G
Graham Turner
Am looking for a few ideas on the correct procedure for recovery from a
"failed" FRS infrastructure.
what has transpired is a copy of a dataset in excess of 300 MB to a folder
in a FRS replica set that was started to have been copied thorughout a WAN
infrastructure,
given the potential impact on WAN operation this was terminated by stopping
the NTFRS service on the host servers and the removal of the remote shares
as replicas.
it is apparent there are a number of servers that still hold a very large
dataset in an "FRS-STAGING" folders.
QU: is the manual removal of this frs-staging folder a supported operation -
am not sure on this given the potential impact on the SYSVOL replication ??
for obvious reasons it is eminently preferable not too have this dataset
replicate to servers that are connected by low bandwidth WAN links !
what has also happened is a number of "orphaned" directory objects as viewed
in the advanced view of AD users and computers -
system / FRS / DFS volumes / dfs / templates / servername
all servers in the previously configured DFS replica set are still listed in
the directory and still with child objects which presumably are the
connection objects of the replica set - these would appear to be "corrupt"
given that "DEL" appears in the properties of these connection objects
QU - is it a supported operation to manually delete these objects from the
directory - presumably there is no other way out of this given that the
directory does not seem to have sorted itself out following the removal of
the replicas using the DFS GUI.
apologies for the essay and will be going to shoot the "hapless admin" !!!
GT
"failed" FRS infrastructure.
what has transpired is a copy of a dataset in excess of 300 MB to a folder
in a FRS replica set that was started to have been copied thorughout a WAN
infrastructure,
given the potential impact on WAN operation this was terminated by stopping
the NTFRS service on the host servers and the removal of the remote shares
as replicas.
it is apparent there are a number of servers that still hold a very large
dataset in an "FRS-STAGING" folders.
QU: is the manual removal of this frs-staging folder a supported operation -
am not sure on this given the potential impact on the SYSVOL replication ??
for obvious reasons it is eminently preferable not too have this dataset
replicate to servers that are connected by low bandwidth WAN links !
what has also happened is a number of "orphaned" directory objects as viewed
in the advanced view of AD users and computers -
system / FRS / DFS volumes / dfs / templates / servername
all servers in the previously configured DFS replica set are still listed in
the directory and still with child objects which presumably are the
connection objects of the replica set - these would appear to be "corrupt"
given that "DEL" appears in the properties of these connection objects
QU - is it a supported operation to manually delete these objects from the
directory - presumably there is no other way out of this given that the
directory does not seem to have sorted itself out following the removal of
the replicas using the DFS GUI.
apologies for the essay and will be going to shoot the "hapless admin" !!!
GT