Print

Print


> 
>> Do a traceroute that has a few hops that haven't set up their PTR
>> records properly.  (Or in a VRF with no DNS, etc).  You will wait a
>> while for each hop to resolve.  You are correct in that it doesn't
>> slow the packets down on the network.  However, it greatly delays
>> the time between packets.  It changes from a few ms to several
>> seconds per hop (default timeout can be changed with "ip domain
>> timeout").  Add this up for a couple of hops and you must be really
>> slow at copy/paste for it to be worse.
> 
> A) if the DNS server is operational but without PTR records, it will
> return immediately.
> 

It is not unusual for the DNS server somewhere in the chain to be non-operational.
Especially when you typo a Cisco command and it starts trying to open a telnet session.

> B) if they are lame delegations then the provider should fix it and
> not leave it broken.
> 

In an ideal world, yes, but, not a big help when you're trying to troubleshoot something and usually outside of the control of the person doing the traceroute.

> C) 5ms is always faster than 2 seconds to copy, change 'windows',
> click your edit box, paste, return (or click submit).  Or if you are
> me, I'd queue up a dig or nslookup in a window then play the copy and
> paste game and hit return - no finding the edit box, and the time
> wasted will still be wasted.
> 
True, but, the DNS timeout I experience from these issues is usually more like 30-90s and 2s is always faster than 30s or more.

> Why your opinion MUST override mine baffles me.
> 

I can't speak for the other guy, but, I'm trying to discuss the issues and ignore the
personal offense/attacks and bickering. I'm not saying my opinion or experience
should override yours. Just pointing out the issues from my experience with the
mechanisms and reasoning you have put forth.

Owen

########################################################################

To unsubscribe from the MICE-DISCUSS list, click the following link:
http://lists.iphouse.net/cgi-bin/wa?SUBED1=MICE-DISCUSS&A=1