Hi everyone,
I'm trying to fix a weird problem an I thought maybe it was a good idea to ask a question or two here. A little bit of context. We migrated our DNS servers from Win2003 to Win2008 R2 not long ago. After doing that, I noticed that DNS resolution on servers using UNIX/Linux based OS had some issues initiating connections from time to time when DNS was required.
Part of my troubleshooting steps brought me to do a traceroute from UNIX to the DNS servers and I found that packets were timing out at destination. None of this was happening on other Windows servers. Here's an example:
traceroute devpwdc01
traceroute to devpwdc01 (10.3.99.10), 30 hops max, 60 byte packets
1 10.3.200.250 (10.3.200.250) 0.731 ms 0.728 ms 1.083 ms
2 10.3.77.251 (10.3.77.251) 0.696 ms 1.410 ms 1.740 ms
3 devpwdc01.pptc.gc.ca (10.3.99.10) 0.380 ms * *
As you probably know, Linux and UNIX use UDP with traceroute as opposed to tracert which uses ICMP. If I use TCP or ICMP instead, none of this would happen.
traceroute -T -n 10.3.99.10
traceroute to 10.3.99.10 (10.3.99.10), 30 hops max, 60 byte packets
1 10.3.200.250 0.600 ms 0.891 ms 0.892 ms
2 10.3.77.251 0.884 ms 1.237 ms 1.565 ms
3 10.3.99.10 0.302 ms 0.305 ms 0.301 ms
Soon after, I found that all Windows 2008 servers where showing the same symptoms. I thought about routing issues or a problem with a switch but there's 2 factors that makes it unlikely
1. Windows 2003 servers are not showing any sign of this particular behavior
2. I've tested the traceroute command from different source and going to different destination, covering different switches and VLANs but no avail.
Anyone got an idea? I just can't seem to figure this one out!
Thank you.