Learn from 0 copy ~ My copy learning History (a) simple snapshot copy processing

---restore content starts---

Different servers before synchronizing data can have a variety of methods, the simplest can choose to use a linked server, build a job to dispatch. It's easy to operate, but if you have more tables and more synchronized environments, then it's harder. So I also learn about this tool called Replication ~

My version

Publisher Microsoft SQL Server R2 (RTM)-My publisher and Distributor are in the same environment, which is easier

Subscriber Microsoft SQL Server R2 (RTM)-This is the virtual machine, because the computer is too low-end ... So only one _ (: З"∠) _

--------------------------------------------------I'm starting to configure the split line on the publishing side---------------------------------------------------------- ------------

First, we'll start with a release.

1, because the admin permissions are too large, so I created a call REPL account, specifically for replication, it only requires basic permissions on OK

2. Then open SSMs and log in to build the published instance. (There is a place to note that you need to log in with the computer name + instance name to log in, can not use the IP address)

Otherwise, this error will explode when the replication is established (don't ask me why I know _ (: З"∠) _)

3, um ~ Then just click on the open instance, replication, publishing, right-click New Publication

Here I chose to set the publisher and distribution to the same machine, because to measure the steps, do not think so much, if you use other machines for distribution, select the following

4, set the snapshot storage path, here to fill the network path, and to share this folder out, if the settings are wrong, you will always be prompted not to access the folder

5, select the database to be published, this step should not say more

6, and then select the publication type, just like the title, I chose Snapshot publishing

7. Then select the object to publish, including the table ah, the stored procedure Ah, and so on, I ordered the table.

8, then filter the data, you can write the statement to synchronize only the specified data. I'd like to skip over here.

9, then again loved the next step.

10, then the security settings, this place still need to note that as a published process account, must have the publication database Dbowner or higher permissions, because the distribution server and published on the same machine, so also need to have distribution data dbowner or higher permissions.

11, then a handsome (dou) gas (BI) name, point completion ~ Good ~ Release

-----------------------------------------------------------------------------then set up the subscriber------------------------------------- --------------------------------------------------------

1, click to open a subscription, right-click New Subscription, the first choice to publish objects.

2. Then select the subscription mode, push and pull 2 modes

Push: It is the initiative to push data from the Distributor (this configuration is relatively simple, the job is also on the distribution server)

Pull: This is the request data from the Subscriber to the Distributor. That's what I'm using.

3. Select a subscription database

4, security settings, this process account, note the need to have a subscriber Dbowner or higher permissions to line

5, and then loved to continue to the next ~ on the finish ~ boil and boil

--------------------------------------------------------------------------------------------------------------- --------------------------------------------

Then we open the publisher's Replication Monitor to see ~

Good ~ normal ~

If there is anything abnormal ~ look at what agent Ah ~ such as there is no normal work

Also, for snapshot replication, the publisher's job can set the frequency of execution and how often it is set to run. Yes, roughly.

Learn from 0 copy ~ My copy learning History (a) simple snapshot copy processing