Kitz Forum
Announcements => Site & Forum Discussion => Topic started by: Weaver on January 11, 2016, 05:56:33 PM
-
Site performance seems very slow this afternoon, a click on a link seems to have a long delay before it takes effect.
-
Perhaps it's just me.
-
No, I had the same earlier. Seems ok now though.
-
Spoke too soon. :(
-
I, too, have had a sluggish afternoon and it is still continuing into the evening. :(
Other sites load promptly and behave normally. I wonder if Eric could go and check "the meters", please? :-\
-
I have checked other sites with the same web browser, as a sanity check. The response time to clicks makes things quite difficult for me using an iPad.
-
I'm getting response times to "mouse" clicks of two to nearly five seconds.
-
Site performance seems very slow this afternoon, a click on a link seems to have a long delay before it takes effect.
Glad you started this, Weaver ....... I wondered if it was just me ??
-
This does seem to happen from time to time. It's most likely to be a routing issue, over which Kitz has no control. Be patient, and it will be sorted.
@b*cat: The only 'meters' I'm able to check don't show anything to be wrong.
-
@b*cat: The only 'meters' I'm able to check don't show anything to be wrong.
Thank you for looking.
-
We will endure patiently, Roseway. Thank you for looking into it.
-
Well I had the same with Google this morning, it was so slow that I used Bing which loaded instantly.
-
The same with me this afternoon - especially when previewing or posting on the PC. Working fine now via a tablet.
-
Now also working well for me, with no change of hardware. :)
-
I noticed it too, it appeared to stall slightly on reading code.jquery.com but when I ran firebug on the forum it said the majority of time was spent on the html page so if it persists kitz may have to look deeper.
-
Server stats & database all look good. No issues with the forum logs which look pretty error free [tapatalk dependencies may have had some issues on the 4th but very minor] Although not much use looking now as load is a real time thing, but the load seems perfectly fine and there is plenty of processing power and RAM to spare.
reading code.jquery.com
There are a few scripts that need external dependencies.. ie the cookie legislation script that by EU law I have to display. (which is what the jquery one relates to). If those are slow or if there is a routing issue, then that will slow down the loading of html pages. I switched the script last year and where the dependency was previously hosted from because that was causing slowness unrelated to the site. The other dependencies I use are google ad related.
Thanks for letting me know, I'll try keep an eye on things.
-
wonder if the bbc ddos attacks have affected other sites, apparently the attack was/is huge.
-
I dont know for certain, but based on what ronski said, I too would guess it may be google related as rather than the previous geo-location dependency which was slowing down lots of sites (I believe MDWS had a similar issue) I started using the google hosted one*. With LL mentioning the jquery.com stall, then that would perhaps suggest to the google cookie legislation script which I use.
I too have noticed issues with google searches over the past few days where it becomes unresponsive or slow... but as I had been using a mobile device (over wifi) I didnt bother looking into it. I certainly dont have any BBC dependencies, and not sure how or if that could affect any routing. Could it cause issues at some peering?
*Blame EU legislation for these sorts of scripts which are the bane of EU sites that comply with cookie compliance laws and which have become a bit of a headache at times. As website owners you are supposed to check if the visitor is from within the EU, then if they are ensure that the relevant cookie script is displayed.
-
With LL mentioning the jquery.com stall, then that would perhaps suggest to the google cookie legislation script which I use.
The stall as displayed in firebug was minor, the html page itself spent over 600ms in waiting state, the jquery one was 50-60ms, which was still a lot greater than anything else.
-
I'm really not sure what caused it :(
The only other thing I have noticed of late is the google ad displayed which says "APNIC. Thank you for helping us measure the internet".
Im not too concerned with it being APNIC but wondered what it was all about. I did a bit of searching and it would appear APINIC are using google ads on some websites to measure the reach & penetration of IPv6 addresses on the internet.
PS they also say in the documentation.
IPv6 measurement
If you see the advert
PLEASE DON’T CLICK ON IT
(it costs us more)
-
Google Maps was dreadfully slow this afternoon, but I've not noticed any problems with the Kitz forum, or any other sites, just Google.
-
I dont know for certain, but based on what ronski said, I too would guess it may be google related as rather than the previous geo-location dependency which was slowing down lots of sites (I believe MDWS had a similar issue) I started using the google hosted one*. With LL mentioning the jquery.com stall, then that would perhaps suggest to the google cookie legislation script which I use.
