Jump to content
superfuture

BUG 001 - SPEED ISSUE

Recommended Posts

This is affecting every user i know on Superfuture.

Here are some old school users complaining about it in a shoe thread:

http://supertalk.superfuture.com/index.php?/topic/13199-shoes-that-look-better-with-age/page__st__28350

Share this post


Link to post
Share on other sites

- i am having speed issues, the site is either super slow or super erratic,

- this started after the most recent update

- happening across the whole site

- using google chrome

- not really dependent on time of day seems slow or erratic whenever I get on

- specifically it takes ages to post/check rep/and for the home page to load

Share this post


Link to post
Share on other sites

i find it okay browsing, but slow as fuck posting. if it doesn't time out, i usually have to stop/refresh before just giving up entirely.

Share this post


Link to post
Share on other sites

edit: must have been just me. Errything good now.

Edited by Shoreman

Share this post


Link to post
Share on other sites

while examining apache processes noticed there was some weird code in a gallery album name which looked a bit suspect. just deleted it and also temporarily turned off gallery.

my connection seems to have just got a lot faster [as a result??] it might just be my imagination [distorted wishful thinking] but did anyone else notice any change?

Share this post


Link to post
Share on other sites

yep it certainly seems faster for me today.....still not back to normal (or what I remember normal speed being)

Share this post


Link to post
Share on other sites

hmm. that's slightly interesting. the server load just dropped to a figure i haven't seen in weeks -

-> 0.70

last pid: 96448; load averages: 0.70, 2.09, 2.72 up 228+11:01:16 01:56:02

though now bouncing back up to around 3.

anyone else notice any change?

+++++++++++++++++++++++

actually if any supergeeks out there who want to translate the following for all the ungeek supertalkers it would be great...

Server Version: Apache/2.2.21 (FreeBSD) mod_ssl/2.2.21 OpenSSL/0.9.8k DAV/2 PHP/5.2.17 with Suhosin-Patch

Server Built: Jan 6 2012 22:58:58 Current Time: Thursday, 23-Aug-2012 02:05:03 UTC Restart Time: Tuesday, 21-Aug-2012 05:44:18 UTC Parent Server Generation: 0 Server uptime: 1 day 20 hours 20 minutes 45 seconds Total accesses: 3093752 - Total Traffic: 18.4 GB CPU Usage: u139.523 s25.3828 cu0 cs0 - .103% CPU load 19.4 requests/sec - 120.8 kB/second - 6.2 kB/request 34 requests currently being processed, 17 idle workers

RC.._KKC.__...C._._WKW..K..K_.RR._KRR.._R.KRR_K.R.W__R_..R_R__RK

.._._R.CK..KK....K..............................................

Scoreboard Key:

"_" Waiting for Connection, "S" Starting up, "R" Reading Request,

"W" Sending Reply, "K" Keepalive (read), "D" DNS Lookup,

"C" Closing connection, "L" Logging, "G" Gracefully finishing,

"I" Idle cleanup of worker, "." Open slot with no current process

Share this post


Link to post
Share on other sites

so frustrating.

it is still reasonably fast browsing for me - but it's the fact that it's so erratic for many other users [when posting, on big threads, checking rep].

all i know is we have tested and checked and re-tested and re-checked, moved servers again, installed sphinx, updated skins, and got so many consultant opinions and opened so many tech tickets, and made so many tweaks, that everyone involved is just exhausted. it's been yet another endless tech nightmare.

therefore...

hosting and server change is looking like the next option.

--> that's what has been recommended by forum software makers.

--> however our server managers are not convinced it will make a difference and suspect it is still something in the code.

...so it is not a completely clear cut case. this is the issue we are dealing with for the past month or so.

who do we believe?

moving hosting is not something we really want to do, as doing ANYTHING with a website of this complexity can open a can of worms [as we have seen so many times before]. but unless i can be convinced otherwise we probably don't have a choice, and with a bit of luck it might actually fix the speed issue like the forum software makers are expecting?

to move hosting requires a good deal of logistical coordination and we also need it to coincide with billing cycles.

if we go ahead it will probably be in the first or second week of september.

after that, we need to look at archiving the big threads as a further measure.

anyway... yeh.

Share this post


Link to post
Share on other sites

its superfuckingslow like rightthefuck now. do it do it move hosting

Share this post


Link to post
Share on other sites

Extremely frustrating using any part of Supertalk right now due to speed issues.

