504 Gateway Timed Out


#1

I have a web page with a form where the “Submit” button triggers a CGI script on draco. It was always working fine, until Saturday. Now it gives me a “Gateway Time-out” message. (“The gateway did not receive a timely response from the upstream server or application.”)

I did some research on this, and discovered this web page, which says “This problem is entirely due to slow IP communication between back-end computers, possibly including the Web server. Only the people who set up the network at the site which hosts the Web server can fix this problem.”

http://www.checkupdown.com/status/E504.html

Dreamhost’s support technicians told me the script was still working fine when they tested it, but I pointed out that they were submitting very short data, and with more data the server seemed to “time out” before completing all the calculations that it used to. They then told me “Your users processes aren’t triggering our process monitor.” (Which seems weird, because if the script is running and accepting their test input and producing a web page, then that process should be visible to Dreamhost someplace).

I’m really just looking for a solution. Regardless of the process monitors, my theory is the server now “times out” too soon – that is, before completing all the calculations that it used to perform with no problems. Since it just started happening this weekend, I’m theorizing that it’s related to the way the new hardware was set up after the work that was done on draco.

Dreamhost ended their e-mail by saying “troubleshooting your script is beyond our general scope of support.” (Which I understand – but it’s Dreamhost’s servers which have suddenly stopped RUNNING the script.) Has anyone encountered similar problems, and/or do you have any suggestions? (Er, and is this a problem which only the network administrators can handle?)


#2

Is draco still on the injured list? Now that dreamhost only posts server issues to affected customers via the panel I can’t tell. But recently, from information obtained via other posts here, draco has been recovering from a hardware failure.

Sadly there is a pattern that can be established during these hardware failures (which typically are attributed to a bad RAID card). They last between 7-10 days generally and support always replies that your site is working when they tested it, or they move your support request to a queue for special assistance or some such thing where it stays until the server has recovered.

Dreamhost really needs a better answer for these issues. 7-10 day outages just don’t fly, maybe some sites work but there is enough complaint in this forum for me to know it’s a recurring problem.


#3

Thanks, Lakerat!