Forum Index >  General >  General Support New Topic Post Reply
 Geeklog Benchmarking
First | Previous | 1 2 3 4 5 6 7 8 9 | Next | Last    |  Printable Version
By: tt0ne (offline)  Jan 05 2008 00:51 am (Read 26579 times)  
tt0ne

Hi Mark and all,

Have any of you done any Geeklog benchmarking? For example, I'm using "ab" and "siege" to run some benchmarks and I'm actually concerned about some of the numbers I'm seeing. Before I share them with you I'm curious as if anyone else has spent some time testing out high load values against Apache/Geeklog?

I'll post some scores tomorrow when I'm not so brain dead!

- Marco

HeavenlySanctuary.com - Surf The Innernet
Forum Active Member
Active Member

Group Comfort
Level:
: +10

Registered: 05/15/07
Posts: 287
Location: Redlands, CA

Profile Email Website  
  Quote
By: Mark (offline)  Jan 05 2008 07:21 am  
Mark

Marco,

I've actually done a lot of benchmarking for Geeklog and the plugins. I've focused more on how many additional SQL queries get called based on the various plugins and Geeklog configurations.

There are some definate areas where certain plugins will place an extraordinary load on the system. There are also some things you can do to help speed things up a little.

I would love to see your numbers and we can walk through what you're seeing and figure out how to improve things.

Thanks!
Mark

Forum Admin
Admin

Group Comfort
Level:
: +110

Registered: 10/21/05
Posts: 6326
Location: The Great State of
Texas

Profile      
  Quote
By: tt0ne (offline)  Jan 05 2008 10:53 am  
tt0ne

Hi Mark,

Hopefully this thread when we are finished will help some other poor sys admin out :-)

Ok, first some background on our servers: CLICK HERE

As you can see, we aren't some major mover and shaker but considering we are a non-profit organization that started in my dorm it's still pretty cool ;-) At the minimum, with proper configuration and attention to details, we should be able to handle a significant load before things started to crash and burn (correct me if I'm wrong).

web.heavenlysanctuary.com is running Centos 3, web2 is running Debian (running the latest testing cycle), and web3 is running Windows 2003 server. All the hardware stats are listed on the above link.

web.heavenlysanctuary.com is the host which runs Apache. When I say that I mean there isn't much else running on the machine, especially since recently I moved the entire database to web2. Furthermore, only www.heavenlysanctuary.com (http and https) are served from that machine. All of our other vhosts (there aren't alot - maybe another 10 - only 2 of them incur any kind of real traffic) are run from the other machines. All three machines are plugged into a full duplex gigabit switch.

When a person goes to http://www.heavenlysanctuary.com they will be served by Apache running on web.heavenlysanctuary.com and Apache will pull all of it's data from the database that sits on web2.heavenlysanctuary.com. I realize this is all elementary - but I just want to be really descriptive so others can help me understand just why I'm seeing the bad numbers.

One last thing - on web.heavenlysanctuary.com I use eaccelerator - a php accelerator. I've turned it off and on just to make sure it wasn't the issue and it wasn't.

To test, I'm using ApacheBench (ab) from one of my workstations at home which sits on a 15 down/3 meg up fiber connection (Verizon FIOS). My workstation runs Gentoo (I'm addicted to Gentoo) and is a dual-core Athlon.

* WHEW *

Now, if a person just goes over to www.heavenlysanctuary.com right now you can see the site is fairly quick and for me it just rendered the frontend in 54 seconds. I'm fairly positive that considering all the plugins we use that is really good.

But, what would happen if we started getting some moderate to major traffic? Well, let's find out: ;-)

PHP Formatted Code

ezra ~ # ab -n 5000 -c 2 http://www.heavenlysanctuary.com/index.php
This is ApacheBench, Version 2.0.40-dev <$Revision: 1.146 $> apache-2.0
Copyright 1996 Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
Copyright 2006 The Apache Software Foundation, http://www.apache.org/

Benchmarking www.heavenlysanctuary.com (be patient)
Completed 500 requests
Completed 1000 requests
Completed 1500 requests
Completed 2000 requests
Completed 2500 requests
Completed 3000 requests
Completed 3500 requests
Completed 4000 requests
Completed 4500 requests
Finished 5000 requests


Server Software:        Apache
Server Hostname:        www.heavenlysanctuary.com
Server Port:            80