I too have noticed issues with google searches over the past few days where it becomes unresponsive or slow... but as I had been using a mobile device (over wifi) I didnt bother looking into it. I certainly dont have any BBC dependencies, and not sure how or if that could affect any routing. Could it cause issues at some peering?
*Blame EU legislation for these sorts of scripts which are the bane of EU sites that comply with cookie compliance laws and which have become a bit of a headache at times. As website owners you are supposed to check if the visitor is from within the EU, then if they are ensure that the relevant cookie script is displayed.
Thats what I meant, the attack may have been big enough to saturate some transit links, but who knows.
You cannot run that script locally? Running scripts on a remote source is bad for security and performance.
-
Its the jquery DOM. Its recommended that you dont host it locally and use an eternal source.
http://encosia.com/3-reasons-why-you-should-let-google-host-jquery-for-you/
Although the code to run the script is on the page itself, the libraries have to also be loaded. I cant recall offhand now but Im pretty sure I refer to google for jquery with the fallback to jquery.com by default. Seeing LL's mention of a call to jquery.com is what made me think it is going to be one of the google scripts.. most likely the cookie one.
I dont host the jquery libraries locally as I dont want to have to keep up with the various versions myself.
-
It took me a good minute to 'post' this afternoon and pages where very slow to load.
-
dont agree with encosia personally, I think their only partially valid point is the caching.
Modern browsers allow more than 2 connections per hostname they all decided to ignore the RFC now, not only that but keepalive (default in all modern browsers) will allow objects from the same hostname to be sent over one connection, so obviously this is a time and resource saver as it means no extra dns lookups, no new connections etc.
I understand the point you made regarding jquery libraries, but I guess if its slowing down your site too much it may need a rethink.
Of course we dont know if this is the reason, personally I have not had any lag on here for many weeks. The lag stopped for me when someone fixed their sig which was causing me hangs. :p
I suggest next time someone see's lag, is to refresh the page with developer mode enabled in the browser and that will pinpoint which request is slow.
-
It took me a good minute to 'post' this afternoon and pages where very slow to load
I was around and using the site most of yesterday afternoon and evening: ie from 2pm onwards and up until quite late, and I didnt see anything or experience any lag. Anyone else have probs yesterday afternoon and if so which ISP? Trying to find out if it could be routing.
-
For me, it was ok on Tuesday. I was using the site in the afternoon.
-
Nothing was noticed by me, yesterday, Tuesday 12th January.
(Like as Weaver mentioned, it was the afternoon and evening of Monday 11th January when I was experiencing the sluggishness.)
-
No problems for me on Tuesday, either.
-
Just been incredibly sluggish .............. 19:30 on 14/01/16
-
It's ok now but was very slow at lunch time today, I'm with Eclipse.
-
I wasnt around at 07:30 so cant comment for that time.. but I was around at lunch time from about just gone 12 to about 2:30ish and it was ok here then, so perhaps looking like an isp/routing issue?
-
I use FireFox to view is there any info I can capture next time it's slow which for me is often :'(.
-
I looked in at various times through the day, and didn't experience any problem.
-
I was a very late arrival this afternoon and I have not experienced any problems today (Thursday 14th January).
-
I use FireFox to view is there any info I can capture next time it's slow which for me is often :'(.
Just a thought. This is something Ive mentioned to a few people over the past couple of weeks, but firefox has become increasingly slow for me and Im even considering permanently swapping over to another browser because of this. Im getting quite used to it now crashing every other day.. and it has become sluggish and slow. There is definitely some sort of memory issue with it which I have noticed will cause slowness. Its not unusual for it to be taking anything up to 4GB of my RAM and Ive even seen it at 6GB once. Its currently at 1.8GB of RAM for me, and thats with me force crashing it from task manager just this morning. Once it consumes about 2-3GB it starts slowing down everything I am doing.
-
Hmm . . .
It is FirefoxESR 38.5.0 that is packaged with RHEL-6. Checking the RAM used, I see it is only 509.7 MB
-
On my system, Pale Moon (spinoff of Firefox 25.x) is currently taking less than 300 MB. I wonder whether some addons are responsible for excessive memory use?
-
I use FireFox to view is there any info I can capture next time it's slow which for me is often :'(.
Just a thought. This is something Ive mentioned to a few people over the past couple of weeks, but firefox has become increasingly slow for me and Im even considering permanently swapping over to another browser because of this. Im getting quite used to it now crashing every other day.. and it has become sluggish and slow. There is definitely some sort of memory issue with it which I have noticed will cause slowness. Its not unusual for it to be taking anything up to 4GB of my RAM and Ive even seen it at 6GB once. Its currently at 1.8GB of RAM for me, and thats with me force crashing it from task manager just this morning. Once it consumes about 2-3GB it starts slowing down everything I am doing.
yeah its broken in my view.
