Offline physical-fragment reclamation. All users of the table must be stopped for deletion; this is not an online garbage collector or a distributed lock.
Offline physical-fragment reclamation. All users of the table must be stopped for deletion; this is not an online garbage collector or a distributed lock.
(collect-fragments! spec)(collect-fragments! spec {:keys [dry-run? quiescent?] :or {dry-run? true}})Scan a dedicated Konserve table and report unreferenced fragment bytes. Defaults to {:dry-run? true}. To delete, pass {:dry-run? false :quiescent? true}, after stopping ALL readers and writers of ALL databases in the table. Keep them stopped until this call finishes and restart readers afterward. :quiescent? is the caller's assertion, not a lock.
Uses strong paginated scans, validates all manifests and referenced fragment presence before deleting, and rescans on every call. A failed deletion can be retried. Reports raw item bytes, not billed storage or an AWS cost estimate. Does not remove logical nodes; run Datahike's logical GC separately.
Scan a dedicated Konserve table and report unreferenced fragment bytes.
Defaults to {:dry-run? true}. To delete, pass
{:dry-run? false :quiescent? true}, after stopping ALL readers and writers of
ALL databases in the table. Keep them stopped until this call finishes and
restart readers afterward. :quiescent? is the caller's assertion, not a lock.
Uses strong paginated scans, validates all manifests and referenced fragment
presence before deleting, and rescans on every call. A failed deletion can be
retried. Reports raw item bytes, not billed storage or an AWS cost estimate.
Does not remove logical nodes; run Datahike's logical GC separately.cljdoc builds & hosts documentation for Clojure/Script libraries
| Ctrl+k | Jump to recent docs |
| ← | Move to previous article |
| → | Move to next article |
| Ctrl+/ | Jump to the search field |