[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [hobbit] http network test
- To: hobbit (at) hswn.dk
- Subject: Re: [hobbit] http network test
- From: "Ralph Mitchell" <ralphmitchell (at) gmail.com>
- Date: Wed, 30 Jan 2008 09:09:17 -0600
- Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:received:received:message-id:date:from:to:subject:in-reply-to:mime-version:content-type:references; bh=5X/yabh+7L4cIdtevLKDWVzOfYbZoaBv/7XlL22Nxdw=; b=bBY2GZu7MRNYj7yirJ6WiNrh2DlDxu691KvlUlh1l2EAUhhtQnaR04RITix5dHPOmi/V+DpCqiOXZ3b6aq4dSdH0ZYLuQts3bVKSb8ueQDTUF32WO/sHEGrkM+qUbJj8I/ADAeMwhYsYWAbSMaPEKNmZ8hCWGqVwjM8ZIQS1PO4=
- Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=message-id:date:from:to:subject:in-reply-to:mime-version:content-type:references; b=au8iSRW/zUSqokK9ytjPUOE69eFWCNKi/3+My96gq0ef3yQS8oOHHdEnk/GlhnUMWKw5xksxkW2BwcpRHUb7V0vo+ZN9/LCk6bKFgrRILJK7X5onDwF5q3xeZEhF+Y/PhlJggOa1M5IoSXdjGKzq5XCFtet6bUhC7xL+bTBX8eU=
- References: <997a524e0801291503j1fbf79c7i9c546e7189eedac5 (at) mail.gmail.com> <1201684661.16044.38.camel (at) localhost> <200801301135.51305.bgmilne (at) staff.telkomsa.net>
On Jan 30, 2008 3:35 AM, Buchan Milne <bgmilne (at) staff.telkomsa.net> wrote:
> On Wednesday 30 January 2008 11:17:41 Anna Jonna Armannsdottir wrote:
> > On þri, 2008-01-29 at 17:03 -0600, Ralph Mitchell wrote:
> > > I've got a few instances of
> > >
> > > a.b.c.d www.server.com # http://www.server.com/
> > >
> > > and the http column shows up just fine, including the time the test
> > > took and a nice graph.
> > >
> > > The question I was just asked is, is it possible to specify "max
> > > response time", so that if the time shown is greater than N seconds I
> > > get a red flag?? I don't see a way to set that in the man pages for
> > > either bb-hosts or hobbit-client.cfg.
> >
> > There is the possibility to use a per host test e.g.:
> >
> > a.b.c.d www.server.com # http://www.server.com/
> > badhttp-06-0000-2359:3:5:8 badhttp-12345-2200-0659:3:5:8
> > badhttp-12345-0700-2159:1:2:3
> >
> > This specifies the number of times the test has to be
> > yellow or red, in order to actually siglal yellow or red
> > (to the alert module- correct me if i am wrong here).
> >
> > I suggest that this may be seen as your time N (in seconds)
> > given by T*yellow and T*red . T is the time between tests
> > or retests.
> >
> > In the working hours, I get a
> > warning if http is down for T*2 or more, and an
> > alert if it is down for T*3 minutes or more.
> >
> > Some clarification may be called for here.
> >
> > The beauty of this, is that it can be configured per
> > time of day or day of the week as is done in the
> > example. BTW. I am actually using this. :)
>
> But, it doesn't allow you to change the timeout (e.g., http server 1 with
> slow
> application always takes 30 seconds to give content back, but http server
> 2
> must respond in less than 5 seconds to meet SLA) before the test is
> considered to have failed.
Yep, that's what I've been asked for. It's not exactly a timeout, because
that implies abandoning the connection. The server could still be
responsive, but slow - i.e. alive and talking to the public, while still
being in breach of SLA.
Ralph Mitchell