jiangphcn closed pull request #1186: [WIP] Clustered purge mixed cluster upgrade
URL: https://github.com/apache/couchdb/pull/1186
This is a PR merged from a forked repository.
As GitHub hides the original diff on merge, it is displayed below for
the sake of provenance:
As this is a foreign pull request (from a fork), the diff is supplied
below (as it won't show otherwise due to GitHub magic):
diff --git a/src/couch/src/couch_bt_engine.erl
b/src/couch/src/couch_bt_engine.erl
index ee0d6d864b..1783c6e2fb 100644
--- a/src/couch/src/couch_bt_engine.erl
+++ b/src/couch/src/couch_bt_engine.erl
@@ -84,6 +84,8 @@
seq_tree_join/2,
seq_tree_reduce/2,
+ purge_tree_reduce/2,
+
local_tree_split/1,
local_tree_join/2
]).
@@ -604,6 +606,13 @@ seq_tree_reduce(rereduce, Reds) ->
lists:sum(Reds).
+purge_tree_reduce(reduce, IdRevs) ->
+ % count the number of purge requests
+ length(IdRevs);
+purge_tree_reduce(rereduce, Reds) ->
+ lists:sum(Reds).
+
+
local_tree_split(#doc{revs = {0, [Rev]}} = Doc) when is_binary(Rev) ->
#doc{
id = Id,
@@ -681,7 +690,16 @@ init_state(FilePath, Fd, Header0, Options) ->
Compression = couch_compress:get_compression_method(),
- Header1 = couch_bt_engine_header:upgrade(Header0),
+ DiskV = couch_bt_engine_header:disk_version(Header0),
+ Header1 = case couch_bt_engine_header:latest(DiskV) of
+ undefined ->
+ % disk_version is higher that the latest version
+ % db must be the format of Clustered Purge,
+ % perform downgrade
+ downgrade_purge_info(Fd, Header0);
+ _ ->
+ couch_bt_engine_header:upgrade(Header0)
+ end,
Header = set_default_security_object(Fd, Header1, Compression, Options),
IdTreeState = couch_bt_engine_header:id_tree_state(Header),
@@ -727,7 +745,7 @@ init_state(FilePath, Fd, Header0, Options) ->
% to be written to disk.
case Header /= Header0 of
true ->
- {ok, NewSt} = commit_data(St),
+ {ok, NewSt} = commit_data(St#st{needs_commit = true}),
NewSt;
false ->
St
@@ -763,6 +781,60 @@ set_default_security_object(Fd, Header, Compression,
Options) ->
end.
+% This function is to rollback pf clustered purge.
+% It replaces purge_tree_state and upurge_tree_state with
+% with purged_docs disk pointer that stores only the last purge request
+downgrade_purge_info(Fd, Header) ->
+ {
+ db_header,
+ _DiskVer,
+ UpSeq,
+ _Unused,
+ IdTreeState,
+ SeqTreeState,
+ LocalTreeState,
+ PSeq,
+ PTS,
+ SecurityPtr,
+ RevsLimit,
+ Uuid,
+ Epochs,
+ CompactedSeq,
+ UPST,
+ _PDocsLimit
+ } = Header,
+ PurgedDocsPtr = case PTS of
+ nil -> nil;
+ _ when is_tuple(PTS) ->
+ {ok, PTree} = couch_btree:open(PTS, Fd, [
+ {reduce, fun ?MODULE:purge_tree_reduce/2}
+ ]),
+ {ok, UPTree} = couch_btree:open(UPST, Fd, [
+ {reduce, fun ?MODULE:purge_tree_reduce/2}
+ ]),
+ [{ok, {PSeq, UUID}}] = couch_btree:lookup(PTree, [PSeq]),
+ [{ok, {UUID, IdRevs}}] = couch_btree:lookup(UPTree, [UUID]),
+ Compression = couch_compress:get_compression_method(),
+ AppendOps = [{compression, Compression}],
+ {ok, Ptr, _} = couch_file:append_term(Fd, [IdRevs], AppendOps),
+ Ptr
+ end,
+ NewHeader = couch_bt_engine_header:new(),
+ couch_bt_engine_header:set(NewHeader, [
+ {update_seq, UpSeq},
+ {id_tree_state, IdTreeState},
+ {seq_tree_state, SeqTreeState},
+ {local_tree_state, LocalTreeState},
+ {purge_seq, PSeq},
+ {purged_docs, PurgedDocsPtr},
+ {security_ptr, SecurityPtr},
+ {revs_limit, RevsLimit},
+ {uuid, Uuid},
+ {epochs, Epochs},
+ {compacted_seq, CompactedSeq}
+ ]).
+
+
delete_compaction_files(FilePath) ->
RootDir = config:get("couchdb", "database_dir", "."),
DelOpts = [{context, compaction}],
diff --git a/src/fabric/src/fabric_rpc.erl b/src/fabric/src/fabric_rpc.erl
index 60526f4950..c1bb31b3ac 100644
--- a/src/fabric/src/fabric_rpc.erl
+++ b/src/fabric/src/fabric_rpc.erl
@@ -21,6 +21,7 @@
delete_shard_db_doc/2]).
-export([get_all_security/2, open_shard/2]).
-export([compact/1, compact/2]).
+-export([get_purge_seq/2]).
