umgeher/leveled
0
Fork 0
Code Issues Pull requests Releases Activity Actions

Handle a Filename not being present in pending_deletes (#357)

Browse source 
This may be due to a situation whereby a second confirm_delete has been sent before the sst_deleteconfirmed has been received from the first request.

Now this will now cause inert action rather than crashing.
This commit is contained in:
Martin Sumner 2021-10-18 11:06:52 +01:00 • committed by GitHub
parent 70ebb62a61
commit beb8ba14a5
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
1 changed files with 9 additions and 6 deletions
Download patch file Download diff file Expand all files Collapse all files

15
src/leveled_pmanifest.erl
View file

@ -545,21 +545,24 @@ release_snapshot(Manifest, Pid) ->
%% A SST file which is in the delete_pending state can check to see if it is
%% ready to delete against the manifest.
ready_to_delete(Manifest, Filename) ->
{ChangeSQN, _ME} = dict:fetch(Filename, Manifest#manifest.pending_deletes),
case Manifest#manifest.min_snapshot_sqn of
0 ->
PendingDelete = dict:find(Filename, Manifest#manifest.pending_deletes),
case {PendingDelete, Manifest#manifest.min_snapshot_sqn} of
{{ok, _}, 0} ->
% no shapshots
PDs = dict:erase(Filename, Manifest#manifest.pending_deletes),
{true, Manifest#manifest{pending_deletes = PDs}};
N when N >= ChangeSQN ->
{{ok, {ChangeSQN, _ME}}, N} when N >= ChangeSQN ->
% Every snapshot is looking at a version of history after this
% was removed
PDs = dict:erase(Filename, Manifest#manifest.pending_deletes),
{true, Manifest#manifest{pending_deletes = PDs}};
_N ->
_ ->
% If failed to delete then we should release a phantom pid
% in case this is necessary to timeout any snapshots
% This wll also trigger a log
% This wll also trigger a log
% This may also be because of i355 - a race condition when a second
% confirm_delete may be sent prior to the first being processed
{false, release_snapshot(Manifest, ?PHANTOM_PID)}
end.