PliotronX
Diamond Member
I first noticed this phenomenon with software such as Sage Masterbuilder and Maxwell Estimation where mapping the requisite network drive using the IP rather than the hostname of the server is more stable, encounters less errors and timeouts. This is within two different companies and locations, very small with simple LAN's. No errors indicate any dropped packets, there is no way the networks are saturated to a point that DNS traffic is constricted.
As a result, I have since always mapped drives and even printers by IP and I see a lot less complaints this way. I cannot get my coworkers to do the same however, so I end up cleaning up their messes.
Anyway, the latest example is within a fairly large corporation with AutoCAD and Flexnet. Now it is possible there are transient TTL issues and whatnot at this company as they are in three different states with MPLS goodness and latencies vary from 40ms to 300-600ms. So I get a call about AutoCAD complaining about the license server, remote on, see that he was told to use the hostname of the license server which is located one state away. I punch in the IP of the license server and boom.
Why is DNS so finicky??
As a result, I have since always mapped drives and even printers by IP and I see a lot less complaints this way. I cannot get my coworkers to do the same however, so I end up cleaning up their messes.
Anyway, the latest example is within a fairly large corporation with AutoCAD and Flexnet. Now it is possible there are transient TTL issues and whatnot at this company as they are in three different states with MPLS goodness and latencies vary from 40ms to 300-600ms. So I get a call about AutoCAD complaining about the license server, remote on, see that he was told to use the hostname of the license server which is located one state away. I punch in the IP of the license server and boom.
Why is DNS so finicky??