Site Health: Critical issue – Median Server Response Time

Hello @spale

Thank you for reaching out and I am happy to help!
I’ve checked your website and as I can see the response time is great:

All the testing suggests that the TTFB is around 330ms
Can you please share where the server is based and can you try to use the Disk: Enhanced caching method for page caching and let me know if you see any difference.
Also, try the Setup wizard in W3TC and see which caching engine offers the best performance and what is speed tested.

Thanks!

Hello Marko,

Thanks for the answer.

Server is based in Nuremberg in Germany, Hetzner is hosting provider.
I have changed the Page Cache settings to Disk: Enhanced.

Also, I ran the settings wizard and it showed me that it is better to use Database Cache – Redis, and not to use Object Cache at all.

The problem that WP site health tool reports is still present:

Do you have any idea how to solve this?

Best regards,
Milos

I feel that “WP site health” is doing something deeply wrong, like accessing index page as a logged-in user (which will not be cached, of course), or something like that.

From Incognito browsing I always get 60 to 80 ms TTFB which is pretty fine for distance from Serbia to Hetzner datacenter.

Hello @spale

Thank you for providing the information.
@dpeca You are right, and there is also the aspect of Cloud flare.
@spale As you can see in the setup guide, the Disk: Enhanced shows similar results as I have provided. You may want to check the Cloudflare configuration, and under the Browser Cache TTL make sure to respect origin settings.

Also make sure to enable etag in Performance>Browser Cache, save all settings and purge the cache.

Thanks!

Hello Marko,

Thanks for the answer.

I have checked the W3TC CloudFlare configuration, the Browser Cache TTL is configured to same time as the origin.

The etag is enabled under the Performance -> Browser Cache. I have purged all cache.

The result is the same, still getting the message that the server response is slow.

Do you have any other suggestions?

Best regards,
Milos

Hello @spale

Can you please clarify that the same settings you have for browser cache TTL in Cloudflare are also set in the W3TC Browser Cache settings? Or have you changed this in Cloudflare to respect the origin headers which was suggested?

Try caching for logged-in users just to see if the WP health status will change.

Make sure to revert it back and once the test is done.
And also, after each setting is changes, make sure to save all settings and purge the cache.

Thanks!

Hello Marko,

I can confirm that the same settings I have for browser cache TTL in CloudFlare are also set in the W3TC Browser Cache settings.

I didn’t change this in CloudFlare, or in the W3TC, it was already set that way when I went into the settings.

I cant turn on caching for logged in users, because its a live e-commerce site, and when caching for logged in users is turned on, buyers cannot log in to their accounts. So its not the option.

Do you have any other ideas how to fix this?

Best regards,
Milos

Hello @spale

Sorry for the late reply. WP was stuck yesterday and I was not able to answer for some reason.
I’ve performed tests on the website using various things to determine the TTFB or if the initial server response time needs to be reduced. Every tests suggests that there is no problem with the server response time. The Pages are cached, the cache-control, expires header, etag, gzip, are all enabled and working as expected.
Cloudflare is also adding a layer of caching so there is reason for the Site health to suggest that the Median server response time should be less than 600ms, as every test and user interaction suggests that it is.

SO the fact that you are still seeing this may be out of the W3TC control and there is some plugin may be causing this.
I understand that this is a live website so troubleshooting may be difficult, however, ?you should make sure plugins, themes and the PHP are updated to the newest versions available, and disable the plugins one by one to see which one may be causing this in the backend.

Thanks!

Hello Marko,

Thank you for the answer and thank you for all the efforts to help me.

I have also ran various tests in incognito tab and all results show that the server is running fast and everything is cached properly. The average time for TTFB is around 100ms, which is great.

There is definitely a possibility that a plugin is causing this issue with the WordPress site health tool. Since it is a live site, def. I can’t turn off plugins one by one to find which one is causing the problem, so I’ll leave everything as it is for now and mark this thread as solved.

Best regards,
Milos