Adar Dembo has posted comments on this change. Change subject: KUDU-1943: Add BlockTransaction to Block Manager ......................................................................
Patch Set 22: (3 comments) http://gerrit.cloudera.org:8080/#/c/7207/22/src/kudu/fs/block_manager.cc File src/kudu/fs/block_manager.cc: Line 45: DEFINE_string(block_manager_flush_control, "finalize", > I don't like the term flush because it doesn't distinguish between a memory Would prefer pre-flush to pre-sync. Also note that this is an experimental flag that largely exists for our own benchmarking, so I think the bar for documenting it is low. http://gerrit.cloudera.org:8080/#/c/7207/22/src/kudu/fs/log_block_manager.cc File src/kudu/fs/log_block_manager.cc: Line 355: // If successful, adds all blocks to the block manager's in-memory maps. > and takes ownership of the LogWritableBlock instances, right? Don't think so; the blocks are just manipulated. PS22, Line 1119: due to KUDU-1793 > That issue is marked as RESOLVED. Is this still an issue? Yes, because KUDU-1793 existed in the wild in at least one release, so we may load on-disk deployments with misaligned blocks. -- To view, visit http://gerrit.cloudera.org:8080/7207 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-MessageType: comment Gerrit-Change-Id: I01f3ed830cc1f86d7bb8a66b2f780914a88884bd Gerrit-PatchSet: 22 Gerrit-Project: kudu Gerrit-Branch: master Gerrit-Owner: Hao Hao <hao....@cloudera.com> Gerrit-Reviewer: Adar Dembo <a...@cloudera.com> Gerrit-Reviewer: Andrew Wong <aw...@cloudera.com> Gerrit-Reviewer: Dan Burkert <danburk...@apache.org> Gerrit-Reviewer: Hao Hao <hao....@cloudera.com> Gerrit-Reviewer: Kudu Jenkins Gerrit-Reviewer: Mike Percy <mpe...@apache.org> Gerrit-Reviewer: Tidy Bot Gerrit-HasComments: Yes