To fix crashes using a 64bit build is probably enough.
Problem is 64bit builds dont fix the slowdowns and cycle collector hangs.
Chrome is not only faster but its also way more stable, if for some reason a tab will crash it wont crash the entire browser either. It does slow down if you have many tabs open like me but only a small amount, not severe like firefox.
Firefox E10 can be enabled even in stable builds now but some extensions break including noscript and it seems to be implemented badly as it doesnt yield the speed gains I would expect, it makes startups way faster and reduces hangs a lot, but general sluggyness remains. I think the core problem is down to aggressive heavy cycle and garbage collection code in firefox but the dev's stubbornly wont add an option to disable it.
Opera is probably the best port of chrome if you want a port not made by google. Opera is 32bit only tho. Also I think none of the ports have the DRM codecs needed for things like netflix, only chrome has it (not 100% sure).
Finally a lot of privacy related stuff can be disabled using command line options, here is my launch batch file I use to move temp files to my ramdisk and adjust lots of security and performance settings.
start chrome.exe --cipher-suite-blacklist=0x009c,0xc009,0xc00a,0xcc15,0x009e --enable-ssl-false-start --disable-3d-apis --disable-backing-store-limit --disable-breakpad --disable-client-side-phishing-detection --disable-cloud-import --disable-direct-write --disable-directwrite-for-ui --disable-java --disable-password-generation --disable-preconnect --enable-remote-fonts --disable-webgl --disable-webrtc --disk-cache-dir=R:/CHROME2 --disk-cache-size=524288000 --dns-prefetch-disable --enable-async-dns --disable-potentially-annoying-security-features --disableOFF-precise-memory-info --disable-strict-mixed-content-checking --enable-strict-powerful-feature-restrictions --enable-smooth-scrolling --enable-tcp-fastopen --processOFF-per-tab --disable-background-mode --no-pings --purge-memory-button --enable-offline-auto-reload-visible-only --disable-sync --disable-offline-pages
-
The kitz forum went down for a few minutes 14/1/2016 just after 8pm noticed this while reading a forum page and then selected the next page this then showed up on the browers as a infinite refresh and then page timed out error.
-
I use the latest Linux Firefox 43.0.4 and it's using about 571MB and I do not have any problems with the other Tabs I have, it does not tend to crash.
-
Had more delay at 14.15 today.
Not sure if below is any help.
Tried refreshing page and did same lots of times.
-
Ive just sat here for about the past 45 mins or so trying to replicate this, but unfortunately I really cant.
Even clearing cache and pulling everything afresh to get the longest load time possible results in the following results (image below) from here, which is quite a bit further north than you.
From your cap, the longest items to call are the Paypal logo (which is hosted on my server) and something which I find rather strange is that tabmenu_left load quite quickly, yet there is a noticeable delay for the tabmenu_right. Im wondering if there is some packet loss going on?
Does anyone else have any suggestions please?
*I cant seem to log into MDWS atm to check your line stats.
-
Hi
You could try pingdom for an external load test and there's another external website testing site but I just cannot remember it's URL right now
This should give you a better idea, of any issues and what potentially maybe delaying your site
If you suspect packet loss, you could use pathping from your PC to site and from server (if you have direct access to server) to see if any packets are lost and where there lost
If it's shared hosting, your provider should have a status page or issue page to have a quick check on
I hope that helps a little and sorry if I'm wrong
Many thanks
John
-
All seems ok for me from here so I didnt previously attempt to load externally. However Ive just completed an external test from New York NY
Test performed from: New York, NY
Test performed at: 2016-01-20 15:57:25 (GMT +00:00)
Resolved As: xx.xx
Status: OK
Response Time: 0.946 sec
DNS: 0.089 sec
Connect: 0.091 sec
I also external monitor using pingdom which should sent me an alert if there is an issue.
Im not on shared hosting, Its my own server with 24/7 managed support. I have WHM and cPanel which show me server status. Server load, memory, disk etc all good and well within what they should be.
Waiting the results of a more intensive test from gtmetrix Canada.
- Results are good except a couple of suggestions.. some of which arent applicable (such as use CDN or make fewer http requests) . The only improvement it suggests is using gzip compression. 3.5 seconds for full page load to Canada from the UK is pretty damn good.
