One VPS, seven blogs: what is better, one or many MySQL databases?

I recently switched from 8 individual installs into one multisite. Didn’t notice any performance changes whatsoever. Database size is almost the same as all 8 databases combined because the same information is there.

The only thing that is now much faster is maintenance – plugins, WordPress core and themes can be updated with a single click instead of going on each dashboard separately. There are some tools that allows to manage multiple WordPress install under one roof without multisite but I still prefer to keep everything in one place.

It all comes down to performan preference if you want to use multisite or not because there aren’t any performance benefits at least I didn’t notice any for highly monitored website.

Hi Silkalns, and thanks for your comment!

I knew that there would be much less maintenance to do, but knowing that there is no performance benefit in going multisite, this is unexpected: Didn’t notice any performance changes whatsoever.

This for me is interesting and useful news.

Thanks

By the way, let me elaborate on my original question:

I am wondering how much the difference in performance will be, of having 7 databases instead of one with seven sets of wp tables. Thoughts, optimization tips?

One can have one single database with seven sets of wp tables in two ways: first is going multisite, second is using seven different worpdress installs, ALL pointing to the same database, but each using a different table prefix.

Silkalns’ answer refers to the first case only, doesn’t it? I mean, maybe the answer is the same EVEN in the second case, but not necessarily. Or not?

seven different worpdress installs, ALL pointing to the same database, but each using a different table prefix.

I only do that as pairs of individual sites sharing users, and I once (for just a short while before catching my own oversight) even had both sites being handled by the same SQL user with no trouble.

OK, to sum it up, we have 3 combinations (because number 4 is NOT possible, is it?):

1. 7 wp installs, 7 databases
2. 7 wp installs, 1 database
3. 1 multisite, 1 database
4. 1 multisite, 7 databases

and:

1. silkalns reports no performance difference between #1 and #3
2. leejosepho reports no problem with #2</>

Is that correct?

Thanks

I think your summary is accurate, so the question really breaks down to the relationships between these three things:

1. actual need (such as when using multisite or even just sharing users between separate sites);
2. convenience in maintenance;
3. operation and sustainability of any single site in the event of server-level intrusion or any overall shared-database corruption possibly cause by one site or another.

Yes, I agree. In MY own case I am oriented towards independent blogs instead of multisite install because, short version, unless there ACTUALLY IS some performance penalty, I prefer it. Longer version:

1. I will do this on a VPS of which I am root, so I can customize it any way it is needed/I prefer
2. so far, nobody (even in other places) has told me that using seven different wordpress installs each with its database, wrt 1 multisite+1db, the seven websites would be slower for users (also because, due to #1 I will be able to manage/consolidate caching, page compression and similar at the httpd server level, ie use less plugins in wordpress
3. this is only my own (strong) feeling, but here it goes: wordpress plugin and themes are, in general, tested more in single installs than in multisite ones, so it is less likely to happen into bugs/untested situations if you stick to single installs
4. Robustness: due to #1 again, I can do manual updates, i.e. check if something goes wrong, only on one blog and then: if it’s OK, I do the same thing on all the others with wp-cli, automatically; if it goes bad, I only have one broken blog to fix

Does it make sense? If I overlooked something, please let me know

wordpress plugin and themes are, in general, tested more in single installs than in multisite ones, so…

I do not know whether that is true, but we still have the matter of whether and/or how well one cook of whatever classification can multitask. Some, I am sure, can do that where others might not do so well.

I can do manual updates, i.e. check if something goes wrong, only on one blog and then: if it’s OK, I do the same thing on all the others…

That is what I do since it is no big deal if my last one (done first) breaks.