https://silo.pgsty.com/tags/data-safety/ Recent content in Data Safety on SILO Hugo en Mon, 09 Nov 2026 12:30:06 +0800 https://silo.pgsty.com/blog/design/restored-copy-expiry/ Wed, 04 Nov 2026 00:00:00 +0000 https://silo.pgsty.com/blog/design/restored-copy-expiry/ Why the expiry of a restored copy of a current version hid the object or stripped an older null version, why addressing the exact version is the core repair, and which storage and native-restore safeguards have to ship with it. https://silo.pgsty.com/blog/design/multi-pool-lifecycle-expiry/ Wed, 04 Nov 2026 00:00:00 +0000 https://silo.pgsty.com/blog/design/multi-pool-lifecycle-expiry/ Why lifecycle rules evaluated on the versions of one pool hide, delete, or resurrect versions stored in another pool, and why a recheck on the merged version list under the object lock is the minimum complete repair. https://silo.pgsty.com/blog/design/heal-dry-run-safety/ Wed, 04 Nov 2026 00:00:00 +0000 https://silo.pgsty.com/blog/design/heal-dry-run-safety/ Why an admin heal run as a dry run can still delete a dangling object version when its metadata has no read quorum, why a five-line early return is the right fix for this release, and why the same fix must also carry the dry-run flag to remote nodes during bucket heal. https://silo.pgsty.com/blog/design/ilm-delete-all-exact-key/ Wed, 04 Nov 2026 00:00:00 +0000 https://silo.pgsty.com/blog/design/ilm-delete-all-exact-key/ Why a lifecycle or batch “delete all versions” of one key also deletes every object stored under that key’s name, and why purging the key’s own versions by ID, under the object lock and after a recheck, is the minimum complete repair.