reply #33 on:** June 15, 2012, 01:37:43 am »

Quote from: amspire on June 15, 2012, 12:44:20 am

Edit: Today, if I go to www.eevblog.com/forum in Chrome, I end up with this in the address bar:
Quote
https://www.eevblog.com/forum/
Yesterday when the forum was really bad, and the pages were stalling, I got this in the address bar instead:
Quote
https://www.eevblog.com/forum/index.php?PHPSESSID=a864ac13390b3ada2a7ecb531b3394be;wwwRedirect

There is a PHP session ID attack that allows a hacker to gain access to a site:

http://www.acunetix.com/vulnerabilities/PHPSESSID-session-fixatio.htm

Hmm, interesting, I need to check this out.
The plugin is this one:
http://wordpress.org/extend/plugins/genki-youtube-comments/

Dave.

HackedFridgeMagnet · « **Reply #34 on:** June 15, 2012, 03:17:34 am »

The Session attack theory doesn't explain my failed direct http request to
https://www.eevblog.com/forum/Themes/default/css/index.css I also tried one of the javascript pages and failed.

trace route seemed ok from memory.

I was thinking at the time Network, Webserver low resources, or DDOS.

I'm glad it's better today.

EEVblog · « **Reply #35 on:** June 15, 2012, 03:28:05 am »

Quote from: HackedFridgeMagnet on June 15, 2012, 03:17:34 am

I was thinking at the time Network, Webserver low resources, or DDOS.

It's not my dedicated server, it's running well under capacity, and no spike in traffic at all.
I don't share this machine with anyone, it's all mine.

Dave.

HackedFridgeMagnet · « **Reply #36 on:** June 15, 2012, 03:44:00 am »

Quote

it's running well under capacity

Just checking,
I guess you would have checked the resources yesterday, when the problem was there.
available memory, cpu utilisation, hard disk swaps.
Maybe that plugin couldv'e sucked up resources for whatever reason.

vk6hdx · « **Reply #37 on:** June 15, 2012, 05:35:23 am »

Seems fine for me today Dave. Everything seems to be loading rather quickly.

Psi · « **Reply #38 on:** June 15, 2012, 05:41:46 am »

Yeah, i've not had any issues for all of today.

EEVblog · « **Reply #39 on:** June 15, 2012, 05:42:18 am »

Quote from: vk6hdx on June 15, 2012, 05:35:23 am

Seems fine for me today Dave. Everything seems to be loading rather quickly.

Yep, me too.
Problem fixed I guess.

Dave.

Joshua · « **Reply #40 on:** June 15, 2012, 03:17:09 pm »

About an hour ago, I also received the url that ampsire did (the php attack thing). Just now I received https://www.eevblog.com/forum/index.php?wwwRedirect

I've never seen the wwwRedirect at the end before.

Joshua


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Poll

Have you seen any forum speed/access issues in the last few days? (slow load, freezing etc)

Author Topic: Forum speed issues (Read 13005 times)

alm

Share me