-
I'm using Chrome and haven't been experiencing any issues tbh? :)
-
Hi
It sounds as though it's a local issue to user then, perhaps an app or plugin causing slowness in browser, might even be antivirus
We use pingdom on 1 minute test and interworx control panels. I might suggest you enable compression but cdn is your choice. We use cloudflare and partnered with them, as we have with interworx
Many thanks
John
-
If you suspect packet loss, you could use pathping from your PC to site and from server (if you have direct access to server) to see if any packets are lost and where there lost
Just noticed this suggestion - thank you.
To clarify, I'm not seeing any packet loss from here and others who responded in the thread on the same ISP as me weren't.. neither are a couple of TT users... hence why at first we wondered if it was a routing issue.
However looking at mikes latest firebug attachment appeared to show that routing was okish, he was reaching my server and it was responding quite quickly for the majority of the html assets.
Re the packet loss suggestion, I just thought it a little strange that tabmenu_left loaded quite quickly, yet there is a noticeable delay for the tabmenu_right, the fact that the majority of assets loaded within a reasonable time made me wonder why just a couple of very small images were being delayed. If for example Mike had a CRC which would cause packet loss, then that could possibly explain it.
Im clutching at straws not knowing why it seems ok for some of us, but Mike is having recurring problems. :(
-
1.
Seems a bit better than lunch time.
Just tried Linux 'Chromium' and it's the same.
Data from FireFox>Tools>Web Developer>Network, Data
-
2.
-
Hi
It sounds as though it's a local issue to user then, perhaps an app or plugin causing slowness in browser, might even be antivirus
We use pingdom on 1 minute test and interworx control panels. I might suggest you enable compression but cdn is your choice. We use cloudflare and partnered with them, as we have with interworx
Many thanks
John
Your post and mine above crossed
Warning - while you were typing a new reply has been posted. You may wish to review your post.
Yep I think it must something local or ISP specific.
Re gzip.. I guess I should.
Re CDN.. The forum is SMF but Ive not used CDN for the main site as there are a lot of pages that wouldnt work within a simple CDN environment because there is so much code (ie the adsl checker is huuuuuge and although you wouldnt realise it is actually 3 separate pages that are displayed depending on which stage you are at). Because so many pages contain written & custom php code, its far easier to write using a template system and site manager that does nice things like colour code and auto indent etc. Ive many times looked at CDN but it would probably need to be something like drupal... which in itself I believe is a learning curve.. and transfer over take a large time investment.
-
1.
Seems a bit better than lunch time.
Just tried Linux 'Chromium' and it's the same.
Data from FireFox>Tools>Web Developer>Network, Data
So again it is mostly ok and responding quite quickly. Just the odd image causing a delay.
What is polling.bbc.co.uk. Any idea? I dont think thats anything to do with me or the site.
-
Sorry I was going to put 7 + 8 are on another 'Tab' for the BBC.co.uk site for a comparison.
-
kitz your working theory is a routing issue, so my next suggestion is for those with issues to report which isp they using and maybe you to add a reverse traceroute on your site so people can paste the trace from kitz to isp at the time of their troubles.
Still no issues here, my isp is sky.
-
Well I just had a hang.
Was posting in my dentist thread, and the post took about a minute to submit, the browser was reporting waiting for forum.kitz.co.uk so not a third party domain. As I was waiting for a POST not GET it wasnt a good idea for me to enable developer mode right at that moment sorry.
-
kitz your working theory is a routing issue, so my next suggestion is for those with issues to report which isp they using and maybe you to add a reverse traceroute on your site so people can paste the trace from kitz to isp at the time of their troubles.
Still no issues here, my isp is sky.
Im not sure now if it is a routing issue, tickmikes attachments appear to show that he is reaching the server in a reasonable time. The odd thing was most of the images came back quick enough, but seemed to be a delay on the odd one or two which is why I wondered about packet loss.
I'm at a loss to know what it could be. The server should be good and all the monitors show no problems. The initial reports in the thread could possibly have been related to the jquery libraries, but the rest I have no idea.
-
by the way my hang occurred during a short period I had enabled http scanning in nod32, that started causing some issues on other sites, so as such I cannot rule that out as the culprit, it is now disabled again.
-
got a screengrab for you kitz, the site took about 20 seconds to load in a new, then I refreshed took another 20 secs, during the second 20 secs I did a traceroute which looked fine, I will paste it, then I enabled developer mode and did some more browsing with cache disabled, there was slow requests but they first party, which means an issue of some kind on your server or config :(
Here is the data.
