iilyak commented on a change in pull request #1370: [5/5] Clustered Purge
Implementation
URL: https://github.com/apache/couchdb/pull/1370#discussion_r197290389
##########
File path: src/fabric/src/fabric_rpc.erl
##########
@@ -298,6 +314,104 @@ with_db(DbName, Options, {M,F,A}) ->
rexi:reply(Error)
end.
+
+read_repair_filter(DbName, Docs, NodeRevs, Options) ->
+ set_io_priority(DbName, Options),
+ case get_or_create_db(DbName, Options) of
+ {ok, Db} ->
+ try
+ read_repair_filter(Db, Docs, NodeRevs)
+ after
+ couch_db:close(Db)
+ end;
+ Error ->
+ rexi:reply(Error)
+ end.
+
+
+% A read repair operation may have been triggered by a node
+% that was out of sync with the local node. Thus, any time
+% we receive a read repair request we need to check if we
+% may have recently purged any of the given revisions and
+% ignore them if so.
+%
+% This is accomplished by looking at the purge infos that we
+% have locally that have not been replicated to the remote
+% node. The logic here is that we may have received the purge
+% request before the remote shard copy. So to check that we
+% need to look at the purge infos that we have locally but
+% have not yet sent to the remote copy.
+%
+% NodeRevs is a list of the {node(), [rev()]} tuples passed
+% as the read_repair option to update_docs.
+read_repair_filter(Db, Docs, NodeRevs) ->
+ [#doc{id = DocId} | _] = Docs,
+ Nodes = lists:usort([Node || {Node, _} <- NodeRevs, Node /= node()]),
+ NodeSeqs = get_node_seqs(Db, Nodes),
+
+ {ok, DbPSeq} = couch_db:get_purge_seq(Db),
+ Lag = config:get_integer("couchdb", "read_repair_lag", 100),
+
+ % Filter out read-repair updates from any node that is
+ % so out of date that it would force us to scan a large
+ % number of purge infos
+ NodeFiltFun = fun({Node, _Revs}) ->
+ {Node, NodeSeq} = lists:keyfind(Node, 1, NodeSeqs),
+ NodeSeq >= DbPSeq - Lag
+ end,
+ RecentNodeRevs = lists:filter(NodeFiltFun, NodeRevs),
+
+ % For each node we scan the purge infos to filter out any
+ % revisions that have been locally purged since we last
+ % replicated to the remote node's shard copy.
+ AllowableRevs = lists:foldl(fun({Node, Revs}, RevAcc) ->
+ {Node, StartSeq} = lists:keyfind(Node, 1, NodeSeqs),
+ FoldFun = fun({_PSeq, _UUID, PDocId, PRevs}, InnerAcc) ->
+ if PDocId /= DocId -> {ok, InnerAcc}; true ->
+ {ok, InnerAcc -- PRevs}
+ end
+ end,
+ {ok, FiltRevs} = couch_db:fold_purge_infos(Db, StartSeq, FoldFun,
Revs),
+ lists:usort(FiltRevs ++ RevAcc)
+ end, [], RecentNodeRevs),
+
+ % Finally, filter the doc updates to only include revisions
+ % that have not been purged locally.
+ DocFiltFun = fun(#doc{revs = {Pos, [Rev | _]}}) ->
+ lists:member({Pos, Rev}, AllowableRevs)
Review comment:
Possibly can use `gb_sets` here.
----------------------------------------------------------------
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
