[cifs-protocol] [REG: 111060784988397] RE: joining domain with IPv6 client

Edgar Olougouna edgaro at microsoft.com
Thu Jun 16 10:30:53 MDT 2011

Hi Tridge,

Just a quick update on this. Hongwei will continue digging through and compare to Windows to Windows behavior.
Like I mentioned yesterday, at one point the client is not being able to resolve the name "BLU" with type=28 (AAAA).
DNS QueryIdentifier == 0x74c5 is timing out and does not show at the server side. Windows client is sending that request on ff02::1:3 which is the link local multicast address.

All the other DNS queries that are successful are going to fdb1:2726:40f9:b3b7::1.

What I observed so far is that it goes through the single name dual address resolution, we should try type=1, and fail with DNS_ERROR_RECORD_DOES_NOT_EXIST. Instead we are timing out, even after one retry on the ff02::1:3  address. Later we bypass AAAA query.
BLU does not match the cache, so we ended up with  status = ERROR_INVALID_NAME (123), this later becomes WSANO_DATA down the call flow.

I am looking at ipv6-4.zip.
Go to Time Travel Position: 455BC0000057. Please use the DNS ETW as well, it has a lot useful information.


More information about the cifs-protocol mailing list