How to test the uptime of a webserver?

[Redd Vinylene]

On Thu, Sep 4, 2008 at 9:05 PM, Redd Vinylene <reddvinylene at gmail.com>wrote:

> On Thu, Sep 4, 2008 at 8:58 PM, Redd Vinylene <reddvinylene at gmail.com>wrote:
>
>>
>> On Sun, Aug 31, 2008 at 4:59 PM, Matthew Seaman <
>> m.seaman at infracaninophile.co.uk> wrote:
>>
>>> Redd Vinylene wrote:
>>>
>>>> On Sun, Aug 31, 2008 at 2:22 AM, Moises Castellanos <m2o7i1 at gmail.com>
>>>> wrote:
>>>>
>>>>>
>>>>> On Sun, Aug 31, 2008 at 7:05 PM, Redd Vinylene <reddvinylene at gmail.com
>>>>> >
>>>>> wrote:
>>>>>
>>>>
>>>  I got this dedicated server which is exposed to DDoS attacks quite
>>>>>> frequently. Say I need to host a website on it, is there any way of
>>>>>> telling how often it is actually online (to the rest of the world)?
>>>>>>
>>>>>> Maybe make some sort of ping script from a remote server?
>>>>>>
>>>>>
>>>     You can install nagios and monitor the web server. It will send you
>>>>> an
>>>>> email when
>>>>> the server is down and when is up again. With this information you can
>>>>> know
>>>>> the uptime
>>>>> of the web server.
>>>>>
>>>>
>>>  I'd have to install Nagios on a different server then, right? I doubt
>>>> the actual server knows when its ISP's link drops (or just slows down)
>>>> due to an attack.
>>>>
>>>
>>> Not necessarily.  You can install nagios on your web server and use it
>>> to monitor a server at the other end of your wan link -- usually a
>>> machine in your ISPs infrastructure[*] -- on the basis that if you can
>>> get packets out, then other people can get packets in.  The trick is to
>>> monitor something that isn't too far away, or you'll end up monitoring the
>>> availability of other people's networks, rather than your own.
>>>
>>> There's a lot more can be done than just monitoring connectivity by
>>> sending ICMP ping packets every so often.  There are any number of
>>> ways a web server can go wrong -- processes can crash, critical disk
>>> partitions can fill up, load spikes can overwhelm the machine's capacity.
>>> You can develop a range of different nagios tests that should tell you
>>> pretty much at a glance just what has gone wrong.  Takes all the fun out
>>> of diagnosing the problems perhaps, but it does mean you'll be back to
>>> bed sooner when the pager goes off in the small hours.
>>>
>>>        Cheers,
>>>
>>>        Matthew
>>>
>>> [*] Some ISPs provide machines specifically for this purpose.
>>>
>>> --
>>> Dr Matthew J Seaman MA, D.Phil.                   7 Priory Courtyard
>>>                                                 Flat 3
>>> PGP: http://www.infracaninophile.co.uk/pgpkey     Ramsgate
>>>                                                 Kent, CT11 9PW
>>>
>>>
>> Thank you guys. I'm looking for the simplest solution though, like a
>> simple oneliner, or a shell script.
>>
>> Anybody have an idea?
>>
>> --
>> http://www.home.no/reddvinylene
>>
>
> I'll try to write a simple shell script and report back to y'all.
>
> --
> http://www.home.no/reddvinylene
>

Perfection is achieved, not when there's nothing left to add, but when
there's nothing left to take away :)

while sleep 555; do wget http://<ip> -O /dev/null -t 1 || mail -s "<email>"
<<< "Host is down"; done

-- 
http://www.home.no/reddvinylene