HostWeb Forums » Microsoft Databases » microsoft.public.sqlserver.replication » Replication Scenario
Topic: Replication Scenario
One of our customers is planning to use merge replication as a way to
replicate information from a centralized sale and distribution system into
remote salesmen (Tablet PC with SQL Server Express as a backend) and
distribution devices (Windows Mobile using SQL Compact Edition as backend).
So far, the performance tests results on merge replication are excellent.
However we have found some limitations in the model that are not aligned with
their business needs. We wish to know if there is some way to manage the
following issues:
1. The customer wants that a remote device contains only a single database
subscribed to different sets of publications using a pull model. Even though
tables contained on the publications target different data (eg. Monthly,
weekly or daily catalogs to replicate), a set of control tables to configure
extended join filtering are shared across publications. We only could
synchronize the first created subscription successfully. The rests of
publications fail to synchronize due to the share of these control tables.
2. Regarding the join filtering clauses suppose the existence of the
“device†table that contains a host_name column and other handy columns to
implement other custom filtering (eg. Status of invoices to deploy). If
device table is filtered using the host_name function, and this table is
joined to filter customer per device; customer per device is joined with
customer and finally customer with invoices. How is possible to use
device.status to further filter invoices information? At least the IDE only
let to refer immediate joined tables.
3. The idea of the previous example is to replicate the invoice table. Does
there is a way to prevent replicating device and device per customer control
tables? The IDE doesn’t seem to permit this.
--
Microsoft SQL Server MVP
MCSE, MCDBA, TS SQL 2005, MCSD, CCNA
Replies below ↓