W
wilsonj
My scenario:
* I have a fax server solution that I need to make fully redundant.
* Cliets in the fax server solution only look at a set of file
folders.
* I need to the failover to be automatic
* I need the clients only to connect to the primary file folders
unless that server is offline. This because of licensing of the fax
software and because we get charged by the call on our backup PRI.
So this is my design:
DC1:
has DFS root \\domain\corp
DC2:
has replica of DFS root \\domain\corp
Primary FAX Server:
resides at our main network operations center
has DFS link \\domain\corp\fax
which contains client folders \requests - for fax request
\users - storage of users faxes
\system - contains shared address
book, coversheets, and letterhead
Backup FAX server:
resides at our co-location facility
has DFS link \\domain\corp\fax
which contains client folders \requests - for fax request
\users - storage of users faxes
\system - contains shared address
book, coversheets, and letterhead
DFS will be configured not to exclude any file types in replication.
I am expecting 2 GB of data to be stored in the link, but only 100MB
should be changed/accessed regularly.
I've set it up this way in the lab, and it has worked, but we were
only testing with small bits of data and one user.
We have very high bandwidth between physical sites (OC3 or better) so
my intention is configure all physical sites to exist in the same AD
site except the co-lo.
My concerns:
1) That replication will not be reliable
2) Clients will connect to backup DFS link when the Primary is
available
3) The large amount of data in the links will cause a problem.
4) If there is a failover, then returning to the previous state will
be difficult.
I've been reading a lot about DFS on the web and in the news groups.
And it has made me nervous about implementing this -- are there any
thoughts from the world?
Jason
* I have a fax server solution that I need to make fully redundant.
* Cliets in the fax server solution only look at a set of file
folders.
* I need to the failover to be automatic
* I need the clients only to connect to the primary file folders
unless that server is offline. This because of licensing of the fax
software and because we get charged by the call on our backup PRI.
So this is my design:
DC1:
has DFS root \\domain\corp
DC2:
has replica of DFS root \\domain\corp
Primary FAX Server:
resides at our main network operations center
has DFS link \\domain\corp\fax
which contains client folders \requests - for fax request
\users - storage of users faxes
\system - contains shared address
book, coversheets, and letterhead
Backup FAX server:
resides at our co-location facility
has DFS link \\domain\corp\fax
which contains client folders \requests - for fax request
\users - storage of users faxes
\system - contains shared address
book, coversheets, and letterhead
DFS will be configured not to exclude any file types in replication.
I am expecting 2 GB of data to be stored in the link, but only 100MB
should be changed/accessed regularly.
I've set it up this way in the lab, and it has worked, but we were
only testing with small bits of data and one user.
We have very high bandwidth between physical sites (OC3 or better) so
my intention is configure all physical sites to exist in the same AD
site except the co-lo.
My concerns:
1) That replication will not be reliable
2) Clients will connect to backup DFS link when the Primary is
available
3) The large amount of data in the links will cause a problem.
4) If there is a failover, then returning to the previous state will
be difficult.
I've been reading a lot about DFS on the web and in the news groups.
And it has made me nervous about implementing this -- are there any
thoughts from the world?
Jason