Steve Wiser wrote:

> We use Nagios to monitor all of our servers. It works pretty well, but
> we did have to write a small program for Nagios to check the 3050
> service. We just have it log into a database that it is monitoring and
> run a simple stored procedure. If it fails then we have it page us with
> a text message so we know what server, what database, and when it first
> failed.

That sounds good, but might be overkill in our situation. I must add
that these servers are not in our datacenter but reachable through the
internet, so I want to keep the traffic as small as possible. I could
also send a single ping, but I would prefer another solution...

regards,
Yves

> -steve
>
> On Tue, 2005-08-30 at 17:36 +0200, Yves Glodt wrote:
>
>
>>Hello,
>>
>>I want to make periodic checks whether an FB server is running or not
>>(the FB service and thus the box itself as well).
>>
>>Is there a way to connect to port 3050 and send a specifiy string or
>>packet and get a well defined answer (to see if server is alive or
>>not)?
>>
>>Or should I make a separate daemon that checks locally whether FB
>>turns
>>(ps -e | grep ...) and gives me my well defined answer on a tcp
>>socket
>>when I ask it for?
>>
>>Regards,
>>Yves