10

Every time I try to open a local application through Pow I am greeted with the following error (* = numbers):

Error: timeout: waiting for /tmp/nack.*.*.sock
    at null._onTimeout (/Users/pain/Library/Application Support/Pow/Versions/0.4.1/node_modules/nack/lib/process.js:416:29)
    at Timer.listOnTimeout [as ontimeout] (timers.js:110:15)

After refreshing a few times the app starts and works just fine. But it is kind of annoying to do every time after a was away for a while.

Is this the expected behaviour? Is there a way to increase the timeout?

Update:

$ time rails runner "puts 1"
4.72s user 1.32s system 28% cpu 21.198 total # first time after reboot
3.73s user 0.68s system 99% cpu 4.424 total # after that

Update:

In fact the same thing applies to Sinatra app that starts pretty quick by itself. And the timeout message being displayed almost instantly, doesn't look like there is an actual timeout at all. Then on the second page refresh it works fine.

Update:

Just had to kickstart a semi-static Sinatra app like 5 or 6 time before it showed up. What gives? The app is almost instantaneous.

firedev
  • 20,898
  • 20
  • 64
  • 94
  • It shouldn't be but maybe your app is going up veeeery slowly. Some heavy initializers, external db or so? – Mike Szyndel Jun 30 '13 at 20:12
  • I don't know, it's nothing much, local DB and everything, once it "heats up" everything is almost instant. Plus it's not just one app. Granted I am not running this on SSD, but I have a Core i7 Mac Mini Server with hard drives in a RAID so it's pretty snappy. – firedev Jul 01 '13 at 06:21
  • And if you run `rails s` how log does it take? – Mike Szyndel Jul 01 '13 at 06:45
  • I have added Rails startup time to my question. Granted 20 seconds is a bit lenghty wait, but nothing out of ordinary for a Rails project. Is there a way to fix the timeout? – firedev Jul 03 '13 at 02:46
  • I just checked for github issues, you may have seen it https://github.com/37signals/pow/issues/383 there's no resp so far... – Mike Szyndel Jul 03 '13 at 06:41
  • Thanks, I thought there is a solution for that, but seems the devs should just up the timeout on their side. – firedev Jul 03 '13 at 10:26

0 Answers0