Data Is Okay, Server Just Doesn’t Come Up
The failure of a server
does not necessarily mean that the data needs to be restored completely
from tape. Often, a server goes down because of a failure with the power
supplies, a motherboard failure, or even a processor failure. In a
situation where the hard drives on a dead server are still operational,
the hard drives should be moved to an operational server or, at the very
least, the data should be transferred to a different server. By
preserving the data on the drives, an organization can minimize the need
to perform more complicated data reconstruction from a tape restore,
which could result in the loss of data from the time of the last backup.
Restoring from tape should always be considered a final option.
This
type of a failure is an excellent example of where you can take
advantage of a Database Availability Group architecture. When utilizing a
DAG, the failure of a mailbox server doesn’t actually prevent the users
from accessing their email. When the master copy of the mailbox replica
fails, the second priority copy becomes the master copy, and
replication to any additional replicas continues. In this scenario,
there would be no need to restore the failed server. Rather, you would
simply build a new server from scratch at their convenience and add it
back to the replica list. When replicated, you can choose to make this
copy the master copy again. Similarly, if the failed system were a
client access server or a Hub Transport server, if there were more than
one server in that role in the site, the overall Exchange Server 2010
services would still be available when a single system failed. Rather
than being restored, this system could be rebuilt and would take up its
share of the load.
Data Is Corrupt—Some Mailboxes Are Accessible, Some Are Not
Data corruption
typically occurs on Exchange servers when the time period since the last
database maintenance is too long or maintenance has been neglected
altogether. Without periodic maintenance, the
databases in Exchange Server are more susceptible to becoming corrupt.
Exchange Server database corruption that is not repaired can make
individual messages or entire portions of mailboxes stored on an
Exchange server to become inaccessible.
When a mailbox or
multiple mailboxes are corrupt, the good data in the mailboxes can be
extracted with minimal data loss. By isolating the corruption and
extracting good data, an organization that might not need to recover the
lost data can typically continue to operate with minimal downtime.
Data Is Corrupt, No Mailboxes Are Accessible
Depending on the condition
of an Exchange Server database, the information might be so corrupt that
none of the mailboxes are accessible. Recovering data from a corrupt
database that cannot be accessed is a two-step process. The first step
is to conduct maintenance to attempt to repair the database; the second
step is to extract as much information from the database as possible.
In the case of a DAG
configuration, if the database became unavailable due to localized
corruption, you could simply make a different replica of the master copy
and services would be restored.
Exchange Server Is Okay, Something Else Is Preventing Exchange from Working
If
you know that the Exchange server and databases are operational and
something else is preventing Exchange Server from working, the process
of recovery focuses on looking at things such as Active Directory,
Internet Information Services (IIS), the domain name system (DNS), and
the network infrastructure, as with site-to-site connectivity for
replication.
Mail Is Not Flowing Between Sites
If users can access their
mailboxes normally and mail can be sent between users of the same site,
odds are the issue is with the Hub Transport server. In larger
implementations of Exchange Server 2010, the Hub Transport server role
is likely to be run on a system that doesn’t host mailboxes. Generally
speaking, backups are not performed on a Hub Transport server as it
contains no unique information. To restore these services, simply
rebuild the Hub Transport server. Installing with a /recoverserver
switch enables the server to recover its configuration from Active
Directory, saving some configuration steps. This assumes the server is
built with the same name.
If you need the transport
services up rapidly, consider adding the Hub Transport server role to an
existing system. To add this role, follow these steps:
1. | From an existing Exchange Server 2010 server, open a command prompt.
|
2. | Navigate to Program Files, Microsoft, Exchange Server, bin.
|
3. | Type exsetup.exe /mode:install /role:hub.
|
Internet Mail Is Not Flowing
If you cannot send mail
to the Internet or receive mail from the Internet, there is a good
chance that the issue is a failure with the Edge Transport server. Most
environments should run more than one Edge Transport server, preferably
in different locations. But if an Edge Transport server fails, it should
be rebuilt as they are typically not backed up. Installing with a /recoverserver
switch enables the server to recover its configuration from Active
Directory, saving some configuration steps. This assumes the server is
built with the same name.
If you need the
transport services up rapidly, consider adding the Edge Transport
server role to an existing system. To add this role, follow these steps:
1. | From an existing Exchange Server 2010 server, open a command prompt.
|
2. | Navigate to Program Files, Microsoft, Exchange Server, bin.
|
3. | Type exsetup.exe /mode:install /role:et.
|
Note
If
you place the Edge Transport role on a new system, you need to make
sure that incoming Simple Mail Transfer Protocol (SMTP) mail from the
Internet reaches this system. This might involve a change in
configuration of MX records, firewall rules, Network Address Translation
(NAT), or your antispam/antivirus gateway. Be sure you understand the
implications of putting the Edge Transport role on another system before
attempting this fix.