###BeginCaseStudy###
Case Study 4
A Datum
Overview
General Overview
A Datum Corporation has offices in Miami and Montreal.
The network contains a single Active Directory forest named adatum.com. The offices
connect to each other by using a WAN link that has a 5-ms latency.
A Datum standardizes its database platform by using SQL Server 2014 Standard edition.
Databases
Each office contains databases named Sales. Inventory, Customers, Products, Personnel, and Dev.
Servers and databases are managed by a team of database administrators. Currently, all of the
database administrators have the same level of permissions on all of the servers and all of the
databases.
The Customers database contains two tables named Customers and Classifications. The
following graphic shows the relevant portions of the tables:
The following table shows the current data in the Classifications table:
The Inventory database is used mainly for reports. The database is recreated every day. A full
backup of the database currently takes three hours to complete.
Stored Procedures
A stored procedure named USP_1 generates millions of rows of data for multiple reports.
USP_1 combines data from five different tables from the Sales and Customers databases in a
table named Table1.
After Table1 is created, the reporting process reads data from a table in the Products database
and searches for information in Table1 based on input from the Products table. After the
process is complete, Table1 is deleted.
A stored procedure named USP_2 is used to generate a product list. USP_2 takes several
minutes to run due to locks on the tables the procedure accesses.
A stored procedure named USP_3 is used to update prices. USP_3 is composed of several
UPDATE statements called in sequence from within a transaction. Currently, if one of the
UPDATE statements fails, the stored procedure continues to execute.
A stored procedure named USP_4 calls stored procedures in the Sales, Customers, and
Inventory databases. The nested stored procedures read tables from the Sales, Customers, and
Inventory databases. USP_4 uses an EXECUTE AS clause.
A stored procedure named USP_5 changes data in multiple databases. Security checks are
performed each time USP_5 accesses a database.
You suspect that the security checks are slowing down the performance of USP_5.
All stored procedures accessed by user applications call nested stored procedures. The nested
stored procedures are never called directly.
Design Requirements
Data Recovery
You must be able to recover data from the Inventory database if a storage failure occurs. You
have a Recovery Point Objective (RPO) of one hour.
You must be able to recover data from the Dev database if data is lost accidentally. You have
a Recovery Point Objective (RPO) of one day.
Classification Changes
You plan to change the way customers are classified. The new classifications will have four
levels based on the number of orders. Classifications may be removed or added in the future.
Management requests that historical data be maintained for the previous classifications.
Security
A group of junior database administrators must be able to view the server state of the SQL
Server instance that hosts the Sales database. The junior database administrators will not have
any other administrative rights.
A Datum wants to track which users run each stored procedure.
Storage
A Datum has limited storage. Whenever possible, all storage space should be minimized for
all databases and all backups.
Error Handling
There is currently no error handling code in any stored procedure. You plan to log errors in
called stored procedures and nested stored procedures. Nested stored procedures are never
called directly.
###EndCaseStudy###
You need to recommend a disaster recovery strategy for the Inventory database.
What should you include in the recommendation?
A.
Log shipping
B.
SQL Server Failover Clustering
C.
AlwaysOn availability groups
D.
Peer-to-peer replication
Explanation:
* Scenario:
/ You must be able to recover data from the Inventory database if a storage failure occurs.
You have a Recovery Point Objective (RPO) of one hour./ A. Datum Corporation has offices in Miami and Montreal.
* SQL Server Log shipping allows you to automatically send transaction log backups from a
primary database on a primary server instance to one or more secondary databases on
separate secondary server instances. The transaction log backups are applied to each of the
secondary databases individually.
C
A. Log shipping is true
SQL Server Log shipping allows you to automatically send transaction log backups from a primary database on a primary server instance to one or more secondary databases on separate secondary server instances. The transaction log backups are applied to each of the secondary databases individually. An optional third server instance, known as the monitor server, records the history and status of backup and restore operations and, optionally, raises alerts if these operations fail to occur as scheduled.
Log Shipping (btw Miami and Montreal) will not achieve an RPO of one hour. Log Shipping is not real-time; there’s always latency between the primary and Secondary servers, and there is no guarantee even if the two servers are situated in the same Data Centre, that Log Shipping would achieve a RPO of one hour.
Option C, AlwaysOn Availability Group seems a better option, since it only requires two server machines.
On the other hand, Automatic FCI would require minimum of three (3) server machines.
I think it is A because it says A Datum standardizes its database platform by using SQL Server 2014 Standard edition. you can’t use with the standard edition
SQL Server 2014 supports 2-Node AlwaysOn Failover Clustering.
So, I’ll settle for option B instead of A, as it is automatic, and can guarantee a RPO of less than one hour.
“A Datum standardizes its database platform by using SQL Server 2014 Standard edition.”
Always on with availability groups is only Enterprise.
A
I want to say failover clustering but…
“You must be able to recover data from the Inventory database if a storage failure occurs.”
Failover clustering operates by mediating control of shared storage. If you’ve lost the storage then this isn’t going to help much is it?
The question seems borked. Missing some options…
e) Implement database mirroring
f) Spend more money on Enterprise edition
If you did implement log shipping and you had lost a data file you would potentially be able to perform a tail log backup and get back up and running quickly(ish) without data loss although with considerable manual intervention.
Hmmm…
“The Inventory database is used mainly for reports. The database is recreated every day. A full backup of the database currently takes three hours to complete.”
and
“You must be able to recover data from the Inventory database if a storage failure occurs. You have a Recovery Point Objective (RPO) of one hour.”
If the database is recreated (blown away and re-established) every night and a full backup requires three hours then I don’t see how it is possible to allow for an RPO of one hour. A backup could not be guaranteed to complete within an hour of creating the database. I doubt that you could create the database instantly either fwiw. This question is all kinds of messed up and quite implausible.
Solution? Ignore this question. It’s not going to be in the exam in this form.
The Case Study says that you use SQL Server 2014 Standard Edition, so no Clustering or AlwaysOn available.
Log shipping would be good, but not as good as Peer-to-peer replication, since the Case Study requires “.. (RPO) of one hour”.
The Case Study [In General Overview] says “…has offices in Miami and Montreal”, and also [In Databases] says “Each office contains databases …”, so Peer-to-peer replication is the best option.