Share this post


Link to post
Share on other sites

not nearly as frustrating as trying to conclusively diagnose and fix this bug...

since we made the post above last friday, there has been yet another round of 50 or so emails flying around between superfuture tech managers, the database server managers, and the forum software managers. [and that's on top of the 500 or so over the past few months...].

in case you can't be bothered reading this entire post the conclusion is that we need to archive the big threads.

here is a brief summary of the diagnostics to date:

+++++++++++++++++++++++++++++++++++++++++++++

earlier observations from superfuture tech managers:

+++++++++++++++++++++++++++++++++++++++++++++

Following are my observations:

· Re images – I think it is unlikely this is where the issue is. While generating re-sized images may take a little CPU time, they are only re-sized once per image (per size) and are usually processed pretty quickly.

· The server status page is most useful for identifying long running requests – from what I can see most of the long running requests (high number in the Req or SS column) are from the supertalk side of things.

· Periods of high CPU usage on your server (when logged in via SSH) don’t necessarily seem to reflect high CPU usage by the Apache server

· Queries are running at extremely variable speed – for example, one query running on your home page takes anywhere from 100ms to 5000ms

· A lot of queries are being run – since the database server last rebooted over 550 million queries have been executed. Dividing this number by the number of page views in this period (around 17 million) gives around 32 queries per page view. This seems like a lot – as a comparison superfuture.com (excl supertalk) averages around 4 or 5 queries per page.

........... I would suggest the issue lies with your database server. It may be that it is not configured to handle a high volume of requests, particularly simultaneous requests. This to me explains most of the above and would explain periods of poor performance.

If you go to phpmyadmin and click on the ‘Status’ tab at the top of the page, you can scroll down to see where the possible issues are (they are highlighted in red). It may be worth getting a MySQL expert to have a look at the server configuration and see what they can tell you.

Re new hosting, to be honest, without diagnosing what is causing the slow performance on your site the issue may well follow you to a new hosting provider.

Some things that can be checked:

· Does IPS have some cache settings which can be turned on / tweaked to reduce the number of queries and improve performance?

· Is Apache configured to handle an appropriate number of simultaneous requests?

· Is MySQL configured to handle an appropriate number of simultaneous requests and to handle the load it currently experiences?

· You should also be able to get CPU and memory usage (for both MySQL and Apache, plus in total) monitored and recorded so that you can check this against periods of poor performance.

+++++++++++++++++++++++++++++++++++++++++++++

FINAL CONCLUSION from superfuture database server managers:

+++++++++++++++++++++++++++++++++++++++++++++

> the server is configured properly

> It is not about hardware or configurations.

> the size of /tmp partition is not enough to run some tasks.

> by making further tweaks or moving servers and hosting we cannot guarantee the issue will be solved, because if the website contain a buggy code it will have issues on all servers.

> database code is causing performance issues.

>for example:

> ----------

> Slow_queries 110

> The number of queries that have taken more than long_query_time seconds.

> ----------

> Handler_read_rnd 4,258 k

> The number of requests to read a row based on a fixed position. This is high if you are doing a lot of queries that require sorting of the result. You probably have a lot of queries that require MySQL to scan whole tables or you have joins that don't use keys properly.

> ----------

> Handler_read_rnd_next 1,903 M

> The number of requests to read the next row in the data file. This is high if you are doing a lot of table scans. Generally this suggests that your tables are not properly indexed or that your queries are not written to take advantage of the indexes you have.

> ----------

> Qcache_lowmem_prunes 23 k

> The number of queries that have been removed from the cache to free up memory for caching new queries. This information can help you tune the query cache size. The query cache uses a least recently used (LRU) strategy to decide which queries to remove from the cache.

> ----------

> Created_tmp_disk_tables 258 k

> The number of temporary tables on disk created automatically by the server while executing statements. If Created_tmp_disk_tables is big, you may want to increase the tmp_table_size value to cause temporary tables to be memory-based instead of disk-based.

> ----------

> Select_full_join 15

> The number of joins that do not use indexes. If this value is not 0, you should carefully check the indexes of your tables.

> ----------

> Opened_tables 2,687

> The number of tables that have been opened. If opened tables is big, your table cache value is probably too small.

> ----------

> Table_locks_waited 287 k

> The number of times that a table lock could not be acquired immediately and a wait was needed. If this is high, and you have performance problems, you should first optimize your queries, and then either split your table or tables or use replication.

> ----------

+++++++++++++++++++++++++++++++++++++++++++++

FINAL CONCLUSION from IP.Board [server software]

--> just received now...

+++++++++++++++++++++++++++++++++++++++++++++

> after again examining all revised and clean slow query logs and full server reports.

all of these queries are related to huge topics (60k+ posts). Unfortunately these types of large topics are a

known issue with MySQL and we are looking at alternative ways of handling them in future versions of IP.Board.

For now, our recommendation is that you split these large topics in to smaller ones which will prevent these

queries from timing out when viewing later pages in the topic.

