Updating the record associated with the Recovery Kit for Route 53™ resource startup may take time
Amazon provides the following information regarding the propagating speed of changes made to the DNS record.
Q: How quickly will changes I make to my DNS settings on Amazon Route 53™ propagate globally?
The Recovery Kit for Route 53™ resource checks the status of updates to the DNS record using the Route53 API. It considers that updates are completed when receiving INSYNC status, and retries the status checking when it receives a PENDING status. As a result, the Route53 resource may result in a startup failure when it takes a long time to propagate updates to the DNS even though the record is updated successfully for the Route53 resource startup process.
If the startup of the Route53 resource fails, check the Route53 management console to make sure that the A record is updated correctly. If it is updated, updates to the relevant DNS service have been completed. Updates to SPS for Windows are required to propagate the updates to the DNS service. Restart the Route53 resource from SPS for Windows GUI.
If you encounter the startup failure of the Route53 resource all the time due to the above mentioned reason, increase the number of the value of “ROUTE53_CHANGEID_TRY_COUNT” in %LKROOT%\etc\default\LifeKeeper to 6 or 7 (the default value is 5). Restarting SPS for Windows or the OS is not required for this change.
Correctly set TTL value of the DNS record
Access from a client after a switchover or a failover uses the DNS information cache that each client holds until the time set as TTL is passed. If the longer TTL value is set, access attempts the address before switching increase and unexpected problems may occur. If the shorter TTL value is set, DNS resolution often occurs and network load increases. Set the TTL value as short as possible according to your environment.
Set the “ROUTE53_TTL” for the TTL value in %LKROOT%\etc\default\LifeKeeper. The unit should be in seconds.