Document Path:          /index.php
Document Length:        67658 bytes

Concurrency Level:      2
Time taken for tests:   2649.47512 seconds
Complete requests:      5000
Failed requests:        4353
   (Connect: 0, Length: 4353, Exceptions: 0)
Write errors:           0
Total transferred:      339689626 bytes
HTML transferred:       338279626 bytes
Requests per second:    1.89 [#/sec] (mean)
Time per request:       1059.619 [ms] (mean)
Time per request:       529.810 [ms] (mean, across all concurrent requests)
Transfer rate:          125.23 [Kbytes/sec] received

Connection Times (ms)
              min  mean[+/-sd] median   max
Connect:       21   23   2.8     24      60
Processing:   693 1035  58.7   1027    2060
Waiting:      590  925  56.6    920    1949
Total:        716 1058  59.0   1051    2082

Percentage of the requests served within a certain time (ms)
  50%   1051
  66%   1059
  75%   1066
  80%   1072
  90%   1115
  95%   1150
  98%   1213
  99%   1267
 100%   2082 (longest request)
 



From what I've read the most important # we are looking for is the Requests Per Second. As you can see 1.89 requests per second is about as bad as you can get. Furthermore, look at the failure rate!!! Something is wrong!

Of course, this isn't really that big of a stress test - I should be doing 50,000 instead of 5,000 but why even bother if even with 5,000 is basically falls apart? Furthermore, even if I change the number to 100 and make it 5 concurrent instead of 2 - it shows pathetic scores.

Now, what happens if I put just a plain old test.html page in the web server root and test it? (it's just a page which prints a line of text - no database, etc.).

PHP Formatted Code

ezra ~ # ab -n 5000 -c 2 http://www.heavenlysanctuary.com/test.html
This is ApacheBench, Version 2.0.40-dev <$Revision: 1.146 $> apache-2.0
Copyright 1996 Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
Copyright 2006 The Apache Software Foundation, http://www.apache.org/

Benchmarking www.heavenlysanctuary.com (be patient)
Completed 500 requests
Completed 1000 requests
Completed 1500 requests
Completed 2000 requests
Completed 2500 requests
Completed 3000 requests
Completed 3500 requests
Completed 4000 requests
Completed 4500 requests
Finished 5000 requests


Server Software:        Apache
Server Hostname:        www.heavenlysanctuary.com
Server Port:            80

Document Path:          /test.html
Document Length:        170 bytes

Concurrency Level:      2
Time taken for tests:   144.765381 seconds
Complete requests:      5000
Failed requests:        0
Write errors:           0
Total transferred:      2115000 bytes
HTML transferred:       850000 bytes
Requests per second:    34.54 [#/sec] (mean)
Time per request:       57.906 [ms] (mean)
Time per request:       28.953 [ms] (mean, across all concurrent requests)
Transfer rate:          14.26 [Kbytes/sec] received

Connection Times (ms)
              min  mean[+/-sd] median   max
Connect:       21   33 424.0     24   21014
Processing:    22   23   3.0     23      60
Waiting:       21   23   2.9     23      59
Total:         44   57 424.0     47   21037

Percentage of the requests served within a certain time (ms)
  50%     47
  66%     47
  75%     47
  80%     47
  90%     47
  95%     49
  98%     69
  99%     79
 100%  21037 (longest request)
 



Now, I'm going to say that this is still pretty abysmal (and I'm hoping that maybe others will help me out in figuring out why) but at least it's 34.54 Requests Per Second with no failures.

Now, I was thinking that to be really thorough I should perform the stress tests from a machine that sits on the same switch as the webserver and then another test right on the server itself. But I figure that's so irrelevant right now since the server is falling apart just from my fiber connection and workstation.

One last thing... Here are the settings from my httpd.conf file and my.cnf (I'm only showing the settings that have anything to with tweaking/performance):

httpd.conf:

PHP Formatted Code

<IfModule prefork.c>
StartServers               16
MinSpareServers         10
MaxSpareServers        40
MaxClients                  150
MaxRequestsPerChild  1000
</IfModule>

<IfModule worker.c>
StartServers               4
MaxClients                  150
MinSpareThreads        50
MaxSpareThreads       125
ThreadsPerChild         50
MaxRequestsPerChild  0
TheadGuardArea off
</IfModule>
 



my.cnf:

PHP Formatted Code

key_buffer              = 512M
read_buffer            = 32M
sort_buffer             = 32M
tmp_table_size       = 134217728
table_cache            = 1024
thread_cache          = 64
thread_stack           = 128K
query_cache_size    = 64M
query_cache_type    = 1
 



So, I hope that I've given enough information for those who know more then I do to help guide me in the right direction. If you need any data just ask and I'll do whatever I can to make it available.

