Proper use of cluster-setup jar ?sync? command

book

Article ID: 40879

calendar_today

Updated On:

Products

EMBEDDED ENTITLEMENTS MGR

Issue/Introduction

Introduction:

The sync command doesn't synchronize the data in CA Directories. Data synchronization happens between the CA Directories automatically.

 

The sync command can only be run from a secondary node in one of the two cases:

-          If the node is being configured for the first time as a secondary node.

-          If a new secondary node is added, then all the other secondary nodes need to be synced to get the new node details updated.

 

Instructions:

What the sync command does is:

-          Synchronize the EEM and CA Directory configuration files from the primary server

-          Empty the CA Directory DB and remove transaction file (which contains the last sync time)

o   so that the data is synced from CA Directory of Primary server, if this node is being configured for the first time

o   Synced from the newly added node, if sync is called after adding a new node

 

Sync command should not be executed in any other case.

Reason is: In a failover configuration, each CA Directory maintains a transaction file. This file contains the last sync time for each other CA Directory in the configuration, and any updates to the directory will be synced between the other directories based on this time stamp, to make sure that only the delta is synced.

               

If you simply call sync from any of the nodes, as explained above, the Local CA Directory DB will be emptied, but it will not get the data from the other CA Directories, as the other CA directories transaction files contain the recent sync time. Only changes after that timestamp will be synced.

               

Note: When a new node is added to the failover configuration, the new node should be synced first and then the other secondary nodes.

 

 

Steps to recover a corrupted node:

                If you run sync on a node for the second time out of the two cases explained above, below listed are the steps to recover the node:

               

(Note: Assume Node2.ca.com is the name of the corrupted node.)

1.       Stop the dsa on Node2.ca.com (su - dsa -c "dxserver stop all")

2.       Delete $DXHOME/data/itechpoz/itechpoz.dp from Node2.ca.com

3.       Remove and add Node2.ca.com, on PRIMARY NODE using the failover tool

4.       Start dsa on Node2.ca.com (su - dsa -c "dxserver start all")

5.       Data will be synced to Node2.ca.com

 

Environment

Release:
Component: ETEIAM