Views Page Trouble Shooting

We encourage users to post events happening in the community to the community events group on https://www.drupal.org.
LittleCoding's picture

Very odd issue with a rather simple views page I am having and I have run out of trouble shooting tricks to get to the root of the problem.

The page list the locations within a region based on the location page being in a external DB and having an active page within the Drupal DB. But with the most populated region the page fails to load. only error message I have managed to pull out is "database has gone away". The part that is making this so odd is that fact I am unable to reproduce this issue on any other server (local machine, standard development) even with the same data set and all files from the production servers. The query works great for all other regions and is working in the Views admin panel preview of the page. No other pages using the same data in a verity of ways is misbehaving. The page errors out in the same amount of time as the page would regularly take to load. And with reviews of the Drupal, MySQL, Apache logs I still have only managed to find the one error message in the logs.

So I am hoping for a few new angles to try and trouble shoot with, who has some ideas?

Comments

Hey Rob, yeah this is a weird

PierreMarcel's picture

Hey Rob, yeah this is a weird bug. I have never had an issue like that, but usually when it says "database has gone away" it means it has lost connection to the server. So maybe because your region has lots of others things loading in the queue your custom external DB query is timing out... try to increasing your php.ini settings may help. Since it's working on other env I'd say it's some minor settings you have diff on you prod env.

The prod env config is where

LittleCoding's picture

The prod env config is where I am ended up. The page will work up to 54 results or paginated to 54, as soon as it goes to 55+ the page false in fast order. The query still works in the admin view preview from views though.

I'm also looking at the php version. 5.3 is running on the prod env and 5.5 on the others.

It is not acting like a timeout error with the connection as the page fails in the same time frame was the pages on the site regularly load. No prolonged pinwheel.

And we have a winner. Updated

LittleCoding's picture

And we have a winner.

Updated to PHP 5.5 = nothing
php.ini memory_limit upped to 2GB = nothing
my.cnf max_allowed_packet upped to 32MB = WINNER!!!!

Good Stuff Rob, oh man I

PierreMarcel's picture

Good Stuff Rob, oh man I actually was going to say max_allowed_packet... lol I'm glad you were able to fix it. Yeah when things like that start happening the best is always check the versions and settings of your env on all machines. I've had few weird issues like that, so now I'm running DrupalVM so I config it just like the server where the prod env is. So if things go wrong they go wrong on both...