Octez_smart_rollup_node.Snapshots
type compression =
| No
Produce uncompressed archive. Takes more space.
*)| On_the_fly
Compress archive on the fly. The rollup node will use less disk space to produce the snapshot but will lock the rollup node (if running) for a longer time.
*)| After
Compress archive after snapshot creation. Uses more disk space temporarily than On_the_fly
but does not lock the rollup node for very long.
Compression strategy for snapshot archives.
module Header : sig ... end
val export :
?rollup_node_endpoint:Uri.t ->
Tezos_client_base.Client_context.full ->
no_checks:bool ->
compression:compression ->
data_dir:string ->
dest:string option ->
filename:string option ->
string Tezos_base.TzPervasives.tzresult Lwt.t
export ?rollup_node_endpoint cctxt ~no_checks ~compression ~data_dir ~dest
~filename
creates a tar gzipped archive with name filename
(or a generated name) in dest
(or the current directory) containing a snapshot of the data of the rollup node with data directory data_dir
. The path of the snapshot archive is returned. If no_checks
is true
, the integrity of the snapshot is not checked at the end. This function will first try to cancel any GC on the target node if rollup_node_endpoint
is specified to communicate with it.
val export_compact :
Tezos_client_base.Client_context.full ->
no_checks:bool ->
compression:compression ->
data_dir:string ->
dest:string option ->
filename:string option ->
string Tezos_base.TzPervasives.tzresult Lwt.t
export_compact cctxt ~no_checks ~compression ~data_dir ~dest ~filename
creates a tar gzipped archive with name filename
(or a generated name) in dest
(or the current directory) containing a snapshot of the data of the rollup node with data directory data_dir
. The difference with export
is that the snapshot contains a single commit for the context (which must be reconstructed on import) but is significantly smaller. If no_checks
is true
, we don't check if commitments are published on L1 (integrity is not checked because it requires rebuilding the context).
val import :
apply_unsafe_patches:bool ->
no_checks:bool ->
force:bool ->
Tezos_client_base.Client_context.full ->
data_dir:string ->
snapshot_file:string ->
unit Tezos_base.TzPervasives.tzresult Lwt.t
import ~apply_unsafe_patches ~no_checks ~force cctxt ~data_dir
~snapshot_file
imports the snapshot at path snapshot_file
into the data directory data_dir
. If no_checks
is true
, the integrity of the imported data is not checked at the end. Import will fail if data_dir
is already populated unless force
is set to true
. if apply_unsafe_patches
is true
and there is unsafe_pvm_patches
in the configuration they will be applied.
val info :
snapshot_file:string ->
(Header.t * [ `Compressed | `Uncompressed ]) Tezos_base.TzPervasives.tzresult
Lwt.t
info ~snapshot_file
returns information that can be used to inspect the snapshot file.
val with_modify_data_dir :
Tezos_client_base.Client_context.full ->
data_dir:string ->
apply_unsafe_patches:bool ->
?skip_condition:
(Octez_smart_rollup_node_store.Store.rw ->
Tezos_layer2_store.Context.rw ->
head:Octez_smart_rollup.Sc_rollup_block.t ->
bool Tezos_base.TzPervasives.tzresult Lwt.t) ->
(Node_context.rw ->
head:Octez_smart_rollup.Sc_rollup_block.t ->
unit Tezos_base.TzPervasives.tzresult Lwt.t) ->
unit Tezos_base.TzPervasives.tzresult Lwt.t
with_modify_data_dir cctxt ~data_dir ~apply_unsafe_patches ?skip_condition
f
applies f
in a read-write context that is created from data-dir
(and a potential existing configuration). The node context given to f
does not follow the L1 chain and f
is only supposed to modify the data of the rollup node. It is used internally by this module to reconstruct contexts from a snapshot but can also be use by the Repair
module to apply fixes off-line.