• Resolved londonnut

    (@londonnut)


    I noticed that it’s quite slow to resolve 301 redirects (even after following the recommended optimisation tips – so there’s no reporting and no Filter Robots etc). (Direct links resolve straight away but my PrettyLinks take quite a few seconds).

    How do PrettyLinks get cached and how can I make sure that they are?

    I currently use Cloudflare and Litespeed cache (but without the object caching), works brilliantly for loading my pages fast but not so great for PrettyLinks at the moment. Any optimisation tips on how I could get the links to be faster to resolve with Litespeed? Is there some setting that should be tweaked? Or would adding an Object cache solution help – such as Docket Object caching? (https://wordpress.org/plugins/docket-cache/)

    Many thanks for the help

Viewing 8 replies - 1 through 8 (of 8 total)
  • Tyler

    (@tylerthedude)

    Hi @londonnut,

    Thank you for reaching out. As long as your pretty links are cached then the time it takes for the redirect to occur will depend on how fast the client’s connection is with the particular caching solution you’re using.

    Would it be possible to provide me with the URL for one of your pretty links where the redirect seems to be taking a while? I’ll try checking whether or not it’s currently being cached.

    I look forward to hearing from you!

    Kind regards,

    Thread Starter londonnut

    (@londonnut)

    Thank you for the information, useful to know!

    The speed is a lot better today so it looks like the issue has been resolved. (The website is StarFreebies.co.uk)
    I’m not sure what was causing the issues before, I ended up having quite a few critical PHP errors and the site crashed with critical errors. It was resolved by resetting PHP versions. I wasn’t able to get to the bottom of the cause just yet.
    It could just have been issues around Litespeed cache?

    The only pretty links error I noticed quite a few times was this one:

    Cron reschedule event error for hook: prli_cleanup_visitor_locks_worker, Error code: could_not_set, Error message: The cron event list could not be saved., Data: {“schedule”:”prli_cleanup_visitor_locks_interval”,”args”:[],”interval”:3600}

    Thanks again for getting back to me!

    Tyler

    (@tylerthedude)

    Hi @londonnut,

    Thank you for getting back to me, and I’m glad to hear the speeds are loading faster now! I’m thinking it might’ve been an intermittent issue with Litespeed or Cloudflare’s CDN as when the link is cached, our code shouldn’t be running at all.

    Regarding the notice you’re receiving in the log file – this can sometimes occur if a cron event was scheduled but never got executed due to the schedule associated with it being removed.

    Does Pretty Links get deactivated from time to time on your site? If it does, then you might’ve deactivated it while one of our cron events was scheduled, but since deactivating the plugin also removes our cron schedules (prli_cleanup_visitor_locks_interval in this case), WordPress will log that notice.

    These are just notices though so they shouldn’t impact anything on your site or within the Pretty Links plugin.

    I hope this helps!

    Kind regards,

    Thread Starter londonnut

    (@londonnut)

    I’m having the same issues at the moment (cached pages are fine and quick but any prettylinks are very slow to resolve). (My back end is also slow FYI).

    I don’t think the plugin is getting de-activated as it does work, but just slow to resolve links as though they’re not cached and a database lookup is needed each time…any idea if the fault could lie with Cloudflare or Litespeed. Or if you have any other suggestions please let me know.

    Tyler

    (@tylerthedude)

    Hi @londonnut,

    It’s nice to chat with you again, and I’m sorry to hear you’re still experiencing this behavior.

    Would it be possible to provide me with the URLs to one of your pretty links that’s having the slow response time so I may see what might be causing the problem?

    I look forward to hearing from you!

    Kind regards,

    Thread Starter londonnut

    (@londonnut)

    Thanks for getting back to me, I’ve been and to resolve this now (hopefully permanently now) by upgrading my hosting to a plan with greater resources!

    thanks again for your assistance!

    Thread Starter londonnut

    (@londonnut)

    Hi Tyler,

    The issue didn’t go away, however the good news is – my hosting and Litespeed were able to get to the bottom of the underlying cause. Litespeed is not caching PrettyLinks because the HTML code sent from PrettyLinks begins with a <meta tag instead of <!DOCTYPE html> – which is a requirement for Litespeed to detect and cache it.

    The fix would require a small code re-adjustment from Litespeed please.


    Details as received, below:

    Regarding the Pretty Links plugin’s redirections not being cached by the LiteSpeed Cache plugin (and so causing high CPU usage when the website is being heavily visited),
    LiteSpeed’s support have finally identified that the Pretty Link plugin is the underlying cause.

    When a web page is redirected, the Pretty Link plugin sends HTML code that starts as follows:

    <meta http-equiv="refresh" content="10;URL='https://starfreebies.co.uk/'" />
    
    <!doctype html>
    <html lang="en-GB">
    <head>
        <meta charset="UTF-8">
        <meta name="viewport" content="width=device-width, initial-scale=1">
        <link rel="profile" href="http://gmpg.org/xfn/11">
            <meta name="th

    But LiteSpeed says:

    per guideline , all HTML page should start with <!DOCTYPE html> , but the output from that plugin is <meta …
    ref: https://html.spec.whatwg.org/multipage/syntax.html#the-doctype

    In other words, the Pretty Links plugin is incorrectly writing <meta http-equiv="refresh" ... before <!doctype html>, and this stops LiteSpeed from being able to detect that it should cache it.

    Could you please assist with this modification?

    Many thanks!

    Tyler

    (@tylerthedude)

    Hi @londonnut,

    Thank you for getting back to me and providing that detailed information. Since you’ve opened a support ticket via our website, I’ll go ahead and close out this support thread as I believe the behavior you’re experiencing with the <meta> tag may be related to one of our Pro features.

    Kind regards,

Viewing 8 replies - 1 through 8 (of 8 total)
  • You must be logged in to reply to this topic.