Skip to content

CQ: Retry opening write file when writing messages #14178

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Jul 2, 2025
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 3 additions & 1 deletion deps/rabbit/src/rabbit_classic_queue_index_v2.erl
Original file line number Diff line number Diff line change
Expand Up @@ -591,7 +591,9 @@ publish(MsgId, SeqId, Location, Props, IsPersistent, ShouldConfirm, TargetRamCou
new_segment_file(Segment, SegmentEntryCount, State = #qi{ segments = Segments }) ->
#qi{ fds = OpenFds } = reduce_fd_usage(Segment, State),
false = maps:is_key(Segment, OpenFds), %% assert
{ok, Fd} = file:open(segment_file(Segment, State), [read, write, raw, binary]),
{ok, Fd} = rabbit_file:open_eventually(
segment_file(Segment, State),
[read, write, raw, binary]),
%% We then write the segment file header. It contains
%% some useful info and some reserved bytes for future use.
%% We currently do not make use of this information. It is
Expand Down
23 changes: 3 additions & 20 deletions deps/rabbit/src/rabbit_classic_queue_store_v2.erl
Original file line number Diff line number Diff line change
Expand Up @@ -194,25 +194,6 @@ maybe_flush_buffer(State = #qs{ write_buffer_size = WriteBufferSize }) ->
false -> State
end.

open_eventually(File, Modes) ->
open_eventually(File, Modes, 3).

open_eventually(_, _, 0) ->
{error, eacces};
open_eventually(File, Modes, N) ->
case file:open(File, Modes) of
OK = {ok, _} ->
OK;
%% When the current write file was recently deleted it
%% is possible on Windows to get an {error,eacces}.
%% Sometimes Windows sets the files to "DELETE PENDING"
%% state and delays deletion a bit. So we wait 10ms and
%% try again up to 3 times.
{error, eacces} ->
timer:sleep(10),
open_eventually(File, Modes, N - 1)
end.

flush_buffer(State = #qs{ write_buffer_size = 0 }, _) ->
State;
flush_buffer(State0 = #qs{ write_buffer = WriteBuffer }, FsyncFun) ->
Expand All @@ -223,7 +204,9 @@ flush_buffer(State0 = #qs{ write_buffer = WriteBuffer }, FsyncFun) ->
Writes = flush_buffer_build(WriteList, CheckCRC32, SegmentEntryCount),
%% Then we do the writes for each segment.
State = lists:foldl(fun({Segment, LocBytes}, FoldState) ->
{ok, Fd} = open_eventually(segment_file(Segment, FoldState), [read, write, raw, binary]),
{ok, Fd} = rabbit_file:open_eventually(
segment_file(Segment, FoldState),
[read, write, raw, binary]),
case file:position(Fd, eof) of
{ok, 0} ->
%% We write the file header if it does not exist.
Expand Down
32 changes: 32 additions & 0 deletions deps/rabbit/src/rabbit_file.erl
Original file line number Diff line number Diff line change
Expand Up @@ -17,6 +17,7 @@
-export([read_file_info/1]).
-export([filename_as_a_directory/1]).
-export([filename_to_binary/1, binary_to_filename/1]).
-export([open_eventually/2]).

-define(TMP_EXT, ".tmp").

Expand Down Expand Up @@ -338,3 +339,34 @@ binary_to_filename(Bin) when is_binary(Bin) ->
Other ->
erlang:error(Other)
end.

%% On Windows the file may be in "DELETE PENDING" state following
%% its deletion (when the last message was acked). A subsequent
%% open may fail with an {error,eacces}. In that case we wait 10ms
%% and retry up to 3 times.

-spec open_eventually(File, Modes) -> {ok, IoDevice} | {error, Reason} when
File :: Filename | iodata(),
Filename :: file:name_all(),
Modes :: [file:mode() | ram | directory],
IoDevice :: file:io_device(),
Reason :: file:posix() | badarg | system_limit.

open_eventually(File, Modes) ->
open_eventually(File, Modes, 3).

open_eventually(_, _, 0) ->
{error, eacces};
open_eventually(File, Modes, N) ->
case file:open(File, Modes) of
OK = {ok, _} ->
OK;
%% When the current write file was recently deleted it
%% is possible on Windows to get an {error,eacces}.
%% Sometimes Windows sets the files to "DELETE PENDING"
%% state and delays deletion a bit. So we wait 10ms and
%% try again up to 3 times.
{error, eacces} ->
timer:sleep(10),
open_eventually(File, Modes, N - 1)
end.
Loading