And hey - despite my obvious problems - this really is fun! :-)

Warmly,

- Marco

HeavenlySanctuary.com - Surf The Innernet
Forum Active Member
Active Member

Group Comfort
Level:
: +10

Registered: 05/15/07
Posts: 287
Location: Redlands, CA

Profile Email Website  
  Quote
By: tt0ne (offline)  Jan 05 2008 11:07 am  
tt0ne

Almost forgot - I meant to include this with my last post.

I realize that DNS issues can cause all sorts of problems and I've put a lot of effort into making sure our DNS servers are fast and resolve properly. We have a premium account with dnsstuff.com (awesome site with awesome tools) and here are links showing that are DNS is rocking:

Full DNS report for heavenlysanctuary.com

DNS timing score for heavenlysanctuary.com

- Marco

HeavenlySanctuary.com - Surf The Innernet
Forum Active Member
Active Member

Group Comfort
Level:
: +10

Registered: 05/15/07
Posts: 287
Location: Redlands, CA

Profile Email Website  
  Quote
By: Mark (offline)  Jan 05 2008 11:14 am  
Mark

Marco,

I really do need more info, that just ins't enough Eek! Just kidding!!!!

I agree what you are seeing with your initial tests are not good. You have a lot going on with the front page, let's try to break it down a bit and see where we might have bottle necks:

Dynamic Content from the Right Blocks
- EVLIST
- What's New (Core, FileMgmt, Media Gallery)

Dynamic Content Center Block
- Forum Lists

Dynamic Content Left Blocks
Nothing really dynamic except the menu.

From testing that I've done, the biggest performance killer I've seen to date for plugins is Evlist. It will drop, create a temp table, copy a table and then compute future events twice for each page load! It does not contain any housekeeping routines so older events simply stay and are continually copied to the temp table, over and over. I would recommend that you disable it, then run your tests again to see what difference it makes, if any.

If you only have a handful of records in the evlist tables, it probably won't make a big difference.

The What's New queries are also very expensive. Due to the way many plugins store their data, the query is a multi join that can be a big drain on the database especially if you have a lot of records. I have never been able to tweak the SQL for either FM or MG where it doesn't walk the whole table.

I've actually started a new What's New plugin to replace the core one where data is cached. The logic here being here that there is no reason to query the DB with each page load to see if something is new. Instead, have the core and plugins push the changes when they are made. This means very few DB updates (only when something changes) and no more monster queries to for each page load. All this is still in development but hopefully I'll get it out sooner rather than later.

The Forum List in the centerblock is a pretty in-expensive query. Blaine and I worked very hard on tweaking that SQL and have it down to where it should not make much of a load difference at all.

First, take EVlist out of the picture, I think it will prove to be a big load item and lets see if that makes any difference.

Thanks!
Mark

Forum Admin
Admin

Group Comfort
Level:
: +110

Registered: 10/21/05
Posts: 6326
Location: The Great State of
Texas

Profile      
  Quote
By: tt0ne (offline)  Jan 05 2008 12:16 pm  
tt0ne

Jeez Mark - I was just about to go have a coffee after writing all that expecting that you may still be in bed! HA!

Ok, I'm going to disable EVList and run the same test. Then I'll disable the "What's New" block, too and see if it gets even faster.

And please don't hold back on explaining all the SQL stuff - I have no business administrating a database - it's my weakest point and I'm very eager to understand and learn more about it - so your explanation about multi joins, etc. are very helpful to me.

Here are the results - now I'm really perplexed.

