Today we've run into an issue with DMS, where after running successfully for weeks, we got this: https://repost.aws/knowledge-center/dms-replication-instance-storage-full
Our instance was out of space, and RDS started to accumulate logs. We were lucky to catch it in time. This is a VERY sharp edge.
What is truly confusing though, is that the explanation on how to deal with this is to update the task and set:
{
"Logging": {
"DeleteTaskLogs":true
}
}
Then, unset it again.
I am in denial about this being truly the only way. Can someone confirm?
What happens to the log while it is set to true? Can we always keep it true? Will the logs still make it to CloudWatch?
This isn't very clear.
Resources:
Then why another call is needed to "unset" this if it is not "set" anywhere?
there is no need to call unset