Hi,
As I understand it, snapshot repos get upgraded as part of the upgrade to 5.x. Does this mean that 2.x snapshots in that repo can be restored into the now 5.x cluster? Also, what will happen to fields with conflicting types? Will they be dropped?
I'm asking in the context of indices that we might drop during upgrade due to mapping conflicts and which we would then want to restore. Unfortunately, due to dynamic mappings and ill-defined log formats we can't be 100% sure that some rogue mapping type conflict might appear prior to the upgrade.
What I'm concerned about is not whether it's possible to restore a 2.x snapshot. But rather, I'm conerned that if we remove an index to complete the upgrade what will happen when restoring an index that contains fields with clashing types. My hope is that they'll be dropped allowing the rest of the index to restore..?
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.