no real gain from parallel, series might reduce the cpu load if it finds the doc early
@@ -19,7 +19,7 @@ module.exports = ProjectRootDocManager =
return cb(doc?._id)
else
return cb()
- async.parallel jobs, (root_doc_id)->
+ async.series jobs, (root_doc_id)->
if root_doc_id?
ProjectEntityHandler.setRootDoc project_id, root_doc_id, callback