F
Fari Fuladi
Hi there,
I am using the Dfs+FRS services of W2K server, but the
results that I am getting is not quite consistent with the
Dfs documentation. I am wondering if anyone has
experienced this problem before:
My setup includes 2 sites:
site 1: DC and foo (both W2K servers)
site 2: bar_1 and bar_2 (W2K servers)
The Dfs root is hosted by the domain controller, and my
only Dfs Link is set up such that it points to physical
locations on servers foo (Primary), bar_1. So, the virtual
path looks something like this:
\\domain_name\dfs_root\dfs_link
In happy day scenarios all seems to be fine, and
replication happens as expected. As soon as a file is
modified on one share, it is replicated almost in real-
time.
However, the results are not as pleasant when replication
falls behind. I stress the CPU on server bar_1 (site 2),
and modify a file on its share. Since the CPU on bar_1 is
really stressed, it'll be a while before the modified file
is replicated to the share on foo (site 1). Since the DC
is on the same site as foo (has the old copy), if I try to
access the file from the DC using the virtual path, I
expect to see the old copy of the file. In otherwords,
when I use Windows Explorer on the domain controller to
access
\\domain_name\dfs_root\dfs_link\filename
I should see the old copy of the file. But, that's not the
result that I get. I see the new file which I can only
assume that it fetches from across the newwork from bar_1
(on site 2) because the share on foo does not have the new
copy of the file yet. Well, this violates Dfs
documentation.
And, if I try to access the same file from bar_2 (site 2)
using the virtual path, I should get the new copy as bar_2
is located on the same site as bar_1. But, it gives me
back the old copy. This is a very weired behavior that I
was not expecting.
I should mention again that I use explorer to access the
files, and I don't know if it's not an appropriate tool to
access the files in these scenarios.
Your comments/suggestions is appreciated.
thanks,
Fari
I am using the Dfs+FRS services of W2K server, but the
results that I am getting is not quite consistent with the
Dfs documentation. I am wondering if anyone has
experienced this problem before:
My setup includes 2 sites:
site 1: DC and foo (both W2K servers)
site 2: bar_1 and bar_2 (W2K servers)
The Dfs root is hosted by the domain controller, and my
only Dfs Link is set up such that it points to physical
locations on servers foo (Primary), bar_1. So, the virtual
path looks something like this:
\\domain_name\dfs_root\dfs_link
In happy day scenarios all seems to be fine, and
replication happens as expected. As soon as a file is
modified on one share, it is replicated almost in real-
time.
However, the results are not as pleasant when replication
falls behind. I stress the CPU on server bar_1 (site 2),
and modify a file on its share. Since the CPU on bar_1 is
really stressed, it'll be a while before the modified file
is replicated to the share on foo (site 1). Since the DC
is on the same site as foo (has the old copy), if I try to
access the file from the DC using the virtual path, I
expect to see the old copy of the file. In otherwords,
when I use Windows Explorer on the domain controller to
access
\\domain_name\dfs_root\dfs_link\filename
I should see the old copy of the file. But, that's not the
result that I get. I see the new file which I can only
assume that it fetches from across the newwork from bar_1
(on site 2) because the share on foo does not have the new
copy of the file yet. Well, this violates Dfs
documentation.
And, if I try to access the same file from bar_2 (site 2)
using the virtual path, I should get the new copy as bar_2
is located on the same site as bar_1. But, it gives me
back the old copy. This is a very weired behavior that I
was not expecting.
I should mention again that I use explorer to access the
files, and I don't know if it's not an appropriate tool to
access the files in these scenarios.
Your comments/suggestions is appreciated.
thanks,
Fari