Query Counts for XF, vB4, vB5, IPB etc.

Digital Doctor

Well-known member
====Query counts =====
vB5: 57 queries, 11.79 seconds.
vBulletin 5 beta 11 - 94-101 queries
vBulletin 5 beta 10 - 110 queries
vBulletin 5 beta 14 - approx. 60 queries
vBulletin 4 - 14 queries
Xenforo 1.1x - 11 queries forum home
Xenforo 1.1x - 11 queries on a 1 page thread
Xenforo 1.1x - 11 queries on a 38 page thread
Xenforo 1.1x - as low as 6-7 queries (with optimizations)
IPB - 123 queries for page 2 with a topic containing 400 posts and 61 followers
IPB - 49 queries on a 2 page thread (note)
- 103 queries to show a thread that is 123 pages. (note)
- 8 queries to show the main forum index page (note)

AndyB said:
vBulletin 5 beta 11 requires between 94 to 101 Asserts to display a thread. This is an improvement from beta 10 which required 110 Asserts.​
AndyB said:
vB4 takes about 14 queries to display a thread.​
-GR- said:
Just turned debug mode on my one IPB site. It is 8 queries to show the main forum index page and 103 queries to show a thread that is 123 pages.​
I don't have any cache turned on other then the built in IPB stuff.​
Important comparison issues: # of mods installed.
ManagerJosh said:
Stock install? No mods?​
-GR- said:
I just checked my XF site, forum index is 11 queries and thread view is 11 queries on a one page thread.​
hellreturn said:
IPB - 123 queries for page 2 with a topic containing 400 posts and 61 followers. Do note: While navigating to next/previous pages count is 18-23.​
-GR- said:
Just checked a 38 page thread and it was still at 11 queries.​
sadikb said:
I just saw Paul M's Area 52, which is running VB5 Beta 14. It has Debug on, so you can see that they have reduced it to 60 queries. Still too (too!) many for an empty site but I guess it can be called an improvement.​
The feeling I get right now from VB is that they are indeed focusing on performance, which is a good thing. Whether the architectural design of the software is capable of getting down to XF level of performance, I do not think so.​
R_A said:
I've seen it stated Xenforo has as low as 6-7 queries.​

Comparison issues: quality of the query. (There are queries that are more "expensive" than others !)(ie. number of queries doesn't explain everything)
 
Isn't just comparing numbers a bit misleading? Doesn't it depend on what they're doing and how optimised they are?
 
You'd need someone with a server to install them all on and perform proper tests. I thought Slavik was doing something like that for XF already?
 
It's definitely better to run 10 simple queries than 1 nasty query. But that being said, anywhere remotely close to 100 queries (of any sort) for a single page view is not viable for anything but the smallest sites.

And like others have said, it really depends on how optimized things are.

Preface: subtract one query from all numbers to not count the query being run to get ads to display (custom to my setup).

Some examples:

http://forums.digitalpoint.com
Logged in user with vB4: 10 queries
Guest user with vB4: 7 queries
Logged in user with XF: 7 queries
Guest user with XF: 7 queries

http://tools.digitalpoint.com (basic informational page)
Logged in user with vB4: 7 queries
Guest user with vB4: 5 queries
Logged in user with XF: 3 queries
Guest user with XF: 3 queries

Thread with 431 replies
Logged in user with vB4: 15 queries
Guest user with vB4: 10 queries
Logged in user with XF: 8 queries
Guest user with XF: 8 queries
 
====Query counts =====
vB5: 57 queries, 11.79 seconds.
vBulletin 5 beta 11 - 94-101 queries
vBulletin 5 beta 10 - 110 queries
vBulletin 5 beta 14 - approx. 60 queries
vBulletin 4 - 14 queries
Xenforo 1.1x - 11 queries forum home
Xenforo 1.1x - 11 queries on a 1 page thread
Xenforo 1.1x - 11 queries on a 38 page thread
Xenforo 1.1x - as low as 6-7 queries (with optimizations)
IPB - 123 queries for page 2 with a topic containing 400 posts and 61 followers
IPB - 49 queries on a 2 page thread (note)
- 103 queries to show a thread that is 123 pages. (note)
- 8 queries to show the main forum index page (note)

Comparison issues: quality of the query. (There are queries that are more "expensive" than others !)(ie. number of queries doesn't explain everything)
original thread

While it is true that no query is created equal.... The more people you have online, the more you'll feel those queries. And I don't know about most of you, but if I have as little as 500 people on the site with about 100 queries for each of those people

500 people X 100 queries = 50,000 database queries :eek:

Those queries maybe small little queries and alone they maybe harmless, but they add up regardless.
 
Highest query of my xenforo forum I ever see: 26 queries

23 add-on, 10 user login. Can't remember how many guest.
 
Is that really how it works? What about caching of queries?
We're talking MySQL here (which all the above are using)....

It really depends on the query being run. But you can't cache every query and usually not even the majority.

Assuming 1/2 of those were cached you'd still have 25,000 queries. And 500 people shouldn't have that many.
 
Top Bottom