Dreamhost works fine for the apps I have. I have to use a modified
dispatch.fcgi file so my app stays stable (no more 500 erros), and I use
cron jobs to keep my applications alive.
Hogan, Brian P. wrote:
Dreamhost works fine for the apps I have. I have to use a modified
dispatch.fcgi file so my app stays stable (no more 500 erros), and I use
cron jobs to keep my applications alive.
I’ve been using cron jobs as a keepalive, as well. Are you using wget
or curl or ? With wget, I cannot seem to keep it from making an output
file, despite using the switches to disable output.
On Feb 15, 2006, at 5:10 PM, Chris S. wrote:
–
Rails mailing list
[email protected]
http://lists.rubyonrails.org/mailman/listinfo/rails
Try this:
wget http://example.com/foo/bar 2>&1 > /dev/null
Cheers-
-Ezra Z.
WebMaster
Yakima Herald-Republic Newspaper
[email protected]
509-577-7732
Ezra Z. wrote:
On Feb 15, 2006, at 5:10 PM, Chris S. wrote:
–
Rails mailing list
[email protected]
http://lists.rubyonrails.org/mailman/listinfo/railsTry this:
wget http://example.com/foo/bar 2>&1 > /dev/null
Cheers-
-Ezra Z.
WebMaster
Yakima Herald-Republic Newspaper
[email protected]
509-577-7732
Hmm, seems like no matter what I do I either have output files
generated, or I get emails from cron.
The full line is something like this:
*/5 * * * * wget HugeDomains.com 2>&1 > /dev/null
Any thoughts would be greatly appreciated 
Chris S. wrote:
Hmm, seems like no matter what I do I either have output files
generated, or I get emails from cron.The full line is something like this:
*/5 * * * * wget HugeDomains.com 2>&1 > /dev/null
Any thoughts would be greatly appreciated
I’ve have to reverse the redirections on some systems for it to work
properly in cron. Not sure why. Anyway try:
*/5 * * * * wget HugeDomains.com > /dev/null 2>&1
-a
At 2/16/2006 11:06 AM, you wrote:
I’ve have to reverse the redirections on some systems for it to work
properly in cron. Not sure why. Anyway try:*/5 * * * * wget Scharfie.com is for sale | HugeDomains > /dev/null 2>&1
-a
The redirections are processed when encountered so “2>&1” means
“connect STDERR to whatever STDOUT is right now” which is probably
being captured by cron and then you change STDOUT to go to /dev/null.
cron probably acts something like:
(command) >watchoutput 2>&1 <&-
if [[ -s watchoutput ]]; then mail user < watchoutput; fi
When you put your actual command in and expand what the shell will do
in the order that it would appear to happen, you’d get:
wget Scharfie.com is for sale | HugeDomains >watchoutput 2>&1
<&- >/dev/null 2>&1
send stdout to watchoutput
send stderr to where stdout goes (and that’s watchoutput)
close stdin (it might be closer to “</dev/null” than “<&-”, but the
point is the same)
call the user’s command
send stdout to /dev/null
send stderr to where stdout goes (and that’s /dev/null)
wait for command to finish
see if there’s any output in watchoutput and email it to user
-Rob