Wow, can’t believe it’s been 4 weeks since I started venturing down the path of reviewing managed WordPress hosting.
Since this still a new post series, things are a bit all over the place, and I’m still working out some kinks. Fortunately, this month is going to have a few insightful points of interest now that I’ve been up and running for the last month.
I’m still running the same instances as I was last month. It doesn’t make sense to tear down instances and rebuild them every month. That may be how you approach VPS servers, but for WordPress sites, they tend to be set and forget.
I’m still running the least expensive plans, and there have been no changes to my pricing. Unlike my current experience with MailChimp ;)
When I first started this series, the installs were all configured exactly the same, with minimal plug-ins installed.
Since then, the instances have diverged in regard to their WordPress version. More about that in a bit.
For each of the benchmarks, I am averaging the results of 3 different runs to try to get more average result.
|Storage||5 GB||5 GB||10 GB|
|Bandwidth||50 GB||50 GB||50 GB|
|Location||US - San Francisco||Oregon (US West)||North America|
You may remember me mentioning sales tax in regard to WPEngine last month. Seems this month, Kinsta will be doing the same thing.
|Operating System||Ubuntu 16.04 LTS||Ubuntu 16.04 LTS||Ubuntu 18.04 LTS|
|PHP Memory Limit||128 MB||256 MB||512 MB|
|PHP Max POST Size||300 MB||128 MB||100 MB|
|PHP Max Upload Size||300 MB||128 MB||50 MB|
|PHP Max Execution Time||1 minute||5 minutes||60 minutes|
Flywheel went from PHP version 7.2.18 to 7.2.19 and MySQL version 5.7.12 to 5.7.14 while WP Engine went from PHP version 7.3.4 to 7.3.5.
Both Flywheel and Kinsta automatically upgraded from WordPress 5.2.1 to 5.2.2. WP Engine did not.
Best as I can tell, WP Engine doesn’t automatically update because they want to encourage you to perform a backup before doing so.
||4.4100 sec||13.3133 sec||4.4400 sec|
||3.3400 sec||1.6000 sec||3.2900 sec|
||0.6867 sec||0.9067 sec||0.6067 sec|
Pretty consistent performance in comparison to last month.
Uptime and Response Time
Last month was a bit lacking because I had only just setup the uptime and response time monitoring, so there wasn’t much data to report on.
To give a bit more context to the following metrics, I am using UptimeRobot on a pro plan to monitor my WordPress instances. UptimeRobot is making requests from Dallas, Texas at an interval of once every minute.
Here’s the data for the last 30 days:
|Maximum Response Time||2448 ms||386 ms||446 ms|
|Average Response Time||202.66 ms||255.06 ms||197.94 ms|
|Time Period||576 hr 21 min||674 hr 53 min||676 hrs 57 min|
I dropped the minimum response time because it’s a pain to pick it out of the chart. Maximum remained because those spikes are easier to identify.
The big thing to note here is that Flywheel did end up going down for a couple of minutes on May 30th. I was able to confirm that my WordPress instance was in fact returning a 504 error, not to be confused with some sort of connectivity issue between UptimeRobot and my instance.
Network speed metrics are simply testing the download speed from the Google CDN edge server nearest the server of the host.
|Test 1||32.06 MB/s||92.91 MB/s||104.18 MB/s|
|Test 2||37.41 MB/s||102.09 MB/s||96.30 MB/s|
|Test 3||37.63 MB/s||105.24 MB/s||120.89 MB/s|
|Average||35.70 MB/s||100.08 MB/s||107.12 MB/s|
Also fairly consistent with last month, the exception being a bit more deviation between each run on both Kinsta and WP Engine.
Amazing to me how consistent the results have been this second month, compared to the first. That said, I still feel like WP Engine is the solid choice, albeit a bit more expensive.
The lack of an automatic update is a bit concerning, but the lack is to help ensure that absolutely nothing can go wrong without you knowing about it, that’s a good thing.
The downtime from Flywheel, while brief, was definitely concerning.
I’m still learning the ropes here of WordPress hosting and benchmarking, so feel free to drop any tips or advice on how to improve these posts in the comments below.
As always, I recommend doing your own research instead of taking these benchmarks at face value. It’s good to weigh your needs and pick the host that’s right for you and you own requirements.
If you did find this comparison helpful and plan to sign up for one of these providers, please do so using one of my referral links (sorry, no promotions at the moment):