-export([get_db_info/2, get_doc_count/2, get_update_seq/2,
changes/4, map_view/5, reduce_view/5, group_info/3, update_mrview/4]).
@@ -246,6 +247,9 @@ update_docs(DbName, Docs0, Options) ->
Docs = make_att_readers(Docs0),
with_db(DbName, Options, {couch_db, update_docs, [Docs, Options, X]}).
+get_purge_seq(DbName, Options) ->
+ with_db(DbName, Options, {couch_db, get_purge_seq, []}).
+
%% @equiv group_info(DbName, DDocId, [])
group_info(DbName, DDocId) ->
group_info(DbName, DDocId, []).
diff --git a/src/mem3/src/mem3_rpc.erl b/src/mem3/src/mem3_rpc.erl
index c2bd58fdf6..22d8afb8ac 100644
--- a/src/mem3/src/mem3_rpc.erl
+++ b/src/mem3/src/mem3_rpc.erl
@@ -20,14 +20,18 @@
get_missing_revs/4,
update_docs/4,
load_checkpoint/4,
- save_checkpoint/6
+ save_checkpoint/6,
+ save_checkpoint/7,
+ get_purge_seq/3
]).
% Private RPC callbacks
-export([
find_common_seq_rpc/3,
load_checkpoint_rpc/3,
- save_checkpoint_rpc/5
+ save_checkpoint_rpc/5,
+ save_checkpoint_rpc/6,
+ load_purges_rpc/2
]).
@@ -43,21 +47,35 @@ update_docs(Node, DbName, Docs, Options) ->
rexi_call(Node, {fabric_rpc, update_docs, [DbName, Docs, Options]}).
+get_purge_seq(Node, DbName, Options) ->
+ rexi_call(Node, {fabric_rpc, get_purge_seq, [DbName, Options]}).
+
+
load_checkpoint(Node, DbName, SourceNode, SourceUUID) ->
Args = [DbName, SourceNode, SourceUUID],
rexi_call(Node, {mem3_rpc, load_checkpoint_rpc, Args}).
-
+% Remove after upgrade to clustered purge
save_checkpoint(Node, DbName, DocId, Seq, Entry, History) ->
Args = [DbName, DocId, Seq, Entry, History],
rexi_call(Node, {mem3_rpc, save_checkpoint_rpc, Args}).
+save_checkpoint(Node, DbName, DocId, Seq, PurgeSeq, Entry, History) ->
+ Args = [DbName, DocId, Seq, PurgeSeq, Entry, History],
+ rexi_call(Node, {mem3_rpc, save_checkpoint_rpc, Args}).
+
+
find_common_seq(Node, DbName, SourceUUID, SourceEpochs) ->
Args = [DbName, SourceUUID, SourceEpochs],
rexi_call(Node, {mem3_rpc, find_common_seq_rpc, Args}).
+% stub function for mixed cluster upgrade of clustered purge
+load_purges_rpc(_DbName, _SourceUUID) ->
+ rexi:reply({ok, {[], nil, nil}}).
+
+
load_checkpoint_rpc(DbName, SourceNode, SourceUUID) ->
erlang:put(io_priority, {internal_repl, DbName}),
case get_or_create_db(DbName, [?ADMIN_CTX]) of
@@ -80,7 +98,7 @@ load_checkpoint_rpc(DbName, SourceNode, SourceUUID) ->
rexi:reply(Error)
end.
-
+% Remove after upgrade to clustered purge
save_checkpoint_rpc(DbName, Id, SourceSeq, NewEntry0, History0) ->
erlang:put(io_priority, {internal_repl, DbName}),
case get_or_create_db(DbName, [?ADMIN_CTX]) of
@@ -111,6 +129,40 @@ save_checkpoint_rpc(DbName, Id, SourceSeq, NewEntry0,
History0) ->
rexi:reply(Error)
end.
+
+save_checkpoint_rpc(DbName, Id, SourceSeq, SourcePurgeSeq,
+ NewEntry0, History0) ->
+ erlang:put(io_priority, {internal_repl, DbName}),
+ case get_or_create_db(DbName, [?ADMIN_CTX]) of
+ {ok, Db} ->
+ NewEntry = {[
+ {<<"target_node">>, atom_to_binary(node(), utf8)},
+ {<<"target_uuid">>, couch_db:get_uuid(Db)},
+ {<<"target_seq">>, couch_db:get_update_seq(Db)}
+ ] ++ NewEntry0},
+ Body = {[
+ {<<"seq">>, SourceSeq},
+ {<<"purge_seq">>, SourcePurgeSeq},
+ {<<"target_uuid">>, couch_db:get_uuid(Db)},
+ {<<"history">>, add_checkpoint(NewEntry, History0)}
+ ]},
+ Doc = #doc{id = Id, body = Body},
+ rexi:reply(try couch_db:update_doc(Db, Doc, []) of
+ {ok, _} ->
+ {ok, Body};
+ Else ->
+ {error, Else}
+ catch
+ Exception ->
+ Exception;
+ error:Reason ->
+ {error, Reason}
+ end);
+ Error ->
+ rexi:reply(Error)
+ end.
+
+
find_common_seq_rpc(DbName, SourceUUID, SourceEpochs) ->
erlang:put(io_priority, {internal_repl, DbName}),
case get_or_create_db(DbName, [?ADMIN_CTX]) of
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
