500 Internal Server Error
-
The plugin has been functioning flawlessly since our site launched back in March. However, as of yesterday the home page that uses the FTS shortcodes consistently creates a 500 Internal Server Error unless the plugin is deactivated.
Any thoughts as to what may be going on? We hope this can be resolved as we are seeing a large discussion across social media due to a company campaign that started just yesterday.
Thank you in advance for any support you can provide.
-
Hi, I happen to work for DreamHost ??
So a few things are at play here, and it makes it a total PITA for everyone. And none of us work for Instagram, who is screwy as all get out :/ I noticed that the plugin was updated 5 days ago (and before that, 3 weeks to fix … Pinterest it looks like). But none of that seems related.
I do plugin reviews for .org ?? I’m pretty good at that kind of check ??
On the DH side… We haven’t made any rate limiting changes, but if we had it would should up as a 418 – I’m a little teapot – error, which would relate to our ModSecurity firewalls. That it’s an error 500 means it’s a ‘pure’ PHP error. Which is not super helpful, I know.
What HAS happened lately is we force upgraded people to PHP 5.5 and 5.6. And some people HAVE been moved to new servers. Both of those could be in play here, since the new servers are on Ubuntu and weird stuff has happened.
One thing we’ve found in error logs that seems relevant is this:
PHP Fatal error: Maximum execution time of 30 seconds exceeded in /REDACTED/wp-content/plugins/feed-them-social/includes/feed-them-functions.php on line 1890
30 seconds seems incredibly short. I also see a lot of this:
Premature end of script headers: admin-ajax.php
My best guess, and I’ve only just started digging into this one, is that we’re hitting a random PHP error simply due to being moved and the PHP details changing.
I would immediately check what version of PHP you’re on (go to Panel -> Domain -> Edit domain, scroll down to PHP). Make sure you’re on at least PHP 5.6 FastCGI but PHP 7 would also be something to check.
Also if you WERE moved, and you happen to have a custom php.ini/phprc file, disable it for now. See if that helps. You may have some old code in there which is unhappy with the new location, or simply missing.
Glad to see someone from Dreamhost weighing in on this thread. However, I have to say… the simple fact that this is not happening at all with any other hosting providers says something.
Support at DH has said something different every time I have been in touch (regarding multiple websites) with this error but ultimately every tech says contact the plugin developer. Which as you can see we all have been harassing Spencer for days now. ??
the simple fact that this is not happening at all with any other hosting providers says something.
You know how no two cars, even the same make/model, are alike becuase of how different people drive? Same thing.
And I’m not trying to be dismissive! This is just kind of a ‘Well … yes?’ factoid here for us. We know it’s ‘just us’ but it’s also ‘just this plugin’ so .. who is it? The trick with debugging is to rule things out.
Is this possibly an issue where this plugin + DreamHost servers/network aren’t friends? Yes of course! But from our end it would be “Well your plugin worked and now it didn’t so what changed on the plugin?” since I don’t see anything in your particular case.
I looked at
plantsomethingma
directly.* You were moved to a new server on 2015-11-15 so I doubt that’s related
* You’re on php 5.5
* You have a custom phprc file
* The first issue was reported 10 days ago, which is shortly after the last code update to the plugin
* Process watcher is killing youI checked something in wp-cli and got this:
Fatal error: Call to a member function get_options() on a non-object in /home/REDACTED/web/wp-content/plugins/simplemap/classes/maps.php on line 110
I didn’t SEE that you’d checked with wp-debug on (WP_DEBUG: Disabled is all I see about it) so maybe turn that on and see if WP spits out errors even with instagram off?
Also I would do the following:
* Make sure EVERYTHING is upgraded
* Turn off your phprc
* Switch to PHP 5.6
* Turn on wp-debug (you can save it to a log if you want – https://codex.www.ads-software.com/Debugging_in_WordPress#WP_DEBUG_LOG )
* Test with all your OTHER plugins disabled.That the issue started 10 days ago and the plugin was updated (codewise) 3 weeks ago seems to imply that it’s not the code that changes, but what ELSE on yoru site updated? Other plugins? The theme?
These three plugins updated on the 16th:
simple-lightbox
social-share-buttons-by-supsystic
video-embed-thumbnail-generatorI’d start with disabling them.
I disabled all plugins (except for Feed Them Social) and switched the theme to Twenty Sixteen and encountered the same error multiple times when I first began troubleshooting the error.
The phprc was setup by Dreamhost once all of this started as one support tech recommended it.
The three plugins in question may have been updated on the 16th but this error began over the weekend on the evening of the 14th or morning of the 15th (I am not certain).
I appreciate your input and wish I could have been in direct contact with someone like yourself from the get go. However, I have spent too much time already troubleshooting this as I won’t charge the client for something out of their control.
We now have alternate feeds in place that are working just fine. I simply plan on avoiding the use of the Feed Them Social plugin on any site hosted by Dreamhost. However, I am still notified of updates on this thread and will continue to comment.
What is your email address Mika, so I can be in direct contact next time something goes awry!? ??
mika.epstein at dreamhost.com ??
I’m actually paid to be a WP guru there so this is totally in my wheelhouse.
(Spencer, if you want to ping me too, I can get you set up with a test site and maybe we can figure out why this is so derpy!)
Oh and FWIW, if process watcher is killing things, then maybe it’s really just a timeout trying to talk to Instagram (and sometimes Facebook) that is spinning up an endless loop? I don’t know where you’d want to put a better timeout, but Spencer, is the code checking to see if it can talk to these services before trying to query the data?
In theory you could just do a quick API availabilty ping (IIRC Jetpack does this, which is why sometimes you have to reconnect publicize) before loading everything else?
@mike appreciate your input. Wish more of the DH team was as knowledgeable as you are, Sir!
I’m no expert, but I’m having the same issue with “error 500” and the site I’m working on is hosted with Go Daddy. I installed the plugin, it worked for a couple of minutes, and then I started getting “Error 500”. I don’t currently have access to the FTP for the site so I can’t check error logs.
I’ve been messing around to try and fix this through the settings. I clicked on “Facebook Options” under the “Feed Them” tab. I changed the “Change Post Limit Quantity” to 7, and instantly the Error 500 stopped. I will let you know if it comes back, but maybe try this settings change.
Darn it – it worked three or four times, and then I got another “Error 500”. Maybe it wasn’t the specific change to the settings, but somehow just making a change?
@jacob.cowan Thanks for inputting too! Is it only the facebook feed causing this problem. Can you try the instagram feed or twitter feed and see if they give you troubles by themselves?
Spencer
@mike, sent you a follow up email to get the ftp url for the test site. Can you hit me back with that?
Spencer
@mike nevermind I got it, thanks again. I’ll keep everyone posted here If I find a solution.
Spencer
I can’t at this time try Instagram or Twitter. I don’t have access to the API keys for either account. It’s my brother’s band’s website, and I think it might be someone else in the band who set up the accounts. I tried again today. It worked the first time I clicked the link to the page I have the shortcode on, and the second time it did not work.
- The topic ‘500 Internal Server Error’ is closed to new replies.