Alexey Serbin has posted comments on this change. ( http://gerrit.cloudera.org:8080/16586 )
Change subject: KUDU-2612: add TxnManager::BeginTransaction() ...................................................................... Patch Set 1: (6 comments) http://gerrit.cloudera.org:8080/#/c/16586/1/src/kudu/master/txn_manager-test.cc File src/kudu/master/txn_manager-test.cc: http://gerrit.cloudera.org:8080/#/c/16586/1/src/kudu/master/txn_manager-test.cc@371 PS1, Line 371: proxy_ > Should be 'p'? Indeed. Thanks! http://gerrit.cloudera.org:8080/#/c/16586/1/src/kudu/master/txn_manager-test.cc@374 PS1, Line 374: > nit: spacing? Done http://gerrit.cloudera.org:8080/#/c/16586/1/src/kudu/master/txn_manager-test.cc@399 PS1, Line 399: txn_initiator(kNumTransactions, nullptr); > Would it make sense to also check that we have kNumTransactions transaction Done http://gerrit.cloudera.org:8080/#/c/16586/1/src/kudu/master/txn_manager-test.cc@426 PS1, Line 426: insert > nit: emplace? I'm not sure emplace() would fit here: that's a range insert. At least, I don't see range std::unordered_set::emplace in the C++ stdlib API: https://en.cppreference.com/w/cpp/container/unordered_set http://gerrit.cloudera.org:8080/#/c/16586/1/src/kudu/master/txn_manager.cc File src/kudu/master/txn_manager.cc: http://gerrit.cloudera.org:8080/#/c/16586/1/src/kudu/master/txn_manager.cc@119 PS1, Line 119: while (true) { : if (next_txn_id_.compare_exchange_strong(stored_txn_id, try_txn_id + 1) || : stored_txn_id > try_txn_id) { : break; : } : } > I don't fully understand this. Wouldn't try_txn_id + 1 = next_txn_id_ in th Ah, I guess there was a swap of the 'try_txn_id' and 'highest_seen_txn_id', indeed. Most likely I messed it up while doing second iteration or so. Good catch! The idea is to make the thread that has gotten a transaction reserved with the highest 'highest_seen_txn_id' so far updating 'next_txn_id_' until it succeeds or bail if there is another thread that received a greater 'highest_seen_txn_id' back from TxnStatusManager. With this approach, max(next_txn_id_, stored_txn_id_) isn't needed since it's done automatically. Is it clearer now with try_txn_id <--> highest_seen_txn_id swap and the comment added? http://gerrit.cloudera.org:8080/#/c/16586/1/src/kudu/master/txn_manager.cc@134 PS1, Line 134: if (s.IsNotFound()) { > Maybe not as a part of this patch, but I'm curious whether you think it's w Yep, good point. I think we can track whether a request to add some particular range has already been sent. I guess that would require tracking the range since we cannot assume that's a single range all over the place because the rate of incoming BeginTransaction request might be high. So, I'd rather put it into a separate patch to 1) add targeted tests as well 2) make reviewing easier. -- To view, visit http://gerrit.cloudera.org:8080/16586 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: kudu Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: I51c476d92bb5b147ffd03fd9f3163ab86d581496 Gerrit-Change-Number: 16586 Gerrit-PatchSet: 1 Gerrit-Owner: Alexey Serbin <aser...@cloudera.com> Gerrit-Reviewer: Alexey Serbin <aser...@cloudera.com> Gerrit-Reviewer: Andrew Wong <aw...@cloudera.com> Gerrit-Reviewer: Hao Hao <hao....@cloudera.com> Gerrit-Reviewer: Kudu Jenkins (120) Gerrit-Comment-Date: Tue, 13 Oct 2020 01:21:55 +0000 Gerrit-HasComments: Yes