+++++++++++++++++++++++++++++++++++++++++++++

so... yeh. ok. right.

so if we accept the conclusions above... then we are now back again to where we started. covered here

it would seem that moving servers and hosting again is not necessary.

we've never been keen to be splitting up WAYWT and other huge threads but now going to have to take another look to see how to do it in a way that still gives users access to older posts. maybe it will be painless? hope so. will it guarantee the problem will be resolved? hope so.

will take another look tomorrow.

unless anyone's got any other great ideas?

hmm. i wonder what the world's biggest internet thread is?

i guess we hit the edge of the internet.

Share this post


Link to post
Share on other sites

after splitting most of the really big threads we had a quiet period during september and october when things seemed to finally be working without any problems...

unfortunately speed issues have been reported again in past week. ;-( ...

it seems for some users speed can be erratic. sometimes fast, other times terminally slow. reports of up to a minute for home page to load, etc etc bleh...

again, it's has not been possible to pinpoint the exact cause...

the speed problem does not appear to be server related but more likely software.

based on feedback from server managers and forum software developers, it has been recommended to upgrade from 3.3.3 to 3.3.4 of the forum software and install a new patch that has been developed for forums with large load topics [such as supertalk].

accordingly we have given the go ahead for this upgrade and now just co-ordinating with support people [server managers / server tweaker/ and skin designer] so that they are on stand-by in case any issues.

will update again once we have a clearer idea of schedule but it is likely to be underway this week.

Share this post


Link to post
Share on other sites

Thats all good to hear, hope this fixes the issue.

Share this post


Link to post
Share on other sites

we have given final go ahead for upgrade from 3.3.3 to 3.3.4 of forum software.

the upgrade and patch is going to be done by the forum software developers and we have 3 other tech teams on standby - server managers / server tweaker / skin designer / in case any unforeseen issues arise. [which as we all know - they always seem to do].

waiting on schedule. not sure if there will be any downtime.

will update again as soon as we know what the deal is.

WE REALLY HOPE ALL PERFORMANCE RELATED ISSUES ARE SORTED WITH THIS UPGRADE AND PATCH

!!!!!

Share this post


Link to post
Share on other sites

scheduled for next tuesday 20th november 'when it would be done sometime after 9am EST'...

once we get a more definite time will post it here.

still no confirmation if there will be any downtime.

stay tuned...

Share this post


Link to post
Share on other sites

dont know if this is related to the speed issue being fixed but the skin (dont know if ive used the right word here) seems to be messed up on supertalk , there is no supertalk or superdenim , supertrash ,ect buttons

Share this post


Link to post
Share on other sites

I'm assuming all the issues I'm encountering with the site today are the result of the upgrades being implemented?

Share this post


Link to post
Share on other sites

not sure what's causing the skin issues...

have just opened a ticket.

dont know if this is related to the speed issue being fixed but the skin (dont know if ive used the right word here) seems to be messed up on supertalk , there is no supertalk or superdenim , supertrash ,ect buttons

Share this post


Link to post
Share on other sites

dont know if this is related to the speed issue being fixed but the skin (dont know if ive used the right word here) seems to be messed up on supertalk , there is no supertalk or superdenim , supertrash ,ect buttons

came to post the same thing (couldnt get into any other threads, just the most recently posted on the main supertalk page)

Share this post


Link to post
Share on other sites

came to post the same thing (couldnt get into any other threads, just the most recently posted on the main supertalk page)

this. can't access anything.

Share this post


Link to post
Share on other sites

^ i can access other threads by using the search engine on top.

Share this post


Link to post
Share on other sites

Use the mobile theme and each forum has a clickable arrow all the way on the right hand side.

You can't see the forum title but you can still enter each forum just fine

Share this post


Link to post
Share on other sites

we expect this should be fixed when they do the upgrade on tuesday morning USA est. though still no explanation of what is causing it.

the upgrade will override this skin anyway.

++++++++++++++++++++++++

update ---->

actually we were hacked...!!

it's now fixed and a security patch has been installed.

please post if you notice any problems accessing any of the forums.

++++++++++++++++++++++++

as far as we know the 3.3.4 upgrade still on schedule for tuesday...

Share this post


Link to post
Share on other sites

really?

shit. well that would be nice...

i was expecting someone to say that after the upgrade... rather than before!?

but it's possible that all the most recent slowdown has been caused by the hacking which has just been fixed?????????????

[we also deleted a few 1000 spam member accounts today that seem to have accumulated over the past few weeks.]

is it me or this shit running back to normal speed/

gj

Share this post


Link to post
Share on other sites

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

  • Luisa via Roma (US)
    Privilege Program