How to: Troubleshoot/Monitor the Rollup Server (9.6 onward - SQL Replication)

Environment:

LDMS v9.6+

Review Date:

11/11/2014

How to:

Since LDMS 9.6 version, to create and maintain the Rollup Server database, Microsoft SQL Replication technology is used (see article Best Known Method To Configure Rollup for LDMS 9.6).

The Replication type chosen for Rollup Server is a "Transactional Replication". You can see the Log Reader Agent (logread.exe) running on the Publisher (A Core Server).

Transactional replication is managed by using two replication agents: the Log Reader Agent and the Distribution Agent. The Log Reader Agent is responsible for moving changes from the transaction log on the publisher to the distribution database. The Distribution Agent is responsible for moving batches of changes from the distribution database to each subscriber.

There are main tools to understand the current activity about the Rollup SQL Replication process; "Replication Monitor" and "Tracer Token" for the Publisher (The Core Servers) and "Job History" for the subscriber (The Rollup Server)

1- Replication Monitor (Publisher - Core Servers DB) - Step by Step:

You commonly use the Replication Monitor to monitor the replication architectures. Within Replication Monitor, you can obtain statistics on the operational state of all publications, subscriptions, and agents. You can also view error and status history information to troubleshoot any issues that might be occurring. As changes are moved by the Log Reader and Distribution Agents, the engine maintains statistics on the rate of data movement and how long it took to move data. By using these  statistics, Replication Monitor can display continuous information that tells you how many transactions still need to be sent to subscribers, as well as approximately how long it will take to catch back up.

1. Go on SSMS where one Core Server's DB is
2. Connect to the instance and right click Replication > Choose "Launch Replication Monitor" (NOTE: From this menu, you have also a lot of options to check with)
3. Have a look then to the printscreen below showing you the error which can prevent the Publisher's DB to be replicated

2- Tracer Token (Publisher - Core Servers DB) - Step by Step:

Although the statistics in Replication Monitor provide good status information to determine how far behind subscribers are, it does not provide any finer details. Replication Monitor displays a single statistic for the latency, but administrators cannot tell whether the bottleneck is in the Log Reader Agent or whether the Distribution Agent is backing up. So, tracer tokens were also introduced to provide this detail. A tracer token is a specialized transaction issued for the replication engine. It is sent to the transaction log like any other transaction. The replication agents move the tracer token through the architecture just as they do for any other transaction. With a tracer token, you can get exact timings of how long it took to move the token to the distribution database and how long it took to be sent to each subscriber

1. Go on SSMS where one Core Server's DB is (The publishers)
2. Connect to the instance and right click Replication > Choose "Launch Replication Monitor"
3. Go on the connection (In the printscreen "[CORE96RTMS]: LDMS")
4. Go on tab "Tracer Tokens" and press on the button "Insert Tracer" > It will give all details wanted.

3- Job History (Subscriber - Rollup DB) - Step by Step:

Job history will give you details about what's going on regarding the distributor and the subscriber (All on the Rollup Core Server).

1. Go on SSMS where one Rollup's DB is (The subscriber)
2. Connect to the instance and go to Replication > Right-click Local Subscription > Choose "View Job History" (NOTE: From this menu, you have also a lot of options to check with)
3. The Log File Viewer will give you all the details about the error or if it works just fine (See printscreen below)

4- Classic issues/fixes:

1. The passwords defined in the Rollup Facility on the Rollup Server (\ManagementSuite\LANDesk.Database.Replication.exe) are wrong (Go on the Rollup Server > Start \ManagementSuite\LANDesk.Database.Replication.exe > Check everything is right and click OK)
2. The SQL server agent service is not started on the SQL instances running the Core Servers DBs and Subscriber DB
3. The Log Reader agent or Snapshot Agent are not started on one or several publishers (Go in SSMS > Replication > Right the Connector > Choose View Snapshot Agent Status or View Log Reader Agent Status)
4. Something went wrong when creating the Rollup's DB with the Rollup Facility (Go on the Rollup Server > Start \ManagementSuite\LANDesk.Database.Replication.exe > Check everything is right and click OK)

If you still have some issues or you can't figure out the issue, please contact support.

See also: How To: Find the SQL scripts used to create a Rollup Server with the tables, fields, stored procedures... (9.6 onward - SQL Replication)