Kind of a chicken & egg problem, seems like. Your network guy is right as far as that technically goes, but I'm with you. If DNS goes down, that needs to be the first order of business, since your business will start grinding to a halt anyway. I'd feel silly, pinging a server and not know that DNS was failing.
On Tue, Jun 1, 2010 at 12:55 PM, David Lum <david....@nwea.org> wrote: > +1, I've smoked my Service Desk guys on that EXACT error before (not that > I've ever done the same bonehead thing myself to burn this into my head) > > Setting up monitoring dependencies follows the same thing - no need to PING > test a remote server if you can't ping a the local switch, or the remote > router, etc. > > Which brings up a question as I've had this debate with my network > architect. He says when monitoring servers to ping by IP instead of hostname > "in case DNS goes down". My point is you should be testing for that > infrastructure anyway so ping by name doesn't get triggered unless DNS > functionality (also tested for) is working. I'm of the "test as you operate" > so if clients connect by hostname, then test by hostname. If only IP addr is > used, then use that. Same for websites, etc. > > Would LOVE to see a whitepaper recommending one way or another. > > Thoughts? > > David Lum // SYSTEMS ENGINEER > NORTHWEST EVALUATION ASSOCIATION > (Desk) 971.222.1025 // (Cell) 503.267.9764 > > > > -----Original Message----- > From: Erik Goldoff [mailto:egold...@gmail.com] > Sent: Monday, May 31, 2010 6:29 AM > To: NT System Admin Issues > Subject: RE: Veering even more OT - was: Re: Big Changes Ahead for IT - > Anyone seen this? > > +1 > Back in the NT 4.0 days when interviewing candidates I'd ask them the first > thing they'd check if a user could not login due to a 'domain controller > cannot be found' type error. > Amazing how many would jump directly to the more 'sophisticated' layers, > check domain controller, IP Stack, WINS, etc .... > To me the ONLY correct answer for the FIRST thing to check is: Check the > Ethernet cable ! ( in my experience over 90% of these type errors were from > the ether net cable either being unplugged or damaged ) > > > Erik Goldoff > IT Consultant > Systems, Networks, & Security > > ' Security is an ongoing process, not a one time event ! ' > > > -----Original Message----- > From: Ken Schaefer [mailto:k...@adopenstatic.com] > Sent: Saturday, May 29, 2010 9:23 PM > To: NT System Admin Issues > Subject: RE: Veering even more OT - was: Re: Big Changes Ahead for IT - > Anyone seen this? > > Normalisation is used for data integrity not efficiency. > > And whilst there aren't many practical implementations of OSI, the concept > of a layered approach to networking (physical link, node addressing, > routing, session control) is very useful in design and diagnosing problems. > > Cheers > Ken > > > ~ Finally, powerful endpoint security that ISN'T a resource hog! ~ > ~ <http://www.sunbeltsoftware.com/Business/VIPRE-Enterprise/> ~ > > > > ~ Finally, powerful endpoint security that ISN'T a resource hog! ~ > ~ <http://www.sunbeltsoftware.com/Business/VIPRE-Enterprise/> ~ -- David _____________________ Are you better off than you were four trillion dollars ago? ~ Finally, powerful endpoint security that ISN'T a resource hog! ~ ~ <http://www.sunbeltsoftware.com/Business/VIPRE-Enterprise/> ~
