Forum Replies Created

Viewing 15 replies - 16 through 30 (of 91 total)
  • Thread Starter yellofish

    (@yellofish)

    Fix #2 fixed it.

    After that had to go back to the sitemap settings and do a ‘save’ on all items.

    All good now.

    • This reply was modified 2 years, 8 months ago by yellofish.
    • This reply was modified 2 years, 8 months ago by yellofish.
    Thread Starter yellofish

    (@yellofish)

    Not sure what went wrong, but it works better now. Forwarding works fine. All is good.

    One question: I redirect 404 through RM to a out-of-Wordpress 404.php page.

    So what is better in terms of server load, 301 to 404.php – or – set to 410 ?

    When I put in 410 I get land on the default ‘page can’t be found’ which seems slower. I wonder what is better.

    Thread Starter yellofish

    (@yellofish)

    Most things fixed. One improvement suggested:

    Problem: Only 25 items/errors are displayed. This makes redirecting slow if you have a lot.

    Solution: Allow users to change the number of items displayed to 50/100/250/ or so.

    A minor change I guess.

    Thread Starter yellofish

    (@yellofish)

    I think that site might have some serious internal damage beyond repair. Most likely because of the user and post removal vial SQL. Thinking of moving posts and pages to a nginx host (but start with a new DB) and let nginx deal with the 404s.

    Thread Starter yellofish

    (@yellofish)

    Just an update on the 404 log. I start with Dashboard > Tools > Redirection > 404s

    I see when I go page to page in the 404s log (focus on total count):
    160 items, 1/7 pages
    129 items, 2/6
    160 items, 3/7
    228 items, 4/10
    228 items, 5/10
    186 items, 6/8
    186 items, 7/8
    118 items, 8/5 (none displays)

    Count in redirection_404 is 126

    Just going page to page and no actions.

    Actually a SQL command would come in very handy 😉

    • This reply was modified 2 years, 8 months ago by yellofish.
    Thread Starter yellofish

    (@yellofish)

    Thanks for the fast reply. I didn’t monitor one site for 2 weeks or and had suddenly 20000+ users and 30000+ posts – and getting plenty of hits on those SPAM posts.

    The hits come from bots. So they have nothing in common. Always new IP, different user agent etc. Hit count is not so high, but it’s a low performance server I guess.

    It looks like that the 404 logs takes a Minute or so to update. But anyway, for now it’s ok. Not very efficient though. Looks like most SPAM post hits are one time hits, not so much activity in the Log yet.

    Need to study your links.

    Thread Starter yellofish

    (@yellofish)

    Thanks for the reply. Well, yes and no. I try that app again.

    The plugin can redirect a SPAM URL to an outside-WP-page (or basically any URL), however, you need to make each redirect individually (or in groups of 25). It seems that the redirection is handled by WordPress rather than the server, putting load on WordPress – or looks like that.

    One annoyance is that the “select all” function and the the delete seem to work very buggy and you need to reload a lot.

    Will monitor it for a while and hope that Cloudflare will pick up those links.

    • This reply was modified 2 years, 8 months ago by yellofish.
    Thread Starter yellofish

    (@yellofish)

    I guess that was too early. I cleaned out most, but I still get when I am on the site near 100% CPU Usage.

    I certainly get too many hits to dead page (going to “page not found”), that triggers also Wordfence often to update.

    I get a large amount of cron jobs too, like: /wp-cron.php?doing_wp_cron=1579495442.6644361019134521484375

    Is there a way to block to dead URL hits? Cloudflare (free version that is) did not help much.

    Thread Starter yellofish

    (@yellofish)

    Thanks Hari,
    in the meantime I cleared those 5000+ entries in a few MySQL tables manually. Now I run a “Yoast optimize”. That seem to take hours. After I will add the site to Cloudflare.

    Will report back.

    Not sure it’s the FAQ somewhere:

    If you have any type of cache you need to purge it frequently while working on your page. Specially just before you logout. And that means any cache, WP, FastCGI, Redis, Cloudflare, ***CDN etc.

    If the Maintenance Mode still ain’t showing, clear your browser data (cache and cookies).

    This worked at least for me.

    yellofish

    (@yellofish)

    Just wanted to give 1 star – but completely disabling (fastCGI) cache did the trick. Don’t need to cache while doing the pages anyway.

    yellofish

    (@yellofish)

    fastCGI is a nginx server cache method and can speed up WP quite a bit. Are you on a nginx VPS?

    I have a few WP sites on different servers, and some on a shared litespeed host. All are good, but strangely the fastest server makes the most problems.

    Here is a fun thing you can try: use a VPN near your server (with the Chrome Windscribe plugin you can chose from a lot of IP locations, most are free) – so get a VPN near your server and check if you still get errors. For me, not, strangely.

    Bunny gives you 14 days free usage and after that for my low volume WP sites it’s like $1/Month for all sites combined.

    https://bunnycdn.com

    yellofish

    (@yellofish)

    I believe there are CF-server issues that are not related to this plugin. The WP CF plugin has also issues:

    https://github.com/cloudflare/Cloudflare-WordPress/issues/275

    Might be related to fastCGI.

    I moved a few sites away from Cloudflare to Bunny and no issues there at all so far.

    Thread Starter yellofish

    (@yellofish)

    Thanks, great link!

    So I got the icon from Font Awesome, however it has only one color and the “flash” is transparent, which is trouble if the background is not white. So I put a white rectangle behind the flash, then optimized it in SVGomg to strip all ballast – and this came out:

    <svg viewBox="0 0 512 512" xmlns="http://www.w3.org/2000/svg" fill-rule="evenodd" clip-rule="evenodd" stroke-linejoin="round" stroke-miterlimit="2"><path fill="#fff" d="M95.105 155.685h327.003v198.677H95.105z"/><path d="M256.55 8C116.52 8 8 110.34 8 248.57c0 72.3 29.71 134.78 78.07 177.94 8.35 7.51 6.63 11.86 8.05 58.23.367 10.663 9.239 19.235 19.908 19.235 2.744 0 5.458-.567 7.972-1.665 52.91-23.3 53.59-25.14 62.56-22.7C337.85 521.8 504 423.7 504 248.57 504 110.34 396.59 8 256.55 8zm149.24 185.13l-73 115.57a37.405 37.405 0 01-31.557 17.353 37.377 37.377 0 01-22.353-7.423l-58.08-43.47c-5.318-3.988-12.682-3.988-18 0l-78.37 59.44c-10.46 7.93-24.16-4.6-17.11-15.67l73-115.57a37.395 37.395 0 0131.554-17.357 37.365 37.365 0 0122.356 7.427l58.06 43.46c5.318 3.988 12.682 3.988 18 0l78.41-59.38c10.44-7.98 24.14 4.54 17.09 15.62z" fill-rule="nonzero"/></svg>

    Works fine.

    Thread Starter yellofish

    (@yellofish)

    It’s seems this method makes trouble in Android Chrome Mobile.

    Anybody has a proven working facebook messenger SVG?

Viewing 15 replies - 16 through 30 (of 91 total)