<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
i guess in reddit case they were dealing with millions of requests, the
system in going to buckle under that kind of demand. few choices when
you have low mem, a hung server, shitty code (opps did I say that) or
errant process, reboot or restart seems the best immediate solution to
maintain a certain quality of service. this helps fight the fire while
you're trying to figure out what is going wrong, if there is a pattern
is will be easier to figure out than those quirky ones that just show
up randomly.<br>
<br>
you're right it should be fairly easy to write a script to monitor some
of the things and send out an alert or take some action, i am not sure
exactly what data can be pulled by those tools for a script to query a
system. i don't want to run and maintain a lot of scripts that are
using more cpu to query, parse and poll said services if there is a
better solution out there, something event driven rather than
poll/parse driven.<br>
<br>
this is really a good education exercise for me, it's helping me learn
more linux from a sys op point of view than from a hacker point of view
=)<br>
<br>
kind regards,<br>
Rajinder<br>
<br>
<br>
<br>
<br>
On 10-08-01 06:01 PM, aaron d wrote:
<blockquote
cite="mid:AANLkTink-w-wjRcm6zNqM7N-1sBpiT0Ac9B0PXEhohEj-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org"
type="cite">If you can program it should be fairly trivial to script
custom monitoring using the aforementioned tools, that takes the action
of your choice. Reboot though, as stated before, is a bad habit to get
into and generally NEVER the answer. We aren't running windows here :)
<div><br>
<div><br>
<div class="gmail_quote">On Sun, Aug 1, 2010 at 5:47 PM, Rajinder
Yadav <span dir="ltr"><<a moz-do-not-send="true"
href="http://devguy.ca">devguy.ca</a>@<a moz-do-not-send="true"
href="http://gmail.com">gmail.com</a>></span> wrote:<br>
<blockquote class="gmail_quote"
style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<div bgcolor="#ffffff" text="#000000">i know some of those tools,
not an avid user of them, will read up.<br>
<br>
as for rebooting or restarting errant process, i was thinking of this
in terms of my own web-server hosting setup, in particular my own rails
apps, since i would be logging error, i would know what was causing any
kind of abnormal memory, cpu, network load, it would be a stop gap
solution till the real issue was rectified in the code or through load
balancing, etc.<br>
<br>
i heard a talk from the reddit founder talking about fail often, crash
early and they were using some of this monitoring techniques, it was
interesting and something to mull over for sure =)<br>
<br>
thanks,<br>
<font color="#888888">Rajinder</font>
<div>
<div class="h5"><br>
<br>
<br>
<br>
On 10-08-01 08:19 AM, aaron d wrote:
<blockquote type="cite">iftop, iostat, vmstat, free, df ? they
aren't monitors in
the sense that you are talking in the second half of your question
(i.e. they don't take action, they simply tell you what is going on).
In my experience with monitoring systems, you generally would not want
something just killing off random PIDs at it's own discretion, let
alone REBOOTING because of a memory leak. We never reboot :) Usually
you would want to receive some form of alert so that you can use these
tools to investigate the issue and possibly take steps to keep it from
happening over and over again. However there may be a time when you
just want to kill a notoriously bad process (google chrome anyone?) if
it gets out of hand.
<div>
<div><br>
</div>
<div>conky (<a moz-do-not-send="true"
href="http://conky.sourceforge.net/" target="_blank">http://conky.sourceforge.net/</a>) is
not
a console-based monitor, but is very nice, and of course if you are
monitoring remotely you could always forward X to your local machine.</div>
<div><br>
</div>
<div>-aaron</div>
<div><br>
</div>
<div>
<div class="gmail_quote">On Sat, Jul 31, 2010 at 6:01 PM,
Rajinder
Yadav <span dir="ltr"><<a moz-do-not-send="true"
href="http://devguy.ca" target="_blank">devguy.ca</a>@<a
moz-do-not-send="true" href="http://gmail.com" target="_blank">gmail.com</a>></span>
wrote:<br>
<blockquote class="gmail_quote"
style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">Other
than
top, that's as far as my linux-fu gets me =P, what other console
based monitors are you guys using that has low cpu load for things like:<br>
<br>
network activity, file activity, storage space, low memory<br>
<br>
if the system is overloaded, i'm memory leaks, errant process is there
a monitor that will force the server to reboot? i am asking more for an
education purpose and not any kind of immediate use<br>
<br>
Thanks,<br>
Rajinder<br>
<font color="#888888">--<br>
The Toronto Linux Users Group. Meetings: <a moz-do-not-send="true"
href="http://gtalug.org/" target="_blank">http://gtalug.org/</a><br>
TLUG requests: Linux topics, No HTML, wrap text below 80 columns<br>
How to UNSUBSCRIBE: <a moz-do-not-send="true"
href="http://gtalug.org/wiki/Mailing_lists" target="_blank">http://gtalug.org/wiki/Mailing_lists</a><br>
</font></blockquote>
</div>
<br>
</div>
</div>
</blockquote>
<br>
</div>
</div>
</div>
</blockquote>
</div>
<br>
</div>
</div>
</blockquote>
<br>
</body>
</html>