• TL;DR If WP Page Load Stats shows 52mb of memory used per page load, 2 would be 104mb, 4 208mb, etc?

    Ok, so I’m thoroughly confused on what the best route for me to go would be with my WP blog. On my local server, I just set WP to use 256mb ram (why not) and off I go. Since I probably wouldn’t have such a good time doing that on WPE/Synthesis/MediaTemple/Linode I figure before I go jumping into a solid host I’d ask you guys.

    I’ve really been looking at WPEngine and Synthesis, and then decided to take a gander at MT and in the meantime remembered Linode has VPS which out of the 4 of those alone, MT’s shared grid apparently has a limit on mysql ram usage of 99mb and that’s where I was kind of taken back. For example, on a live site right now (that I host locally) my WP Page Load Stats is:

    98 queries in 0.339 seconds.
    Average load time of 0.339 (1 runs).
    52.14 out of 256 MB memory used.
    Peak memory usage 52.63 MB.

    So I said to the guy at MT, does that essentially mean I couldn’t even have 2 visitors hit that one site at the exact same time, let alone ~5, 10, 50, 100 at the same time? His reply was yes (and I’m sure he misunderstood the question as that seems extremely odd) and I simply said “that makes no sense at all to me. you guys do run WP sites right?” considering the fact, I don’t feel any of my WP sites are that poorly designed and configured…

    So, first question; was he “right” about if they only allowed mysql (on shared plans) to be 99mb of ram, I couldn’t even have 2 visitors hit my site concurrently without going over?

    Second question; would that then also mean I’d only be able to reliably have 15-25 concurrent visitors be able to hit a site (concurrently of course) on a 1gb linode VPS?

Viewing 1 replies (of 1 total)
  • I lost a little bit the thread of what the first question was, but, the “exact same time” is just that. The EXACT same time. At the same instant. Of course, if you don’t have the resources for this, your system will attempt to “queue” whatever request it is serving.

    Now, in your example, it shows that the page loads in 0.339 seconds. Assuming client 1 and client 2 hit your website at the EXACT same instant (though this isn’t really possible in a world of infinite instants?!) then client 1 would be first served, then client 2 (if your system can only cope with 1-at-a-time. Client 1 loads your page in, say, 0.4 seconds. Client 2 loads it in 0.8 seconds. Both are tolerable.

    In reality, client 2 would hit 1 second later and wouldn’t notice any difference.

    Try looking up benchmarking tools, like apache benchmark (ab), to test your website under concurrent loads. There is another tool which I use that has gone from my mind, with a web-based interface. Use sparingly, and not on production sites, as you can break things!

    No – this does not mean 15-25 concurrent visitors. I load your page in 0.4 seconds, and, presumably, I then spend 5 minutes or so loading NOTHING while I read your fascinating blog post ;). It DOES mean you’d only be able to handle that many concurrent LOADS, but these can (and will) be queued.

    A minute system can handle 100 visitors over 5 minutes on WP.

    Also recommend you investigate caching systems, like WP Super cache and/or W3 Total Cache, though note the latter is quite complex.

Viewing 1 replies (of 1 total)
  • The topic ‘Questions about the right hosting solution for my WP sites…’ is closed to new replies.