Unsuccessful Page loads (especially on first try)

I’m trying to deploy a stand-alone discourse docker instance, but I’m finding that I’m getting unsuccessful page loads (blank pages). When I load the page the first time as an admin, sometimes I see miniprofiler report a up to 10 seconds for DOM Loaded Event. Yet, once the site is cached, everything seems fine. I’ve actually found I cannot get a successful page load if I use firefox’s inspector, and hit the ‘disable cache’ button in settings.

Looking at the network profiler, I see huge load times for some files (though, it doesn’t seem consistent); sometimes it’s application-.js, othertimes it is vendor or en.

Similar effects happen if I try wgetting application js:

$ wget http://foo.internal.com/assets/application-1914811d144d39813a1f2689a90390b0.js
--2015-06-09 16:16:55--  foo.internal.com/assets/application-1914811d144d39813a1f2689a90390b0.js
Resolving foo.internal.com (foo.internal.com)...  x.y.22.219
Connecting to foo.internal.com (foo.internal.com)| x.y.22.219|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 3218361 (3.1M) [application/javascript]
Saving to: ‘application-1914811d144d39813a1f2689a90390b0.js’

 0% [                                                                                                                             ] 13,455      --.-K/s  eta 2h 8m

The transfer is completely stalled.

I’m sure this is a local problem, however, I’m not seeing any thing in the logging that indicates any way to pursue this. Any suggestions?

Sounds to me like there are some odd networking configurations in place…

  1. Do you have iptables custom rules? Can you disable?
  2. Can you repro the issue from local? (in case its a network proxy causing the pain)
1 Like

:facepalm: I totally forgot to check the firewall. Local wget also works 100% and is screaming fast as expected.

Issue reproduces even after doing iptables --flush on the Ubuntu 14.10 host, so I’ll open a ticket with Company Support to see what’s going on.