Tracing route to forum.kitz.co.uk [185.24.98.37]
over a maximum of 30 hops:
1 <1 ms <1 ms <1 ms home.gateway [192.168.1.253]
2 * * * Request timed out.
3 11 ms 11 ms 11 ms 02780842.bb.sky.com [2.120.8.66]
4 10 ms 10 ms 11 ms uk.slo.ld5.201.r1.misp.co.uk [195.66.237.228]
5 11 ms 11 ms 11 ms switch-004.sl5.misp.co.uk [185.52.26.251]
6 9 ms 9 ms 9 ms kitz.servers.eqx.misp.co.uk [185.24.98.37]
Trace complete.
-
I had Firebug running and noticed it was taking about 10 secs for the forum to respond (FTTC topics listing). So did a refresh.
The attached is a screen grab, just two 45ms images missing at the bottom. Normal overall response is about 500ms..
-
got a screengrab for you kitz, the site took about 20 seconds to load in a new, then I refreshed took another 20 secs, during the second 20 secs I did a traceroute which looked fine, I will paste it, then I enabled developer mode and did some more browsing with cache disabled, there was slow requests but they first party, which means an issue of some kind on your server or config :(
Here is the data.
Tracing route to forum.kitz.co.uk [185.24.98.37]
over a maximum of 30 hops:
1 <1 ms <1 ms <1 ms home.gateway [192.168.1.253]
2 * * * Request timed out.
3 11 ms 11 ms 11 ms 02780842.bb.sky.com [2.120.8.66]
4 10 ms 10 ms 11 ms uk.slo.ld5.201.r1.misp.co.uk [195.66.237.228]
5 11 ms 11 ms 11 ms switch-004.sl5.misp.co.uk [185.52.26.251]
6 9 ms 9 ms 9 ms kitz.servers.eqx.misp.co.uk [185.24.98.37]
Trace complete.
Thanks for that chrys.
If Im looking at those results correctly (which I may not be), but is the failure with show_ads.js
In which case that is the googlead script which is called from pagead2.googlesyndication.com
The longest item other than that was 75ms, which is 0.075 seconds.
-
I had Firebug running and noticed it was taking about 10 secs for the forum to respond (FTTC topics listing). So did a refresh.
The attached is a screen grab, just two 45ms images missing at the bottom. Normal overall response is about 500ms..
Thank you but cant see anything in there that identifies a problem.
I've a feeling Im missing something here about overall response time 500ms - where has that figure come from? :-\
-
the show ads is due to me blocking google syndicate as that hostname is constant lag in browsing.
Check topicseen.html, its over 5 seconds for the request.
When there is no visible pause the page loads between 500-600ms like tony said, personally I actually think thats slow unless its quite low spec hardware, so I think you have a database problem, perhaps either in corrupt data or an optimisation problem.
-
I had Firebug running and noticed it was taking about 10 secs for the forum to respond (FTTC topics listing). So did a refresh.
The attached is a screen grab, just two 45ms images missing at the bottom. Normal overall response is about 500ms..
Thank you but cant see anything in there that identifies a problem.
I've a feeling Im missing something here about overall response time 500ms - where has that figure come from? :-\
No it just tells you the server took a long time to respond, not why... The 500ms is the normal response time for that page to finish loading rather than 11000ms+ here.
Edit:
Forgot I have all Google related links blocked on the site so if that was a problem then it wouldn't show in the listing.
-
Hang on Ive just spotted something. I didnt check the SMF forum logs because on previous occasions there has never been nothing to show. Eric can check these logs too and will verify that on all previous occasions there hasnt been anything
However at around 6:30 this evening there was a problem. I can also see exactly who had problems with some slowness (ie BaldEagle had one slow load at 06:31:15 PM).
The problem is something to do with tapatalk (mobiquo), although why this affected non tapatalk users too I have no idea. Tapatalk is an SMF extension for mobile users.
According to the logs it should have cleared soon after and didnt last very long. Does 6:30 is tie in when you saw this?
Below is the SMF error log, showing just a small sample of users who experienced an issue at about this time.
Normally the only things that show in there are banned users or people trying to hack into admin.
-
Can't say I saw anything at 18:30 as was having dinner - the forum is on automatic refresh anyway. The screen grab I did was at 19:05. Don't know anything about Tapatalk I'm afraid.
