"Andy3691" <> wrote in message
news:9E95A106-23D4-4238-A49D-...
> Hope I am posting this in the correct catagory and that you folks can help
> clear the confusion. I have 2 DCs in my network, DC1 & DC2 with the FSMOs
> spead out over both DCs - forest MOs on DC1 and domain MOs on DC2. These
> are
> also my 2 DNS servers and the are both Catalog servers. I also have a
> DHCP
> server on DC1 for my client workstations. All my servers are manually
> configured with static IPs. I had DC1 freeze on me the other day and all
> network funtionalities came to a crashing halt - outlook could not find
> Exchange, no browsing, no VPN from outside, internet worked fine if you
> by
> IP address but not by resolving names . Luckly this happened after
> business
> hours so the mayority of users where oblivious to the outage. My
> confusion
> is that if I have 2 DCs also functioning as DNS with zone tranfers
> functional
> and up to date, why did'nt DC2 just take over automatically? My DHCP
> server
> has both DNS servers listed in it by IP - DC1 first then DC2. Can
> someone
> explain this process in laymans terms?
Good question. This question has been asked numerous times. It's not the
actual DNS server that "takes" over, rather it is a client-side function.
It's based on the client-side resolver algorith. Once a client has resolved
a query (such as to the DC, the Exchange server, or an internet address),
the client-side resolver will keep the answer in it's local cache for the
TTL (time to live) of the record. The TTL of a record is set on the record
in DNS, so the client can't change that. Therefore if one DNS goes down, and
the client already has that server cached in it's record and tries to
communicate to it (such as Outlook, etc), it won't query the other one.
Read up on it in more detail in my blog:
DNS, WINS & the Client Side Resolver, NetBIOS, Browser Service, Disabling
NetBIOS, Direct Hosted SMB (DirectSMB), If One DC is Down, Does a Client
logon to Another DC, and DNS Forwarders Algorithm
http://msmvps.com/blogs/acefekay/arc...algorithm.aspx
If Exchange is on a DC, no need telling you that if you search on it, you
will find numerous topics by many engineers (including Microsoft) stating
Exchange is not recommended to be installed on a domain controller.
Exchange's database transactional logging system is different than AD's.
Once a machine is promoted to a DC, it disabled the write-behind cache
function on the controller. Exchange needs this, however it's done to allow
AD's database system properly work. A huge drawback of this scenario is that
it can cause Exchange to lose emails during certain scenarios, as well as
with the write-behind cache disabled, it drastically reduces performance on
the machine.
Exchange by default, will also consume all memory resources, for example,
the store.exe process and will drag down the OS it is installed on. If the
OS is a DC, it will hinder DC processes, such as the DC's Lsas.exe process.
This *may* result in other issues, possibly with replication.
Read more on it:
This Exchange server is also a domain controller, which is not a recommended
configuration
http://technet.microsoft.com/en-us/l.../aa997407.aspx
--
Ace
This posting is provided "AS-IS" with no warranties or guarantees and
confers no rights.
Please reply back to the newsgroup or forum for collaboration benefit among
responding engineers, and to help others benefit from your resolution.
Ace Fekay, MVP, MCT, MCITP EA, MCTS Windows 2008 & Exchange 2007, MCSE &
MCSA 2003/2000, MCSA Messaging 2003
Microsoft Certified Trainer
Microsoft MVP - Directory Services
If you feel this is an urgent issue and require immediate assistance, please
contact Microsoft PSS directly. Please check
http://support.microsoft.com
for regional support phone numbers.