Firstly, we apologise for any inconvenience this has caused our customers.
We have been using the excellent Elasticsearch for our back-end data store for 5+ years and this is the first time we have seen an issue like this. The issue occurred in a routine update of the AddressBase Premium data. This appears to have been an issue with an Elasticsearch index that was snapshotted with “include_global_state” then restored with “include_global_state” which wiped out some key security settings on the destination cluster. In future indices should only ever be snapshotted and restored using the pre-prepared scripts and we will ensure this is enforced in our operations.
We are happy that the issue was identified and resolved within 5 minutes and will take steps to ensure this does not happen again.
Mark Varley, CEO, Addresscloud 12/12/2021