Thursday, February 16, 2012

can replication works for a Data Warehouse ?

Hi ,
I am trying to set replication for an exsiting Data Center (let's call it
DC) from the various databases in Server1
my problem :
1) i set up the publication with the snapshot to keep existing data in the
table(there's already data) if same table name is found , however, the
synchronization failed becoz of duplicate index/key
ques : shldn't it just ignore those duplicates ?
and even if there's new data for that table it couldn't replicate over due
to the duplicates
2) i have also tried to use this option "to delete those matching with row
filter" but i realised that those that are the same record it would be
deleted and re-replicated over but those in the destination that not in the
source table have been deleted
3) i finally tried using this option "to delete data and re-create the table
and it works but the issue here is i have many databases to be replicated
over to the data warhouse , i shldn't be forced to purposely removed a table
with data to somewhere else and then after replicate is successful then copy
that data over
could any one kindly advise
tks & rdgs
Yes replication can be used in this scenario. Snapshot replication performs
a period refresh of all of the data. It sounds like transactional
replication might be a better option for you (incremental changes).
HTH
Jerry
"maxzsim" <maxzsim@.discussions.microsoft.com> wrote in message
news:91170637-D104-4CCD-AEC9-F2458A0C282A@.microsoft.com...
> Hi ,
> I am trying to set replication for an exsiting Data Center (let's call it
> DC) from the various databases in Server1
> my problem :
> 1) i set up the publication with the snapshot to keep existing data in the
> table(there's already data) if same table name is found , however, the
> synchronization failed becoz of duplicate index/key
> ques : shldn't it just ignore those duplicates ?
> and even if there's new data for that table it couldn't replicate over due
> to the duplicates
> 2) i have also tried to use this option "to delete those matching with row
> filter" but i realised that those that are the same record it would be
> deleted and re-replicated over but those in the destination that not in
> the
> source table have been deleted
> 3) i finally tried using this option "to delete data and re-create the
> table
> and it works but the issue here is i have many databases to be replicated
> over to the data warhouse , i shldn't be forced to purposely removed a
> table
> with data to somewhere else and then after replicate is successful then
> copy
> that data over
> could any one kindly advise
> tks & rdgs
|||Hi,
If Replication does work for a Data Warehouse , how shld i set it up as i
have tried the settings below but it somehow did not work for me
appreciate any advise
tks & rdgs
"Jerry Spivey" wrote:

> Yes replication can be used in this scenario. Snapshot replication performs
> a period refresh of all of the data. It sounds like transactional
> replication might be a better option for you (incremental changes).
> HTH
> Jerry
> "maxzsim" <maxzsim@.discussions.microsoft.com> wrote in message
> news:91170637-D104-4CCD-AEC9-F2458A0C282A@.microsoft.com...
>
>
|||Replication can be a difficult. If I were you I would step back from the
whole data warehouse thing and implement both snapshot and transactional
replication on a test box and become familiar with the varioius offerings,
options, agents, icons and settings (BOL is pretty good at explaining
replication types). Once you have a better understand of how replication
works and what are the basic and advanced feature sets of repliction, I
think you'll be able to determine which replication type (if any - maybe DTS
would work better or BULK INSERT) would be best for your scenario.
HTH
Jerry
"maxzsim" <maxzsim@.discussions.microsoft.com> wrote in message
news:A7FEFCA8-8244-45D7-B664-6A5656121451@.microsoft.com...[vbcol=seagreen]
> Hi,
> If Replication does work for a Data Warehouse , how shld i set it up as i
> have tried the settings below but it somehow did not work for me
> appreciate any advise
> tks & rdgs
> "Jerry Spivey" wrote:

No comments:

Post a Comment