Bare Metal vs. High Frequency? (Vultr)

frm

Well-known member
I'm currently running 1 forum with a HF 2vCPU with a dedicated HF 1vCPU database server and a couple other HF 1vCPU instances for other websites that get trickle traffic.

My forum is not large, nor are my other forum(s) and websites. The "main" and largest forum is on 2vCPU/1vCPu right now and displays this as an execution time:Width Queries 30 Time 0.1729s Memory 9.31MB

Would it perform better all on one bare metal instance of 4 core/8 threads & 32GB memory or am I better splitting them on separate instances?

As it stands, my costs are about the same ($100) to host as is, but if I can get better performance out of the bare metal ($120), it would behoove me to change.

I run @eva2000's CMM with PHP 7.4 on all instances, if that helps.
 
I'm running 5 sites (1 xf, 4 wp) on one 4gb/2cpu HF vps. it chums along at 5% steady state on cpanel. Total traffic is about unique 1800/day

I've also offloaded all my attachments to s3 and route and cache dns with cloudflare and cloudfront which greatly reduces root server impact.

I don't think you need the bare metal or the additional servers that you have now. Bringing it all on one machine will reduce your internal network IO (and charge for private internal ip) and without the cpanel cost, you can probably be out the door for under $50.

You can always increase your space there without much adu. start small, add a gig/proc when necessary.
 
  • Like
Reactions: frm
I don't think you need the bare metal or the additional servers that you have now. Bringing it all on one machine will reduce your internal network IO (and charge for private internal ip) and without the cpanel cost, you can probably be out the door for under $50.
My thinking behind this was security related. If 1 instance is compromised somehow, nothing else goes down.

I've been using WordPress since 2010ish and it seems like there's always some security flaw in it or a plugin where if 1 site is compromised, everything is wide open. Ever since then, I've put 1 site I "cherish" on its own hosting and combine 5-10 on 5-page sites on another VPS/instance (as it pays for itself and then some) and just take snapshots to back up here and there as they are merely there for AdSense until the domain is sold (for the name and/or with content for passive income as parking is somewhat dead over a 5-page WordPress install with good content).

Therefore, I was wondering if I should go "all in" on bare metal and see if the same is true with WP+XF on the same box and hope that a WP install is not compromised where one can access the XF directories and compromise it too (though they'd have to edit the file and update the hashes.json file or it's be obvious).
 
@briansol, can you perhaps post the queries and execution time for the 1 XF instance on that and also confirm whether you are running MySQL on the same instance or have a separate database for it or all setups (XF and WP).

I'm intrigued to see whether 1 XF instance with a separate database server outperforms your setup on essentially the same 2vCPU (plus the WP sites) setup with the 1vCPU database with my time being around Time 0.1729s (but have seen it as low as 0.12).

I'm trying to get a very low execution time as I seem to rank higher day by day tweaking it on its current setup and unsure whether to take a plunge to have multiple MySQL databases on the same server that is bare metal if it can even cut the low of 0.12 or 0.09 or greater. That may give me an edge over the competition and well worth the extra money if it allows me to rank higher.

Also, NY/NJ allow for attaching blocks, which I believe is cheaper than AWS (but I would have to revisit this and see the pricing per GB).

Thanks!
 
I have everything on 1 box.

home:

  • Width

  • Queries 9

  • Time 0.0563s

  • Memory 4.50MB


thread:


  • Width

  • Queries 8

  • Time 0.0508s

  • Memory 4.48MB
 
  • Wow
Reactions: frm
looks like you have a lot of addons or extra bloat contributing to more queries... i have a few, but mostly resource improvements eg, xon's redis cache, image optimizers, etc
 
looks like you have a lot of addons or extra bloat contributing to more queries... i have a few, but mostly resource improvements eg, xon's redis cache, image optimizers, etc
I'm using memcached. Looks like I might make the switch and then decide. But then again, if I can get to 0.025 on bare metal, that is still significant savings that can have an effect on SERPs. Inching closer to higher positions day by day... by trying to cut down the speed to get to "3G acceptance" (AMP is next).

I run AMS (blog), RMS (review), Classifieds, Event Calendar, and XFMG. Others are small tweak add-ons to make it easier (change author/change post date). I have 26 add ons with the aforementioned critical to running a CMS. The others can be easily dropped.
 
my numbers are pretty good.

FCP is missing benchmark by .1
CLS still shows up at .15. I can't figure out where this happens but i'm guessing it's from ad code.
all others are in the green.

And that gives my mobile score a 47. lol
 
And that gives my mobile score a 47. lol
Yeah, I can't seem to get this up much either.

I'll do investigating on that after possibly testing Redis over Memcached and moving to a bare metal server, provided I can have security from 1 site accessing others as with CentminMod, nginx is the owner and if a WP can be exploited, see whoami, and then browse around to other files owned by nginx... they are all compromised (if someone has time).
 
Top Bottom