You can use the disaster recovery feature to migrate FTL operations to a different site, even though no disaster has occurred.
This task illustrates how you can
temporarily use the disaster recovery feature to facilitate moving FTL operations to a new geographical location. (This process does not result in actual protection from a disaster.)
Note: This task requires that you are
not already using the disaster recovery feature for its usual purpose.
Complete these task steps during scheduled maintenance time so that enterprise-critical operations are not affected.
-
Define the persistence services you intend to deploy at SiteNew.
In the FTL server GUI persistence clusters grid, define a second service set for the cluster, and define new persistence services in it.
Deploy the updated realm definition.
-
Copy the FTL server configuration file from SitePrev to SiteNew.
Make a reference copy of the SitePrev configuration file.
-
Modify your
working copy of the
SitePrev configuration file.
Configure the
drto parameter for each SitePrev core server to contain a list of the SiteNew core servers.
In this example, bold type indicates modifications to the existing SitePrev configuration file:
globals:
core.servers:
ftl1: host1:8585
ftl2: host2:8585
ftl3: host3:8585
servers:
# Core servers with realm services
ftl1:
drto: >
http://host-new1:8585
|http://host-new2:8585
|http://host-new3:8585
- realm:
data: /myhome/ftlserver/data
ftl2:
drto: >
http://host-new1:8585
|http://host-new2:8585
|http://host-new3:8585
- realm:
data: /myhome/ftlserver/data
ftl3:
drto: >
http://host-new1:8585
|http://host-new2:8585
|http://host-new3:8585
- realm:
data: /myhome/ftlserver/data
# Auxiliary servers with persistence services
ftl4:
- persistence:
name: psvc4 #Psvc numbering is parallel to aux server names.
ftl5:
- persistence:
name: psvc5
ftl6:
- persistence:
name: psvc6
-
Modify the
SiteNew configuration file.
-
In the
core.servers parameter and in the servers section, replace the SitePrev core servers with the SiteNew core servers.
-
Configure the
drfor parameter for each SiteNew server so it contains a list of the SitePrev core servers.
-
Configure a realm service for each SiteNew core server.
-
Configure a persistence service for each SiteNew auxiliary server.
For each auxiliary server, specify a unique service
name that you defined in step 1.
-
If authentication is required, configure the
user and
password parameters, with credentials that the SiteNew primary servers can use to authenticate themselves to affiliated servers.
Ensure that the user name is in the authorization group
ftl-internal.
-
Duplicate the other relevant persistence parameters from SitePrev servers to SiteNew servers.
In this example, bold type indicates modifications to the existing SiteNew configuration file (which was a copy of the original SitePrev configuration file):
globals:
core.servers:
ftl-new1: host-new1:8585
ftl-new2: host-new2:8585
ftl-new3: host-new3:8585
servers:
# SiteNew core servers with realm services
ftl-new1:
drfor: >
http://host1:8585
|http://host2:8585
|http://host3:8585
- realm:
data: /myhome/ftlserver/data
ftl-new2:
drfor: >
http://host1:8585
|http://host2:8585
|http://host3:8585
- realm:
data: /myhome/ftlserver/data
ftl-new3:
drfor: >
http://host1:8585
|http://host2:8585
|http://host3:8585
- realm:
data: /myhome/ftlserver/data
#SiteNew auxiliary servers with persistence services
ftl-new4:
- persistence:
name: psvc-new4 #Numbering follows aux server names.
ftl-new5:
- persistence:
name: psvc-new5
ftl-new6:
- persistence:
name: psvc-new6
-
Stop and restart the
core servers (with realm services) at
SitePrev, one at a time.
After restarting each server, and before restarting the next server, wait until the status of the restarted server becomes Running.
Each FTL server will read the updated configuration file when it restarts.
-
Stop and restart the
auxiliary servers (with persistence services) at
SitePrev, one at a time.
After restarting each server, and before restarting the next server, wait until both the server
and the persistence cluster to reach equilibrium:
- Wait until the status of the restarted server becomes Running.
- Wait until the status of the persistence cluster becomes Running.
Each FTL server will read the updated configuration file when it restarts.
-
Start disaster recovery FTL servers at SiteNew.
-
First start the SiteNew core servers (with realm services).
-
Then start the SiteNew auxiliary servers (with persistence services).
At this point you have arranged SiteNew as a temporary disaster recovery backup for SitePrev.
-
Stop all publishing activity at SitePrev, and wait for the persistence services to finish replicating the data.
Gracefully stop all client applications at SitePrev.
In the persistence clusters status table and its services list sub-tables, compare the history values of all the services at both sites. When they all match, replication of persistence data is complete.
At this point the disaster recovery mechanism has copied data from SitePrev to SiteNew.
-
Stop all FTL servers at SitePrev.
At this point you have disabled SitePrev, leaving all FTL operations at SiteNew.
-
Make another reference copy of the SiteNew configuration file.
-
Modify your working copy of the SiteNew configuration file to eliminate the temporary use of the disaster recovery feature.
Delete the
drfor parameter from each SiteNew server, so they become
primary FTL servers. For example:
globals:
core.servers:
ftl-new1: host-new1:8585
ftl-new2: host-new2:8585
ftl-new3: host-new3:8585
servers:
# SiteNew core servers with realm services
ftl-new1:
- realm:
data: /myhome/ftlserver/data
ftl-new2:
- realm:
data: /myhome/ftlserver/data
ftl-new3:
- realm:
data: /myhome/ftlserver/data
#SiteNew auxiliary servers with persistence services
ftl-new4:
- persistence:
name: psvc-new4 #Numbering follows aux server names.
ftl-new5:
- persistence:
name psvc-new5
ftl-new6:
- persistence:
name psvc-new6
-
Stop and restart the core servers at SiteNew,
one at a time.
Each FTL server will read the updated configuration file when it restarts.
After restarting each server, and before restarting the next server, wait until the status of the restarted server becomes Running.
-
Toggle the persistence primary set to SiteNew.
-
Using the browser GUI of a primary FTL server at SiteNew, navigate to the clusters grid, and ensure that the Primary Set column is visible.
-
Change the primary set of the persistence cluster so that the persistence services at the SiteNew will become the primary set.
-
Delete the persistence services at SitePrev from the persistence cluster.
-
Delete the (now empty) service set that contained the defunct persistence services at SitePrev.
-
Deploy the new realm definition.
-
Wait until the status of every FTL server becomes Running.
-
Wait until the status of the persistence cluster becomes Running.
-
Direct application clients to FTL servers at SiteNew.
Start application clients at SiteNew. Explicitly supply the host and port locations of FTL servers at SiteNew.
SiteNew is now the active site of FTL operations.