-
Do you have a tapatalk account?
I can see from the logs that you experienced several issues just after 6:30 all related to tapatalk mobiquo files and topicseen.html
tbailey2 Today at 06:31:08 PM
http://forum.kitz.co.uk/index.php?/topic,16825.0/topicseen.html
8: Undefined offset: 1
File: /home/kitzco/public_html/forum/mobiquo/lib/classTTConnection.php
Line: 680
Ive no idea how tapatalk works, but what it does do is try and sync your read posts with any posts you also read whilst on the PC.
Saying that there were also a lot of requests for unregistered users seeing the same issue, I'm guessing (and it is only a guess) that topicseen.html may be checking for any messages read via tapatalk before loading each and every thread.
What the log above appears to be telling me is that tonight some users - in fact tbailey2, chrysalis & BaldEagle were the only reg users experiencing a problem - all the others were just IPs.
-
Do you have a tapatalk account?
As I said in my previous post, I don't know anything about Tapatalk - or didn't until you just explained it ;)
-
Ive posted a support question to tapatalk to see if I get any answers, it looks like when I installed tapatalk:
Everytime a user requests a page then it checks if its been read or not, part of this process is checking and syncing with tapatalk for SMF read posts.
ClassTTConnection.php (where it opens and communicates with the tapatalk servers) is where tonights errors have been occurring. Tapatalk becomes quite tightly integrated into SMF when installed and why it probably looked like the issue was at this end.
I stress that this problem is only tonight and it is unrelated to previous reports.
Eric can view the SMF logs and could confirm that on previous occasions nothing showed in the logs.. and that the issue with TTConnection only started at around 6:30pm this evening.
-
Do you have a tapatalk account?
As I said in my previous post, I don't know anything about Tapatalk - or didn't until you just explained it ;)
Thanks tony, I wanted to confirm if you had a tapatalk account or not.
One of the questions Ive asked is why TTconnection should also affect non tapatalk users
-
Eric can view the SMF logs and could confirm that on previous occasions nothing showed in the logs
That's correct. I've looked on several occasions in the past and never seen anything amiss.
-
As said in PM mine was around the 6.30 time period. I can get the exact time by checking the post I made before replying to this thread, I will edit this post with the time I found.
Also offered you some advice in the PM.
Actually I think I was just reading posts as the post I did here http://forum.kitz.co.uk/index.php/topic,16818.msg309976.html#msg309976 was after the time on my screenshot.
-
Hi
I use tapatalk and only login to make a reply or start a new thread
I did not notice any slowness in tapatalk but if you remember, I have previously asked about tapatalk displaying last message posted in a thread. It still does not do this but more importantly, unless I login first, it does not show unread posts, it shows all as read.
I also thought of a few ideas, like chrysalis, possible db issue, but more in terms of queries queued.
I am not sure if this a vps or full dedicated, but if vps and everything looks good, it could be a neighbor on the server using resource perhaps - guess work as you all know the setup/server, so sorry if I'm wrong
Lastly, I did a few tests to Kitz.co.uk, which tested on various connection speed from London test server, 5mb connection showed average load 8 sec and 20mb showed just under 3 sec.
I hope that helps and please ignore my post if I'm wrong sorry, I don't mean to introduce red Herrings
Many thanks
John
-
Thanks.
Im going to install the latest version of tapatalk. I dont usually like updating it that often because its manual and nine times out of ten something usually goes wrong :( SMF for some reason stopped approving the plugin and it can no longer be installed via the package manager. :(
One of the changelog updates since my last is this
solved issue that cause in forum view not showing correct unread items
Anyway I'll get everything together then take the forums offline to update.
-
Ive updated to the latest tapatalk and whilst offline I also undertook some forum maintenance.
The database tables have been optimised
Ive also cleared out a pile of old logs and pruned some posts.
I ran a check to see if there were any issues with any of the SMF files which came back "Congratulations, no errors found"
The main site is busier than the forums as thats what gets the most hits and there are several other databases - the adslchecker is probably one of the most busy pages, and Ive temp turned off a tracker on there. (I was just logging for excess access from certain IPs - such as BT CS running several checks per minute).
The front page calls from the forum, but some of the static pages such as this should always run quite quickly http://www.kitz.co.uk/adsl/
Ive no doubt in my mind that yesterdays issue @ 6:30pm was caused in some way by the tapatalk plugin and was unrelated to mikes issue.
-
ok fingers crossed its resolved, but in case it isnt I have sent you more info in PM.