PHP Formatted Code

ezra ~ # ab -n 5000 -c 2 http://www.heavenlysanctuary.com/index.php
This is ApacheBench, Version 2.0.40-dev <$Revision: 1.146 $> apache-2.0
Copyright 1996 Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
Copyright 2006 The Apache Software Foundation, http://www.apache.org/

Benchmarking www.heavenlysanctuary.com (be patient)
Completed 500 requests
Completed 1000 requests
Completed 1500 requests
Completed 2000 requests
Completed 2500 requests
Completed 3000 requests
Completed 3500 requests
Completed 4000 requests
Completed 4500 requests
Finished 5000 requests


Server Software:        Apache
Server Hostname:        www.heavenlysanctuary.com
Server Port:            80

Document Path:          /index.php
Document Length:        65413 bytes

Concurrency Level:      2
Time taken for tests:   2774.798396 seconds
Complete requests:      5000
Failed requests:        4817
   (Connect: 0, Length: 4817, Exceptions: 0)
Write errors:           0
Total transferred:      327218712 bytes
HTML transferred:       325808712 bytes
Requests per second:    1.80 [#/sec] (mean)
Time per request:       1109.919 [ms] (mean)
Time per request:       554.960 [ms] (mean, across all concurrent requests)
Transfer rate:          115.16 [Kbytes/sec] received

Connection Times (ms)
              min  mean[+/-sd] median   max
Connect:       21   23   2.5     23      58
Processing:   736 1086  84.3   1079    4135
Waiting:      572  919  78.9    912    3909
Total:        759 1109  84.4   1102    4157

Percentage of the requests served within a certain time (ms)
  50%   1102
  66%   1112
  75%   1122
  80%   1134
  90%   1183
  95%   1227
  98%   1288
  99%   1351
 100%   4157 (longest request)
 



There must be something wrong with my MySQL settings or maybe my database is borked (I seem to remember looking at a bunch of MySQL runtime information stats that suggested I had index problems - but of course - I'm not completely sure what is the difference between an index and a finger - so who knows?

I'm going to run it without the "What's New" block now to see what happens. This takes quite a while - so I'll put it in the next post.

- Marco

HeavenlySanctuary.com - Surf The Innernet
Forum Active Member
Active Member

Group Comfort
Level:
: +10

Registered: 05/15/07
Posts: 287
Location: Redlands, CA

Profile Email Website  
  Quote
By: jmucchiello (offline)  Jan 05 2008 12:21 pm  
jmucchiello

Mark, I actually had a hack where php blocks could be cached just like rdf blocks. You could set the update frequency and it would pull the data out of content rather than run the php until it expired. I don't know if I still have a running copy of the hack, but it would be pretty easy to write again. Problem is what's new isn't a phpblock. That's yet another hack I once had. Turning the builtin geeklog blocks into phpblocks. But I know I've since lost that one. Again, not hard to write but since it's so hard to get stuff into core, I've given up on such hacks.

Forum Active Member
Active Member

Group Comfort
Level:
: +2

Registered: 05/15/07
Posts: 445

Profile Email    
  Quote
By: jmucchiello (offline)  Jan 05 2008 12:53 pm  
jmucchiello

Quote by: tt0ne

ezra ~ # ab -n 5000 -c 2 http://www.heavenlysanctuary.com/index.php

This is probably your bottleneck. It takes .5 sec to render a page and you only grab 2 at a time so naturally you won't get more than 2 pages per second. Raise the -c parameter to actually cause some load. Try it at 10 and see if you get around 20 pages per second. You max load is when the average pages/sec is no longer close to concurrency number / secs per page.

Forum Active Member
Active Member

Group Comfort
Level:
: +2

Registered: 05/15/07
Posts: 445

Profile Email    
  Quote
By: tt0ne (offline)  Jan 05 2008 13:17 pm  
tt0ne

Here is the results from disabling EVList and the What's New Block:

PHP Formatted Code

ezra ~ # ab -n 5000 -c 2 http://www.heavenlysanctuary.com/index.php
This is ApacheBench, Version 2.0.40-dev <$Revision: 1.146 $> apache-2.0
Copyright 1996 Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
Copyright 2006 The Apache Software Foundation, http://www.apache.org/

Benchmarking www.heavenlysanctuary.com (be patient)
Completed 500 requests
Completed 1000 requests
Completed 1500 requests
Completed 2000 requests
Completed 2500 requests
Completed 3000 requests
Completed 3500 requests
Completed 4000 requests
Completed 4500 requests
Finished 5000 requests


Server Software:        Apache
Server Hostname:        www.heavenlysanctuary.com
Server Port:            80

Document Path:          /index.php
Document Length:        64377 bytes

Concurrency Level:      2
Time taken for tests:   2727.319397 seconds
Complete requests:      5000
Failed requests:        4931
   (Connect: 0, Length: 4931, Exceptions: 0)
Write errors:           0
Total transferred:      322022198 bytes
HTML transferred:       320612198 bytes
Requests per second:    1.83 [#/sec] (mean)
Time per request:       1090.928 [ms] (mean)
Time per request:       545.464 [ms] (mean, across all concurrent requests)
Transfer rate:          115.31 [Kbytes/sec] received

