The "Bad Gateway" error you are getting from NginX is caused by concurrency setting mismatches between NginX and whatever you are running on the back-end (Apache with mod-php, or php-fpm?). NginX is reporting this error because the back end is busy and cannot accept any more incoming connections.

Make sure that the NginX config and whatever you are using for back-end PHP processing have the same concurrency settings.

Signed,
-Smokin'

Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.

Understood,
-99% of this post's readers

Seahawk Sailor wrote:Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.

Understood,
-99% of this post's readers

Not sure what you just said. It's all Greek to me.

All I know is that the front end and the back end aren't properly tuned for high numbers of concurrent users. The front-end is configured to handle more connections than the back-end, so once it overflows, the "bad gateway" error is churned out.

SmokinHawk wrote:Not sure what you just said. It's all Greek to me.

Bingo!

(but it's actually Latin)

Seahawk Sailor wrote:

SmokinHawk wrote:Not sure what you just said. It's all Greek to me.

Bingo!

(but it's actually Latin)

I know. Was trying to make a funny.

Rockhawk and Anguish were talking about this problem in the mod forum today. Trying to plan long term what to do about it.

They are planning on making the next move in time for draft day which till now has been the most posters at once.

Over the weekend I saw over 2,800 posters at once just on the forum plus what ever more in the chat room, which is seperate.

Think Sailor saw even more.

The Radish wrote:Rockhawk and Anguish were talking about this problem in the mod forum today. Trying to plan long term what to do about it.

They are planning on making the next move in time for draft day which till now has been the most posters at once.

Over the weekend I saw over 2,800 posters at once just on the forum plus what ever more in the chat room, which is seperate.

Think Sailor saw even more.

Same but different. Smokin was right, a couple config tweaks needed to be made. Unfortunately I was in an aeropuerto in the middle of nowhere Mexico yesterday afternoon.

Seahawk Sailor wrote:

SmokinHawk wrote:Not sure what you just said. It's all Greek to me.

Bingo!

(but it's actually Latin)

Actually, it's not really latin. It's mostly gibberish. If I'm remembering right, this is the passage they use to test typefaces and spacing and stuff, correct? There's quite a few Latin words in there, but quite a few non-Latin words, too.

Anguish wrote:

The Radish wrote:Rockhawk and Anguish were talking about this problem in the mod forum today. Trying to plan long term what to do about it.

They are planning on making the next move in time for draft day which till now has been the most posters at once.

Over the weekend I saw over 2,800 posters at once just on the forum plus what ever more in the chat room, which is seperate.

Think Sailor saw even more.

Same but different. Smokin was right, a couple config tweaks needed to be made. Unfortunately I was in an aeropuerto in the middle of nowhere Mexico yesterday afternoon.

Webserver tuning can be such a pain in the ass. I host a couple high traffic community sites and use NginX to serve static content while Apache (via mod-php) handles the PHP, which is what I suspect you are doing here (perhaps php-fpm instead of Apache). APC is used to cache php opcodes while memcached is employed to store session data such that multiple web servers can be implemented and subsequently load balanced.

In all of that configuratin', the hardest thing to tune was PHP concurrency settings as setting values too high will cause you to run out of RAM in high load situations (very bad as it causes paging to disk and horrible slowness while disk queues lengthen) while setting it too low will cause server busy or gateway unavailable issues. php-fpm is a lot better than mod-php in terms of memory usage, but can bottleneck as it relies on sockets or TCP for piping data in and out.

If the site is presently hosted on a single server solution, moving up to a multi-server solution will cost a pretty penny (slightly more than double the cost unless "smaller" virtual servers are employed), and will require a lot of mojo to configure properly (multi-master database replication, distributed PHP session data, rsyncing of user data, etc). If you guys want any recommendations or an extra brain to help with some design issues, by all means feel free to ask. I'm a busy guy these days, but I have no problem advising on a solution.

Latin has always been Greek to me... even when I was an Altar Boy.
Yes, Rocket was a little angel.