Sidebar
Browse Our Articles & Podcasts

Bug Hunt

It’s driving me crazy.

We have this technical glitch that has been happening over the past couple of weeks, and no matter what we try, we can’t seem to find what’s causing it.

You’ve probably seen it. It’s a database error that comes up every so often when you load our website. It’s either a simple error “Database failed to load” or “Error establishing database connection” or it’s a table with a bunch of information that looks like this:

It’s not just you – I’m seeing it too. I’ve been working with technical support at our web host to try to pin it down, with no luck. The error is somewhat frequent, but also random. It’s not the kind of thing I can force to happen, so if I make a change to the site, the only way to test and see if it fixed it is to wait.

I’m bringing this to your attention for two reasons:

  1. I don’t want you to think it’s just you, or that the website has been hacked, etc. We’ve got a cylinder misfiring here somewhere, and that’s the problem. We just have a lot of cylinders, so we have to isolate it.
  2. I would like to ask that you let me know if you continue to see this error. You can leave a comment on this thread (as simple as, “saw it just now”) or send us an email through our contact form, etc. Some of you have already been doing this, and it’s been a big help, because it doesn’t always happen when I’m using the site so I don’t know it’s still an issue until someone tells me, and then I know we have to try something else.

We’ve disabled a couple of the things behind the scenes on the website that might be causing the hiccup, so it’s my hope you just won’t see it again, period. Either way, I appreciate your patience and your help.

40 thoughts on “Bug Hunt”

  1. Just a random thought Steve – but are you load-balancing across multiple web servers or database servers? Maybe try to direct connections down one path to see if it is just a bad individual server?

    Reply
  2. I get that display when I go to the Canon212 website. Other websites, almost exclusively Catholic ones load very slowly or ‘hang’.

    Reply
  3. Interesting indeed. I’ve not had it happen yet and usually log onto your site within an hour or two of your posts.

    Reply
  4. Thanks Steve for the alert because I have been seeing it. Thought I was about to crash. Will alert you if it crops up.

    Reply
  5. Hi, Steve……I get it very randomly. Normally it is “there” and if I refresh the page, it is gone and the real you opens. Other times, like now, it did not appear at all. I thank you for your post as I was thinking it was my Mac.

    Reply
    • Are you living out there at the moment or just on a trip? I have family who have just re-emigrated from the place – hopefully I never have to go there again!

      Reply
    • It’s fine when using Windows 10 and Chrome on my computer. However, it’s different on my smartphone. When I click on the 1P5 icon, the bug report shows up at least once or twice. The second or third time, it’s OK. (And I use Chrome on my phone too.)

      Today (5/29, 11:57 am EDT) it’s OK.

      Reply
  6. I never seen the issue. I access the site from either a Windows 10 PC using Edge or a Windows 7 PC using Chrome. All the those who are seeing the issue should mention what type of computer they are using and what browser they are using. Maybe a pattern will emerge that will help you better figure out where the problem is.

    Reply
  7. The first error typically occurs when sending packets that are too large from the web end to the database server. The second error is simply because the first error was output before the application attempted to send headers, as the headers were already sent by the first error being output.

    I would check your database server configuration to see what the maximum packet size is and either look at increasing it or trying to debug exactly what WordPress is sending to cause a larger than normal packet.

    Reply
  8. Haven’t seen it so far Steve, and thanks for the info. However it will take a lot more than bugs to keep me from following you and read your excellent and highly professional articles. Thanks Steve, you’re in my prayers always.

    Reply
  9. Had it happen a couple days ago, but it eventually disappeared and I was able to hop on. I knew it was something to do with an error in the site, but wasn’t sure what and knew someone would be working on it. I am the farthest thing from a ‘Techy’ probably in the entire country, but knew that you guys had some kind of a glitch somewhere. Be patient Steve, someone will figure it out.:)

    Reply
  10. Since this is intermittent, I’d assume it’s an out of memory issue on the PHP side. I’ve seen this sort of thing happen on WordPress sites that have lots of plugins. Each plugin used adds to the render pipeline, and has a good bit of overhead. It’s best to minimize the number of plugins used, and for simpler ones, build the functionality directly into the theme.

    Of course, this could also happen with intermittent network issues to the DB, or a number of other causes.

    Reply
  11. I have occasionally seen this on my Win10 laptop, but don’t remember the exact times. It has always resolved itself after one refresh of the screen.

    Given the intermittent nature and difficulty in tracking down, this has all the earmarks of a resource limit being hit, such as memory size allocated to PHP or php-fpm (if used) or MySQL or nginx or apache, or something like the limit of database connections being reached. If you can, first increase the database connections limit and memory allocations to MySQL. Then add memory to PHP or php-fpm. Lastly increase the memory for apache or nginx.Recommended is using php-fpm instead of any apache PHP module as that allows better memory management.

    Reply
  12. I’ve seen it on quite a few occasions. Seems to come in phases. Last week I saw it on one day, then just gave up for the day but today, all’s fine again.

    Reply
  13. I’ve turned off some more site plugins to help with resource allocation. We’re running in a pretty robust environment and the servers are set up to handle our pretty heavy traffic. I haven’t heard any new reports since the latest round of shutoffs, so I’m hoping we have found the problem. (It could also be a PHP compatibility issue with one of them, because we got bumped up to a higher version than we were previously on.)

    I’m only tech savvy enough to be dangerous, so here’s hoping we squashed the bug…

    Reply
    • I haven’t seen it in a couple of days, but I haven’t been around that much either. Take that for what it’s worth.

      Reply
  14. Like others commenting here, the problem seemed to disappear several days ago, after it happening numerous times to me. I hope you were keeping notes on your fixes, because you seemed to have found the problem, and might be able to undo some’fixes’ that weren’t the cause. Thanks for the extra effort you put in to get the site back online!

    Reply

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Popular on OnePeterFive

Share to...