Connection Times (ms)
              min  mean[+/-sd] median   max
Connect:       21   23   2.5     23      58
Processing:   722 1067  88.0   1061    4139
Waiting:      560  896  84.9    895    3909
Total:        744 1090  88.1   1084    4163

Percentage of the requests served within a certain time (ms)
  50%   1084
  66%   1094
  75%   1103
  80%   1114
  90%   1166
  95%   1213
  98%   1269
  99%   1317
 100%   4163 (longest request)
 



So basically nothing has changed... I'm going to run a test now with Mucciello's advice and I'll post the results as soon as it is done. Thanks for the help guys!

- Marco

HeavenlySanctuary.com - Surf The Innernet
Forum Active Member
Active Member

Group Comfort
Level:
: +10

Registered: 05/15/07
Posts: 287
Location: Redlands, CA

Profile Email Website  
  Quote
By: Mark (offline)  Jan 05 2008 13:22 pm  
Mark

Marco,

I think Joe is on the right track, basically ab is a simple tool that has very little explanation about what it is really doing.

How I've used it in the past is to start with a small number (say 10) and run -n 10 -c 5 to simulate 10 concurrent users making 5 requests. Then grow the -n number by5 until you start seeing failed requests. This now tells you that you can get approximately XX concurrent users with no performance issues.

You can then add the -t parameter to increase the time to wait and continue to grow the -n parameter. Then you'll get an idea of how many more concurrent users you can support with a bit slower response.

The problem with benchmarking is there is nothing that is truely black and white. It is very difficult sometimes to fully understand and get a good representation of what your site can really handle.

Thanks!
mark

Forum Admin
Admin

Group Comfort
Level:
: +110

Registered: 10/21/05
Posts: 6326
Location: The Great State of
Texas

Profile      
  Quote
By: tt0ne (offline)  Jan 05 2008 14:03 pm  
tt0ne

Mark & Joe,

Here is the latest test results:

PHP Formatted Code

ezra ~ # ab -n 5000 -c 10 http://www.heavenlysanctuary.com/index.php
This is ApacheBench, Version 2.0.40-dev <$Revision: 1.146 $> apache-2.0
Copyright 1996 Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
Copyright 2006 The Apache Software Foundation, http://www.apache.org/

Benchmarking www.heavenlysanctuary.com (be patient)
Completed 500 requests
Completed 1000 requests
Completed 1500 requests
Completed 2000 requests
Completed 2500 requests
Completed 3000 requests
Completed 3500 requests
Completed 4000 requests
Completed 4500 requests
Finished 5000 requests


Server Software:        Apache
Server Hostname:        www.heavenlysanctuary.com
Server Port:            80

Document Path:          /index.php
Document Length:        67929 bytes

Concurrency Level:      10
Time taken for tests:   2215.759160 seconds
Complete requests:      5000
Failed requests:        4853
   (Connect: 0, Length: 4853, Exceptions: 0)
Write errors:           0
Total transferred:      339852152 bytes
HTML transferred:       338442152 bytes
Requests per second:    2.26 [#/sec] (mean)
Time per request:       4431.519 [ms] (mean)
Time per request:       443.152 [ms] (mean, across all concurrent requests)
Transfer rate:          149.78 [Kbytes/sec] received

Connection Times (ms)
              min  mean[+/-sd] median   max
Connect:       21   23   2.8     23      55
Processing:  1142 4358 11704.6   3864  463498
Waiting:      950 4094 11584.0   3601  459917
Total:       1166 4382 11704.7   3887  463522

Percentage of the requests served within a certain time (ms)
  50%   3887
  66%   4372
  75%   4628
  80%   4791
  90%   5207
  95%   5506
  98%   5853
  99%   6158
 100%  463522 (longest request)
 



Well, now we are doing a whopping 2 pages a second ;-) So now that we've increased the number to 10 and we are running the benchmark properly what does everyone think? I agree, Mark - benchmarks can be extremely misleading - but I think in this case it is accomplishing exactly what they should be designed for - to reveal not so much what real world performance would look like but in this case - to reveal there is a major problem!

Is there any data I can mine from mysql server - any runtime info - that would give me an idea if something is borked between apache and mysql?

- Marco

HeavenlySanctuary.com - Surf The Innernet
Forum Active Member
Active Member

Group Comfort
Level:
: +10

Registered: 05/15/07
Posts: 287
Location: Redlands, CA

Profile Email Website  
  Quote
By: tt0ne (offline)  Jan 05 2008 14:05 pm  
tt0ne

I'm concerned about the failed requests - does it mean those requests timed out? Is it possible the real time out is happening between the apache server and the mysql server?

Just thinking out loud.

- Marco

HeavenlySanctuary.com - Surf The Innernet
Forum Active Member
Active Member

Group Comfort
Level:
: +10

Registered: 05/15/07
Posts: 287
Location: Redlands, CA

Profile Email Website  
  Quote
By: tt0ne (offline)  Jan 05 2008 14:08 pm  
tt0ne

My Handler_read_rnd_next = 38 M and according to the explanation within phpmyadmin it means the server is doing a lot of table scans and that they are not properly indexed.

My question is - is 38 Megs a lot? I feel dumb. :-)

- Marco

HeavenlySanctuary.com - Surf The Innernet
Forum Active Member
Active Member

Group Comfort
Level:
: +10

Registered: 05/15/07
Posts: 287
Location: Redlands, CA

Profile Email Website  
  Quote
By: tt0ne (offline)  Jan 05 2008 16:16 pm  
tt0ne

Ok, I had lunch and while I was gone I ran this test with "-c 5" and it performed will but this was just with a static page - so it would seem like the real problem must be the transactions between the webserver and the mysql server, right?

PHP Formatted Code

ezra ~ # ab -n 5000 -c 5 http://www.heavenlysanctuary.com/test.html
This is ApacheBench, Version 2.0.40-dev <$Revision: 1.146 $> apache-2.0
Copyright 1996 Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
Copyright 2006 The Apache Software Foundation, http://www.apache.org/

Benchmarking www.heavenlysanctuary.com (be patient)
Completed 500 requests
Completed 1000 requests
Completed 1500 requests
Completed 2000 requests
Completed 2500 requests
Completed 3000 requests
Completed 3500 requests
Completed 4000 requests
Completed 4500 requests
Finished 5000 requests


Server Software:        Apache
Server Hostname:        www.heavenlysanctuary.com
Server Port:            80

Document Path:          /test.html
Document Length:        170 bytes

Concurrency Level:      5
Time taken for tests:   200.908429 seconds
Complete requests:      5000
Failed requests:        0
Write errors:           0
Total transferred:      2115000 bytes
HTML transferred:       850000 bytes
Requests per second:    24.89 [#/sec] (mean)
Time per request:       200.908 [ms] (mean)
Time per request:       40.182 [ms] (mean, across all concurrent requests)
Transfer rate:          10.28 [Kbytes/sec] received

Connection Times (ms)
              min  mean[+/-sd] median   max
Connect:       21  176 2898.0     24   92983
Processing:    22   23   3.1     23      51
Waiting:       22   23   2.9     23      50
Total:         44  200 2898.0     47   93009

Percentage of the requests served within a certain time (ms)
  50%     47
  66%     47
  75%     47
  80%     47
  90%     49
  95%     52
  98%     76
  99%     88
 100%  93009 (longest request)
 



- Marco

HeavenlySanctuary.com - Surf The Innernet
Forum Active Member
Active Member

Group Comfort
Level:
: +10

Registered: 05/15/07
Posts: 287
Location: Redlands, CA

Profile Email Website  
  Quote
By: Laugh (offline)  Jan 05 2008 17:40 pm  
Laugh

What traffic do you have now?

What do you consider as major traffic and what are the chances of you meeting this number?

Also would joe's template cache system help here?

Forum Chatty
Chatty

Group Comfort
Level:
: +1

Registered: 03/17/07
Posts: 64

Profile Email Website  
  Quote
New Topic Post Reply

First | Previous | 1 2 3 4 5 6 7 8 9 | Next | Last

 All times are CDT. The time is now 07:20 pm.
Normal Topic Normal Topic
Locked Topic Locked Topic
Sticky Topic Sticky Topic
New Post New Post
Sticky Topic w/ New Post Sticky Topic w/ New Post
Locked Topic w/ New Post Locked Topic w/ New Post
View Anonymous Posts 
Able to Post 
HTML Allowed 
Censored Content